Warning:
This wiki has been archived and is now read-only.
Krems/Scribe
Contents
- 1 Tuesday Project Meeting
- 2 Day 1
- 2.1 Keynotes
- 2.2 Plenum Talks
- 2.2.1 Open Data and the global enterprise; Thorsten Skalla; IBM Austria
- 2.2.2 An ongoing Open Dialog in an Open Data Ecosystem; Toon Vanagt, data.be & www.openknowledge.be and Noel van Herreweghe, Flemish eGovernment Co-ordination Unit
- 2.2.3 Open Data in practice; Andreas Woditschka; Eversport.at
- 2.2.4 Business models for Linked Open Government Data: what lies beneath?; Nicolas Hazard; PwC
- 2.2.5 Q & A session
- 2.3 Parallel Sessions 1
- 2.3.1 Current achievements and suggested actions on planning and implementing a Government open data strategy; Nancy Routzouni & Thodoris Papadopoulos; MAREG
- 2.3.2 Towards A Sustainable Austrian Data Market; Michela Vignoli; AIT
- 2.3.3 Context-Specific Critical Success Factors for Open Data Publication and Use; Anneke Zuiderwijk, Iryna Susha, Yannis Charalabidis, Peter Parycek, Marijn Janssen
- 2.4 Plenum Talks 2
- 2.4.1 RDB Rechtsdatenbank - Legal database for free research; Peter Guggenberger; Manz
- 2.4.2 The Development of Commercial Weather Services in Europe since 1970; R. E. W. Pettifer; General Secretary PRIMET
- 2.4.3 PSI and Digital Content of Public Broadcasters; Philip Etzlinger; uugot.it
- 2.4.4 Results of Opendatamonitor.eu Survey; Amanda Smith; opendatamonitor.eu
- 2.4.5 Podium Discussion
- 3 Day 2
- 3.1 Plenum Talks 3
- 3.2 Parallel Sessions 2
- 3.2.1 FINODEX. Open data, open for business; Miguel Garcia;
- 3.2.2 Publishing and improving the quality of open data with Open Data Certificates; Amanda Smith & Sumika Sakanishi; ODI
- 3.2.3 University Business Accelerators on Open Data: Activities, Challenges and Best Practices; Yannis Charalabidis; University of the Aegean
- 3.3 Parallel Sessions 3
- 3.3.1 Italian National Guidelines for the Valorization of the Public Sector Information; Gabriele Ciasullo, Giorgia Lodi, Antonio Maccioni, Francesco Tortorelli; Agency for Digital Italy
- 3.3.2 Linked Data Business Cube – Modelling Semantic Web business models; Tassilo Pellegrini; FH St. Pölten, Christian Dirschl & Katja Eck; Wolters Kluwer
- 3.3.3 OpenMove: How Trentino opened public transportation data and benefitted of a mobile ticketing solution for free; Lorenzo Modena; CEO OpenMove
- 3.4 Parallel Sessions 4
- 3.4.1 re3data.org - Making research data visible and discoverable; Robert Ulrich & Hans-Jürgen Goebelbecker, Karlsruher Institute of Technology; Michael Witt, Purdue University;,Heinz Pampel, German Research Center for Geosciences
- 3.4.2 Service innovation: the hidden value of open data; Muriel Foulonneau, Slim Turki; Luxembourg Institute of Science and Technology
- 3.4.3 Extracting Structured Data from Unstructured Open Data; Uldis Bojars, Renars Liepins;
- 3.5 Bar Camp Sessions
- 3.6 Project Meeting Friday
Tuesday Project Meeting
Scribe: PhilA
The process is that we'll have to analyse the documents that we have into smaller parts. We tried to do X, this is the way we did it.
PW: We have to look at what we have already. Some are straightforward, some are longer, with multiple stories. We want a bit of good practice that focusses on a small number of elements of the PSID. People will be audited on whether/how they have followed these BPs.
... We also need to identify where we don't have sufficient guidance so we can build on that.
... We need to simplify the story. So it can be given to managers "this is what our org needs to do" and tech people to be able to say "yes, we did that"
... So there's some distilling and pruning. And then stand back and see what's missing.
MD: We haveb't decided how we're going to move this forward. We need a group of people (5 - 8) so that we have a common way of working. So we need a task force with bi-weekly conf calls etc.
MD: By end September, we need to show how we're progressing. By end of the year, we need to be pretty fixed in the BPs so we can focus on the localised guides in the final stages.
MD: We need help.
MD: One BP might be how do you find a licence - currently buried in some of the docs we have.
... We really have until September to reformat what we have.
... That group can, of course provide input on how we're doing things.
YC: Did I get this right... you want to assemble a team to analyse the good practices and giving them some categorisation.
MD: We've made a start on that
PW: And also to rework them -
YC: So why don't we go back to the authors. Let's say I analyse your short story and I'm trying to understand a few things.
PW: Yes, the authors need to be part of it, but it needs to be driven.
PW: It's a team effort. Everyone has to be involved, but a small group has to take responsibility.
MD: If you look at Johann's BP (for example) you get the outside view and then Johann can see if you understood the message.
YC: I don't propose losing the editorial team. What I could do though, is answer an e-mail that asks me to do some categorisation - a few clicks. That I can do, I won't re-do my case.
YC: We made a decision in Timisoara not to have a controlled vocab for every step. Then we might come back with classification. I don't think I coujld do that for myself.
NvH: If it needs to be applicable to multiple states then you need to make it transferable etc.
MD: Did I hear you say you'd be willing to work on this?
YC: I'd be happy to work on the classification scheme. I can extend what Peter has started to make a form that just needs someone to click no more than 15 times to confirm our understanding.
YC: then editors can do things like name clouds.
NvH: You'll have to tranbslate it into your local colour
BK: I wonder whether we should have a paper-review like process. As you have done, we can define what a BP has to follow, and then we can assign reviewers and the review can be sent back to the authors for adaptation.
PW: The more shuttling around, the longer this will take. I think it's important to get through the work. I'm worried that this could take forever. We need people to give info in. We currently have 30+ BPs. If we get those in, then we need to crank through, look where the gaps are etc. We need balance. So I think it's better that we get the job done with a small team.
... Of course the next issue is that we're heading into holidays. Sending documents around the place now will fall into the holiday hole.
MD: But we now have 2 proposals. We have a small group doing the reading and classification, and then the volunteers is all of us and everyone gets a few. Maybe we put the reviews to everyone and then name and shame people who don't.
SS: How about we do it in the workshop. working together. It's harder to do it at home - better while we're all together. So maybe tomorrow, we stay 1 or 2 hours longer tomorrow and crank through them.
PW: I think maybe Sebastian is suggesting if we have a room... we could bring the pizzas and booze in and work through it.
JH: Maybe we do some on Friday morning.
General discussion about how to advance the BPs while in Krems.
EB: How about putting a maximum word count on each element of the BPs.
PW: You need to get all your notes together and then edit it down.
Fatemeh: What sort of format are you looking for? Free text only? Semantics? We can run some text analysis.
PW: You could take the raw text from the wiki. It would be good to do that and see how things are partitioned
FA: If I can use tools then you can sign me up.
Summary
Bar Camp - we can look at the classification that Yannis was talking about. Come up with a reasonable set.
And then on Friday morning we can take the Krems BPs and apply it.
PA: raises issue of showing agreeemnt, MD thinks it's not a problem as we will agree among ourselves for each one.
More Discussion
CH: raises issue of consensus vs voting ... Consensus takes longer but is a better route (used by TOG)
- Scribe change to Benedikt
- Discussion about Best Practices Consensus process
- Nancy: It is not realistic that we all give consensus to each of every best practice online. Rather have a session where someone presents the best practice and we discuss and find consensus.
- Phil: Question: Are we doing this as part of workshop or project meeting?
- Makx: Half a day extra for workshop meeting to do exactly what Nancy says would be very useful.
- Johann: I would suggest, first pick those point thats have to be done and include it in the workshop. We should sacrifice part of the workshop for this, since such best practice presentations would be interesting for everybody.
- Phil: We need to do this before September.
- Makx: We only need to show something until September (for the review), until the end of the year.
- Phil: It seems we have consensus: At Berlin, we will have two sessions during the day of the workshop. We say wich Best Practice (bp,BP) we have come up during all workshops (WS) so far. At discussions, we are looking for consensus about the BPs.
- Chris: Will we have more BPs afterwards?
- Phil: Only those that come out of Berlin.
- Phil: How do we know that we have all that we need to cover?
- Makx: We cannot be complete. We do our best.
- Phil: Timing is crucial. When should be our two sessions? How do we make it intersting to people?
- Heike: How about a "speed" run.
- Phil: Could we use the timer. 5minutes per bp, two slides per bp.
- Sebastian: There are mobile TED devices (voting devices). We could lend it for Europe. Could speed up voting process.
- Johann: Would overcomplicate things. We have not decided whether other people are allowed to vote. Why not have a tool afterwards where only project members can vote afterwards?
- Nancy: Why do we need a voting at all? We could collect feedback during the session on the wiki.
- Heike: We could ask questions such as "would it be applicable to your country?", "is therey anyone who thinks this is really stupid?".
- Jan: Agrees to gather both positive and negative feedback during the slots.
- Peter: Would it be useful to add anti patterns to the mix?
- Jan: There is a risk to spend to much time with discussion.
- Nancy: We could limit discussions. E.g., only yes/no questions.
- Yury: I do not believe that there will be negative comments.
- Makx: Identify things that could be problematic and to follow-up on it separately. "If BP is not passed, then either dropped or discussion follow-up separately".
- Noel: Most intersting question: "Applicable or not applicable?"
- Johann: Expect a lot of overlap between BP.
- Ingo: Do not think it is a problem to have overlaps between BP. Even to have partly contradicting BP is ok.
- Phil: OGC, Open Group, can you describe the OGC process?
- Ingo: At OGC, someone stands up with standard. Audience has time for three weeks to vote NO (but have to give arguments).
- Chris: At OK, 75% of members have to be positive.
- Phil: At W3C, while someone has something to say against, it is not adopted (if that takes too long, TBL decides).
- Johann: We could start at Thursday and have first 5min about BP (pitches, no full-fledged BP).
- Ingo: 40 BP per 5min will take some time. People should collect objections.
- Nancy: The template/structure of BP presentation (two slides) is important.
- Emma: Concerned about effort put by single persons in several BP, forms + slides. Problem that no feedback in between.
- Johann: Do not thing that any BP will be rejected.
- Phil: Do not think that anyone will present a BP that we are not happy about.
- Noel: I think we have to be careful about the term Best Practice. Not about "crap", they are asking about something that has been done in one country and we are voting whether it can be applied in other countries.
- Phil: Should we change the term?
- Noel: No. We should only not overemphesise the name.
- Phil: Who volunteers to be part of the task force to have two-weekly calls (not always obligatory) to do some editing work about the best practices. Also coordination part. Drive that process.
- Emma: yes.
- Nancy: yes.
- Heike: yes.
- Jan: yes.
- Valentina: yes (only technical part).
- Fatima: not sure.
- Phil: yes (of course ;-)
- Coffee Break
- Phil:
- Please give feedback to CFP for next workshop.
- Passing through attendance sheet (every sign once)
- Carola was not happy about attendance at some workshops by some partners.
- If noone is present of a partner at a meeting, EU might not give all the money.
- Slovaks have only participated in one workshop so far. Maybe coming to Krems.
- Noel will be take video. Please tell him if he should not record you.
- Pictures: At Timisoara there were pictures, but unfortunately not too great ones.
- Please take pictures at Krems!
- Phil presenting some people
- Fatima from Insight
- Sebastian from init
- Yury from Fraunhofer Fokus
- Bernata from Slovenia government
- Micha from Open Knowledge, Web Foundation
- Nikolas from PwC
- Name Game :-)
Day 1
Keynotes
International Challenges to Transformational eGovernment
Shauneen Furlong
Talked about using tech to change business world, government world. We have made great strides but there is a lot more for us to do.
eGov quenched thirst for online public services. Service delivery has been successful but we have not embraced this fully within the public sector.
We are more transactional that transformational. It's easier, safer. Based on known parameters. Has current business processes behind it. Non-threatening to senior executives. Those are the people who determine what eGov strategies will advance, so they will have a disincentive to do something that will affect their careers.
We thought eGov would be the catalyst to reform the public sector when we started in 2000. 2014 was the first time that every UN member had a Web presence.
Transforming government is harder. Working across departments, systems and cultures. In Canada.... we picked a small group of deputy ministers, the heads of departments - there were 12-30 of them - and we said we wanted to make Canada the most connected government in the world, they'd never met before. It's easier if you just stick within your own four walls.
Gives example of shopping malls where shops have a great deal in common but they don't communicate or interoperate, yet we're expecting governments to do this.
We're asking departments not to take what you did before and automate it. Transformation says do something new.
We had a challenge called 'lost wallet' - how do you get back info from all the different departments that you had stuck in your wallet. We haven't managed it yet - it would be transformational as there's no single driver/department.
What impedes our progress?
Perhaps we should drop eGov and see government as a digital state.
Talks about the importance and impact of comparative studies. gave example of Turkey that went from 190 to 180 which helped secure more funding. (Ref session in Timisoara.
Current success KPIs for individuals don't support transformation.
People are risk-averse and transformations are risky. You don't want to be on the front page for a failure.
The business processes we're following are more or less the same now as they were 10 years ago.
Singapore's first project was to do something that you could not do before. The biggest problem they had was traffic and parking for appointments. So they linked doctor's appointments to parking spaces. That's a great example of creativity and transformations.
Gives example of Amazon drone delivery as being creative use of tech.
Academics are not yet connected in this process. In Canada we wanted to be the first in the world to make it happen.
My research on transformation... through working with an organisation in Washington - what are the challenges, what are the impediments, how do you say "who else in the world is designing an e-health card, Maybe I could talk to them before I design mine? In Ontario, it's one of our failures. We don't have an electronic health card, but maybe we can have a shared repository.
One way out of this might be project management. The PM processes do not currently meet the needs of a transformational eGov environment. Our own business process world doesn't have the tech to make it happen. You're busy reporting to someone else who is not involved in delivering that project. You're serving the wrong master. We need a PM methodology that can work with our organic, fluid world.
- The role of PM as a science. Not using PM as the checklist, use it as a tech enabler to bring the project to life.
- We need a PM environment that it is informationally enhanced. The PM has no way of accessing the lessons learned, the different stakeholder interests. Imagine saying to a soldier and en environmentalist and a doctor that they had to share info.
- I did surveys etc. to find out what is holding you back - why are you not using tech. You use it at home and your kids use it - how can we make a difference. We came up with a set of 10 holistic challenges, ones that apply across the board. Synergistic because they're interlinked. IT support was not involved in the decisions and planning, it was a separate department.
Slide shows the 10
No 1. Conflicting stakeholders. We're asking very different people to agree. We're demanding specificity in an IT system for something that doesn't exist yet. How long did it take you to get to work in the morning - oh 20- 25 minutes. But the system says "it will be 20.3 mins" - which is farcical. If you don't meet something specific then you get criticised for it.
No 3. Outdated business models. These are the ones that say if you accurately forecast what will happen with great specificity you'll be rewarded. We should have business models for being more creative.We need a right to fail.
No 5. Lessons learned. How can we stand upon the shoulders of others, not meeting accounting steps. We haven't made progress on this. There's no way to share lessons learned internationally.
No 8. Project manager as the driver. One outstanding challenge is collaboration.
Q about reluctance to give up power.
A. Managed to get Canadian Prime Minister to agree that the individual who managed to do something transformational across several departments would be personally promoted.
Q. Did you research ROI? It's hard to find studies to quote.
A. Orgs like the UN, World Bank etc. have indicators. Those benchmarks are the ones I'm objecting to for transformational eGov environment. One of the criticisms I have when people say if you do this as an eGov service, you will save money. That's automating an existing process. Transformatioanl means that the ROI is different. How do you measure ROI for a Singapore citizen who can find a parking space when you go to a doctor's appointment? We don't have a way of measuring that. In Canada, we had a programme around veterans. There's a page showing who was killed in service when - that's something you can see in parliament if you visit. We developed a means of people seeing the pages of that book changing - remember how big Canada is. What is the ROI on changing our anachronistic models?
Wazing the Information Super Highway: Linking the World's Open Data Resources
Alon Peled
Begins by showing video
Regrets saying negative things about open gov data... glad I overcame misgivings. I needed a worldwide catalogue of the open data assets from municipal to supranational levels. Someone must have developed such a catalogue. Started with things like data.gov. Socrata and Juno have alternatives to CKAN etc. Not going to do what I want. The OD Barometer was of limited use as it only touches a few datasets. Best one was the European ENGAGE project. But I still needed more data from more places.
So we created a catalogue of our own. Software crawls the Web for OGD and when it finds it it extracts metadata (shared of Peter W's session in Timisoara)
Shows screenshot of PSIE database
Showed this to NASA. It shows different things about the corpus as it grows each day. This isn't human-mediated. It works like Google. It showed some mistakes.
I wanted to know what is the oldest open data asset. We had a result from 1833 which we thought was a mistake but it wasn't (it was the Lewis and Clarke mission across America).
In list of who has released most data, Canada comes out top.
Shows search function that returns datsaets about the search term, not things like wikipedia pages. Works across languages.
data.gov turns out to be a story of individuals, moving from one department to another to get data published. Are public authorities actually engaging or are they just paying lip service to the idea of open data.
Rather then relying on the metadata as the source, we should see it as a clue to help software do the processing and make sense of the data.
Now look at what the data is actually about, not what the officials say it's about.
Talks about breadth of data, not just quality. i.e. different domains, not just a few. The visualisation shows that countries on the graph go up, then flatten, then go up to top right of the graph.
Set up challenge to add new non-space keywords to NASA data. Data Treasure Hunting.
Biggest problem in my view is... agencies don't like OGD because they know it's worth something so they don't want to make it available for free. if you force us to open data for free, will pretend to do it but will keep some data back. Problem of recording usage of data. In his view, data that isn't free is also open.
Plenum Talks
Open Data and the global enterprise; Thorsten Skalla; IBM Austria
scribe: PhilA
Thorsten is in the public sector space. Here at request of corporate peers in the US (we're looking at you Steve Adler!)
Slides have the details.
Shows tag cloud... mostly soft factors, not tangible business benefits. No numbers. So what is the value of data? It has no value by itself, especially if no one is using it.
Makes analogy with mobile Web 10 years ago. Hundreds of different devices. Mobile internet existed but had little use. Nice to see reference to Mobile Web Best Practices work
Calls for standards to make datasets comparable.
Talks about Digital Delta project. They don't publish all the data that they could in case people are more scared of flooding than they need to be.
Talks about need for consistency in referring to locations. Portals always have a restricted view of data, hard to compare data from different sources.
Refers to BBC use of LOD.
Austria's open data portal. IBM acted as a launch partner and provided some datasets. We would have liked to have published more but the same restrictions apply to IBM as to public sector - want to make money out of it.
Hope examples show some positive use cases. As indicated, I don't have any hard numbers. Easier to find example where it has benefited society.
IBM setting up a Chief Data Officer?
Repeats that there are no relevant standards for comparison of data.
An ongoing Open Dialog in an Open Data Ecosystem; Toon Vanagt, data.be & www.openknowledge.be and Noel van Herreweghe, Flemish eGovernment Co-ordination Unit
scribe: PhilA
A serial entrepreneur and a civil servant.
Toon is the businessman, Noel can talk about the ecosystem
Noel revisits the 4 sectors. Wasn't sure it would work... didn't want a Flemish gov entity to say they were being forced to do stuff but there are no tools, support, infrastructure etc. Had to make sure there was a plan (see Samos).
We had a framework, we had standards like DCAT...
Toon...
Name data.be doesn't restrict what we can do.
Hard to find business data in Belgium.
Shows current Belgian (not Flemish) government websites (marbled background), no auto complete etc.
So data.be scrapes the sites, which we can do, and curated the data. Using things like OCR on PDFs.
One way you can make money from OD is to refine the data, e.g. with text mining, named entities. Created some REST APIs
It's very hard to make money from open data.
Spoke to an agency who were saying that since they had 7 customers paying €100K so that marginal cost was €100K. Some saying, Ok, we are PSI compliant - sue us if you disagree.
Government is our biggest customer. Mini Finance etc. use our services - based on their data. Thousands of gov people using our service without having to pay for it.
Opening up doesn't immediately monetise. Need to see the money/benefits all the way through the chain.
Stories from the trenches.
Lady wanted to buy a castle, cost €11M. Bought it - didn't like this being public as she'd created a legal entity to buy the castle -> became public.
Toon had his phone number on the website - got good feedback.
There was a shelter for vulnerable children. The address was in official docs that were online - which was a problem of course.
Parents not wanting their children to know they'd sold a company for lots of money.
Noel refers to Joel Guerin's work on the open data compass.
Noel: Asked people like Toon what the issues were, problems faced etc. See slide:
- Demand side needs to be involved
- data isn't interesting
See slides.
Talks about future Open data day, round tables etc. Inviting many stakeholders, including citizens' groups, industry, public authorities. Planning to continue these ecosystem meetings every 6 months.
Toon highlights funding like FINODEX
Open Data in practice; Andreas Woditschka; Eversport.at
Scribe: PhilA
Explains what Eversport is. Want to be able to book a tennis court as easily as booking a hotel room.
No online marketplace. Sports people don't know where to look, facilities providers don't know how to use IT to do it.
Eversport has platform for searching for facilities - tennis courts, camps etc.
Business model is quite simple. See slides.
SEO pages for customers as big benefit
Content strategy ... take open data where we can find it. Vienna had good CSV file - nowhere else.
3rd party data - some info is missing, like opening hours, look at the customer site and add it. look for tennis is Vienna, we're top of the list. That takes effort.
No open data in large parts of Austria. Where we found it, it was old and inaccurate. No commonality of data types.
DBF? QGS??
Came up against local coordinate reference systems.
Lots of tools that you can use for free. There are issues handling Excel. Open Refine and Notepad++ are important.
Eversport has a vision. Wants to include weatehr data for example, so you can see what the forecast is for the court you want to book.
Business models for Linked Open Government Data: what lies beneath?; Nicolas Hazard; PwC
Points to original doc on Joinup.
Slides are informative - and offer a full summary
Q & A session
Toon emphasised need for involvement of potential users of the data. Work bottom up, not top down.
NvH always being asked for results - show me who is using the data.
Parallel Sessions 1
Current achievements and suggested actions on planning and implementing a Government open data strategy; Nancy Routzouni & Thodoris Papadopoulos; MAREG
- scribe: Martin Alvarez
- chair: Nancy
- 50+ attendees
Nancy: [Introductions]
... I work for the Hellenic Ministry of Interior and Administrative Reform.
... We have many ingredients to define an Open Government Data strategy:
... (1) Legislation, (2) technology, and the the most important to me (3) culture.
... We will discuss these three pillars during this workshop.
... [Showing the PSI scoreboard] We are proud to be in this good position (Greece is 4th position)
... This scoreboard shows that Greece has low score in some specific indicators
... [Showing the OK's Web Index] (31st position)
... Another important aspect for us is corruption, difficult to measure.
... [Showing a slide with the index of perception of corruption in the world] We see that Greece is at the bottom of the ranking in Europe. So we should focus on this.
... [the Open Government ecosystem]
... All gov't institutions have to publish their decisions under the program @clarity.
... 4221 public bodies publishing information in this platform.
... 60000 documents uploaded in the platform in order to get the transparency in public administrations in Greece.
... Many companies are using this information of the transparency programme.
... We believe that this policy is really needed for get over this crisis we have living during these years.
... Those interested parts will use the information to their own purposes.
... We believe in Open Data and we try to encourage the use of this PSI.
... Greece is also committed to the OGP. In Oct 2013, our Prime Minister announced publically that Greece will apply the right to reuse PSI.
... This sentences from the top politicians are really relevant.
... We have organised the Open Data last year. Phil attended and gave a talk. (Promotion)
... Oct 2014 we voted for the law 4305 -> act that states the right to reuse by default.
... This law is also the implementation of the EC 2013 Directive on PSI reuse.
... We have cooperated with a dozen of multi-stakeholders to deliver this law.
... We organized a couple of workshops (Mar Apr 2014) to deliver two policy policy papers.
... The process was completed with a public deliberation on the right to reuse by default, as well as about the Law 4305/2014.
... Changed the model from 'request information to be reused' to 'open proactively' the information and enabling the reuse by default.
... Greek Public Administrations will have to open all the information unless they prove that it cannot be released (and notify the reasons) .
... We are prioritizing the datasets to be released. Citizens can suggest and vote on datasets using a consultation platform.
... With all the obligations that Public Administrations have, we are trying to provide incentives for those that follow our recommendations.
... We have a hierarchical network of 4000+ public bodies that are responsible to implement the new Open Data policy in the public sector.
... We know who is the owner of the implementation of this policy in each public body.
... We organise an annual hackathon on Open Data, also an award to promote and give recognition to the "best" users.
... The cooperate with the Open Dada Support project.
... Now is time to Thodoris. Thanks for your attention.
Thodoris: I'm a programmer working for the Greek Government.
... I built the Open Data portal based on CKAN.
... The Transparency initiative in Greece is the fist Open DAta initiative in Greece.
... Since Oct 1st 2010, all gov. public bodies are obliged to expose their decisions on the web.
... There is a institutional framework that fosters transparency and enables public participation [slides]
... Now, the acts are only valid once are published. The online version is the official one.
... In June 2014 we released the new Open Data portal , including accessible features to include all users.
... More communication channels to enable a dialogues with potential reusers.
... This is important to monitor fraud and fight corruption.
... [who is publishing] Different phases in time.
... [what is published] Legal & regulatory, financial, organizational and administrative, environmental information.
... [how is published] We classify all the documents according a 50-topic scheme. We define special metadata and define specific vocabularies to represent the information.
... [slide with screenshots of the civil servant's backoffice to upload the information and classify it]
... [Architecture of the platform] All the infrastructure is based on independent APIs.
... We collected up to 16m decisions since the programme started!
... Some features of the new Open Data portal: we implemented a faceted browser, there is versioning of the documents/datasets.
... The portal enables users to pinpoint issues on the documents or metadata.
... Open REST APIs to query the catalogue of decisions.
... It offers social network communication/login.
... 30 applications were created after the release of the API. For instance some visualizations shown.
... We use open technologies, also licensing (CC-BY), also an open architecture based on Java, Ubuntu,
... Thoughts: technology is important. The law does not state more than publication (not adding metadata) <- we added this.
... Some parts need to include technological aspects such as digital signature. So we have to provide mechanisms for that.
... This will stimulate transparency and creating more value from data.
... Crowd sourcing is really important so we should believe in it.
... Thanks
Noel: Our Open Data policy...
... PSI Directive is a positive thing, but if you read it carefully... The legal people detect many gray areas.
... We have problems with scores and indexes. All are referred to Belgium. Flanders is not Belgium and vice versa.
... The Belgium Federal Government is the one involved in the EC conversations but not Flanders.
... Another challenge is the special arrangements that GLAM institutions are able to make. There are many exceptions that can be applied to public bodies.
... What I did was creating a binding agreement (that all ministries signed):
... (1) Having regulation (aligned with the PSI Directive), some bindings needed.
... (2) Marginal costs: define what it is, that implies investment. Also we should talk about the good-quality service we provide.
... (3) We should define an ombudsman figure.
... All public bodies believe that they have special cases. Academia is even worse with the Open Access.
... Licenses are based on standard licenses (CC-BY but we had to identify the source of the information).
... Also the agreement have to avoid discrimination or exclusive agreements.
... 10 years is too much as date to review the exclusive agreements of the GLAM institutions.
Phil: What does "machine readable" means for you? Is this regulated in Greece and Flanders?
Noel: No. No regulated
Nancy: Very difficult to define.
Noel: The worst thing is that the PSI Directive says: "*preferably* in machine-readable formats".
Noel: ¿Questions?
Phil: Publishing decisions is good, but court decisions are better.
... If I had time, I would write something to the EC to encourage governments opening up court decisions. Who breaks the rules.
Nancy: Actually, we've tried to open the court decisions in Greece. It was not a data protection issue, but it wasn't approved :-(
Ales Versic: On the importance of data quality.
... I'll show you our infrastructure in Slovenia, valid for all "e-System".
... The information in the system could be used for internal analytics and also published.
... We need to control the quality of data. (attributes, geoposition, etc.)
... Public bodies can take good/bad decisions depending on the data quality.
... People can provide useful feedback but improvements have to be notified to the right person.
... We can use an ISO standard to measure the quality of the data.
... Is there an obligation for public bodies to provide good quality data?
Nancy: No. It's not our case.
Ales: [An example of usage of open data represented on arcgis.com]
... This example shows meteo information from various sources showing different layers with different values (a bridge shown in three different locations)
... Presenting examples of the meteorological portal of Slovenia, geospatial datasets shown on maps, also in WFS formats.
... Other example of GIS: http://gis.arso.gov.si
Noel: Anyone heard about data maturity model?
Amanda: We (ODI) did it.
Noel: An example: Flanders bus company releases data on bus stops.
... I received a spreadsheet from a guy saying that hundreds of stops were not well geo-located!
... Who measures the quality of data before releasing it? [a couple of hands arose]
Towards A Sustainable Austrian Data Market; Michela Vignoli; AIT
Scribe: Daniel Pop
Michela - presents the outline of the workshop: there will be 3 presentations + open discussion
Firstly, Michela gives an overview of Open Data
- What open data is
- Open Data sector is under development
- Entrepreneurs hesitate to enter the market
- Austria progress a lot as latest ePSI-Scoreboard shows
- Some slides from Big Open Data Report are presented highlighting predictions of the potential growth of GDP thanks to Big and Open Data as well as the current state of things
- An Open Data Market Framework would regulate on different directions
Gregor Eibl, BKA (Bundeskanzleramt Oesterreich)
Developments of and Best practices from data.gv.at & opendataportal.at
- On data.gv.at
- 1600 metadata sets and 271 apps
- 2014: 117,000 unique clients (+67% on 2013)
- contain only meta-data; datasets remain on their original portals; this is a good practice as it makes it easier to publish the data
- the portal incorporates metafiles from local OGD portals
- Good practice: use of open-source software for providing the service (CKAN, Wordpress)
- The portal won the UN Award for the whole national portal integrating different local sources
- New features of the portal are presented (coming from CKAN platform)
- On Open Government Documents
- License: CC BY 3.0 AT
- On the pilot for Linked data (linkeddata.gv.at)
- data from data.gv.at converted to RDF and enriched by 3rd party developers
- SPARQL endpoint
- usage examples are missing for now
- On opendataportal.at
- target groups: business, ngo/npo, research, community
- cooperation between OKFN, Wikimedia Austria, Cooperation OGD Austria
- same rules as for data.gv.at (metadata, CKAN catalogue etc)
- opening up business data is even more difficult: it is more probable that already available data will be made available more structured and licensed, but new data will be difficult
- on governmental level it will be necessary to open up more data due to the PSI directive
Andreas Woditschka, Eversport.at
Developments of and Best practices from Eversport.at
- sport facilities
- some information from opendataportal.at, but enhanced with 3rd party data + Own data
- Eversport is one of the apps listed in the presentation before
- no data about sport facilities available from most regions in Austria (Lower Austria, Upper Austria, Styria, etc.)
- technical limitations
- different data formats and sources
- ETL Process, Regular expressions
- Tools: google refine, notepad++
- mashing-up with other Open Data providers: they will look into integrating dynamic data (weather data)
Karl-Heinz Leitner, AIT Open Innovation: Oportunities and challenges
- users are important innovators
- Henry Chesbrough (2003): Open Innovation
- innovation-futures.org project collected success stories, scenarios for open innovation
- Open Source Society (OSCAR = Open Source Car)
- Open Innovation and open science (E.g. PatientsLikeMe, 23andme, whole brain catalogue etc)
- OI offers huge opportunities for business and governments
- New challenges:
--- how to involve citizens to use this data (incentives and motivation).
--- what is the optimal level of participation?
--- how to control the processes? A thread for decision people (governments).
--- companies still have to define their core competencies and protect their knowledge
--- business model is key for success, so usually are not disclosed.
--- new, hybrid business models are necessary (3P model: People, Profit, Planet)
- Business models by Gassmann and Osterwalder
- Open Innovation, an evolutionary process
- increased competition between private, public, ngo, etc
- rif2030.eu, innovation-futures.org
Open Discussion
- Open Source Society --- 53% of Austrians are providing their energy for free, i.e. voluntary work
- Johann: a distinction between tangible and virtual goods is needed: digital goods are just marginal costs
- people are eager to collaborate
- Patents
- Karl-Heinz: successful companies who provide free services still have lots of secrets (e.g. design of new Apple products, Google search algorithm)
- Karl-Heinz: new monopolies
- Karl-Heinz: investments need to be paided back
- How much costs to open datasets in Austria? A: Gregor: we don't know exact investment and it’s difficult to asses
- Gregor: pointed out that data can be re-sold
- Johann: in 2013 -- a study showed that a considerable effort has been invested in the creation of the data.gv.at portal. Nobody knows exactly because there were no direct, specific reporting of people effort working on opening-up data tasks. In Austrian case, the already well-established ICT infrastructure eased a lot opening the data. In other countries, the costs of ICT infrastructure may be a considerable cost
- Karen Halford: private sector to improve the open data? A: In case of eversport.at, Vienna tennis federation benefit of the data from eversport.at and updated their own database with new facilities
- Karl-Heinz: for the 271 apps on data.gv.at, it would be good to have an evaluation of these apps: are just for fun, are a start for start-ups, what business models they use?
- In case of Eversport.at, initially there were business angels and money coming from City of Vienna; now, it has 10 employees and it’s planning to ‘attack’ the German market
- Is important to find the market and how to address the market
- Adding services govs cannot provide (adding value to data); this provides business opportunities as well as for outsourcing
- Contribute back the data to governments (for money)
- Gregor: The BKA thought about that, but as it is no normal procurement process this is a legal issue
Michela summarizes: - it will be more about selling services than selling data
- money will not come directly from users, but from other sources (re-distribution of costs)
- rethink the capitalistic approach to develop new models, to shift around costs to other places
Peter Guggenberger; Manz (Publisher of legal papers)
- MANZ Publishing house is working with open data since 87
- in 97 the government decided to free data for the market. The first reaction was: protection! But in the end they had to accept it, and they are still in business
- What do the customers want? What do they need?
- If something is free, doesn't mean that we cannot earn money
- Open data helped us to transform the mindset
- Clients accepted it
- Demand is driven by the customers
Context-Specific Critical Success Factors for Open Data Publication and Use; Anneke Zuiderwijk, Iryna Susha, Yannis Charalabidis, Peter Parycek, Marijn Janssen
- Scribe: Benedikt Kämpgen
Executive Summary
Critical success factors for Open Data implementations were presented:
- implementation = any efforts to improve Open Data publication and use.
For Publication:
- Legislation, regulation, licenses
- Strategy and political support
- Management support
- Training of and support for civil servants
- Sustainability of the open data initiative
- Collaboration
- Open data platforms, tools and services
- Accessibility, interoperability, and standards
For use:
- Legislation, regulation, and licenses
- Success stories
- Training of and support for open data users
- Feedback and sustainability
- Research and education
Questionnaire about such factors and aspects were participants assessed the relevance of CSF in their open data projects.
People had difficulties to understand:
- "success" - "having a lot of stuff out"
- "Open Data" - license, public sector information, scientific data
- "publication" - does it include the curation process? What difference to release.
Nevertheless, very important discussions (one said that now after the discussion he knew why he had difficulties) and one best practice can be identified:
BP: Check which critical success factors fit to your Open Data publication or usage project and consider them.
Detailed Minutes
- Attendants: 20-25 people
- Idea of this workshop was to collect crucial success factors.
- Question to the audicence: Who are practitioners? Who has been involved in publishing, consuming Open Data?
- CSF = success factors for OD implementation.
- OD implementation = any efforts to improve Open Data publication and consumption.
- Factors in 9 categories of open data publication
Publication:
- Legislation, regulation, licenses
- Strategy and political support
- Management support
- Training of and support for civil servants
- Sustainability of the open data initiative
- Collaboration
- Open data platforms, tools and services
- Accessibility, interoperability, and standards
For use:
- Legislation, regulation, and licenses
- Success stories
- Training of and support for open data users
- Feedback and sustainability
- Research and education
Contextual-aspects
- geo-graphical level (regional, international)
- type of published data
- stage in the development of open data initiative
Questionnaire is handed out
Discussion is started
- Yannis: Question to the audience: What are the important things in publishing/consumption of Open Data. Currently, we are creating categories.
- Makx: Important dimension using in the framework: Obligation of public sector to open up data.
- Makx: Missing one word in the framework: Cost/Funding
- Frank: Had problem to understand the meaning of "success". Also how confident are you with your assessment.
- Iryna: Is discussed in the paper.
- Frank: Other people would define with "having lots of stuff out".
- Morton (Denmark): Having civil servant hat on: Data concept. Core data, and data that cannot be released, and then there is Open Data that is not limited by the legislation.
- Some (health information, income, corporate registries) data would put additional costs on institutions/public sector to make it openly available; for instance, personnal related data that would first be need to be anonymised.
- Yannis: Is there a hidden success factor here?
- Morton: The question is, from which tax is effort for anonymisation paid?
- Alon: Incentives for data open use. For whom should be the incentives?
- Teresa: I have a problem with understanding "publication". Is publication releasing/preparation/curation? Should be specified more. I had problems with some of the questions in the questionnaires. Would you include visualisation as curation? As a survey to go to many people, be clear about the terms.
- Yannis: Depends on the life cycle that you use.
- Robert: What is the relationship between Public Sector Information and Open Data?
- Proscovia: Creation of data is not sufficiently considered. Do we need to include more information about the process of how data was captured and created and managed for reuse?
- Proscovia: Is there any other critical success factor?
- Yannis: We assume public sector HAS certain information. What about if data is required that is not collected, yet (e.g., putting sensor data somewhere).
- Heike: Different between transactional and transformational processes of publishing data from the speaker this morning. At the moment data is driven by internal processes and less of external organisations.
- Petyr: Take the example of sensor data. Helpful would be to include the publication processes in the projects.
- Makx: That people that create the data do not know about Open Data. But they should. CSF: Take away to explain Open Data but instead give them the tools at hand so that it is done automatically.
- Petyr: XML document.
- Yannis: But this it not was ? said. Think more about what you want to create? Of course much more difficult.
- Teresa: Data management capabilities are really important critical success factor.
- Petyr: Similar to skandinavian application at Samos showing Green/Yellow/Red visualisation.
- Teresa: We are talking about insufficient architectures inside the public sector. Less about attitude more about capability.
- Victoria: Agree with Teresa. Different epistemic communities. Legislation still often a problem.
- Yannis: Some countries have a legislation being more concrete than others.
- Yan: There needs to be someone at the organisation to drives the publication process; a kind of Open Data Coordinator. Example was Czeck telecomunication office.
- ?3: I also had a hard time with the questionnaire, but now after the discussions I know why. What would be the phases? Who are you addressing? Which of these factors were important to you, and which weren't?
- Benedikt: Can you give a wrap-up of the session towards best practices?
- Yannis: Idea would be to have one BP from this work: "Identify critical success factors in your project.".
- Makx: "Use these critical success factors. And check which are important for you."
Minutes by the organisers
Notes from CeDEM workshop 28 participants in the workshop (including us), about half of them considered themselves to be practitioners (mainly those from the PSI-share project). About half of the people were from PSI-share. Most people had worked with open data. When we started distributing the questionnaires, some people left. 23 people answered the questionnaire.
- Question raised by participant: How do you measure the advancement of open data publication? Some aspects of an initiative may be more advanced while others are less.
- Max: the obligation on the public sector to publish data is important. Some organisations have to publish data.
- Max: one missing critical factor concerns cost and funding.
- Question by Mark Frank: the idea of what you meant with success is not clear. You should clearly
- Mark Frank: add: how confident are you in your assessment. It is important to know whether respondents
- Mark: success definition: it is realy difficult to know about the third dimension of your success definition, namely benefits and impact. W
- Yannis: we clearly need to define what is a successful open data initiative. And why is one initiative more successful than another one.
- Morten Meyerhoff: civil servant perspective: you need to differentiate between data when you talk about open data. Interest groups would like to have anything, but especially personal information. To anonymise data is costing money. Maybe you can differentiate between data that you can release and can potentially be useful and other types of data. More sector specific. E.g. health information, property information. (Iryna: additional success factor!) E.g. Denmark is focused on certain types of datasets.
- Morten: success factor can be: if a very valuable datasets that is usually not opened is actually opened, then this can show the success of an initiative.
- Alon Peled: there is a category about incentives for open data use, but no category about incentive for open data publication. Public sector organisations need an incentive to publish the data.
- Theresa Pardo: we need to think more about publication. You are using it very comprehensively. Is publication preparation? Is it also release? Publication: does publication also include the processes for opening data. Break down publication or clearly define it. Some questions in the questionnaire you refer to preparation of opening datasets. There is a process from decision making about the relevance of a particular dataset in publication, preparation. There is mainly a challenge between the decision making and the release.
- Theresa Pardo: visualising and analysing are seen as curation tasks, so I was surprised to see that this was referred to as curation. Is curation use? (from a librarian view) If you are going to send the survey to many people, curation has to be clarified.
- Yannis: I would see visualisation as curation. Maybe not from a library view.
- … what exactly is Public Sector Information? How does it relate to open data? there is definition in the PSI directive.
- …: reflection on critical factors. Creation of information. Management of reuse. An additional success factor: the management of creating information. Think about creating data Yannis: we consider that the data is already created. Critical: to have the data and to understand this.
- …: (Canadian): there is a difference between transaction and transformation. What we really want to get to: know what we will do with it. The creation and availability the data is also important.
- Peter: most data collected with sensors is not automatically opened. If is would be done automatically, then this could be a success factor.
- Max: the people creating open data may not know that it is open data. It is just their workflow. Workflow processes to automatically release data is important (a success factor).
- Yannis: not only automate, but also think about what you want to publish.
- Yannis: CSF is finding the right data to provide to end-users, to just publish the data that you have.
- Theresa: factors to add: underlying data management and data governance. Regardless of whether data will be opened or closed:. Civil servants need to have data management capabilities. Data governance is also a CSF. Civil servants may realise that the infrastructure is not there’re to open the datasets that they work with. And they also don’t have the skills to make the data available to the public. It’s not about attitude but about capability.
- Peter: Different levels of openness. Traffic light.
- … Information governance/management. The OD community and the … community: very little interaction between these two communities.
- … there are different types of information legislation about data release.
- Yannis: UK with mandatory publication of certain types of data for local authorities.
- …: another CSF is: insight an organisation someone needs to. Finding a person to coordinate open data publication is critical. There are so many different parties within organisations, someone needs to collect their opinions, respond to them. CSF: an enthousiastic data publishing coordinator.
- …: another contextual factor. You might have to make a difference between a project phase and an established open data initiative. Who are you actually addressing with this questionnaire? What is success? What is use? We see different types of use?
- …: Maybe a better question would be: Which of the factors play an important role in your project? Yannis: this is what we do.
- What are the best practices from this study? Identify CSFs in your project.
Plenum Talks 2
RDB Rechtsdatenbank - Legal database for free research; Peter Guggenberger; Manz
Scribe: Claudius Determann File:Notes presentation Manz RDB.pdf
The Development of Commercial Weather Services in Europe since 1970; R. E. W. Pettifer; General Secretary PRIMET
Scribe: Claudius Determann File:Notes presentation Richard Pettifer.pdf
PSI and Digital Content of Public Broadcasters; Philip Etzlinger; uugot.it
Scribe: Claudius Determann File:Notes Talk uugot.it Etzlinger.pdf
Results of Opendatamonitor.eu Survey; Amanda Smith; opendatamonitor.eu
Scribe: Emma Beer File:Amanda Smith ODI Open Data Monitor.docx
Podium Discussion
Scribe: Hannes Kiivet
Participants:
- D = Dieter Zoubek CMC; Head of Open Data working group, Business Consultancy and IT (UBIT), Austrian Economic Chamber
- A = Assoc.Prof. Dr, Alon Peled; Department of Political Science, The Hebrew University of Jerusalem , Israel
- T = Toon Vanagt; data.be & www.openknowledge.be
- G = Dr. Gregor Eibl; E-Government Legal Aspects, Organisation International Affairs, Austrian Chancellery
- W = Wendy Carrara; CapGemini
- P = Phil Archer, W3C
- Audience
Discussion:
- D: real job is to be a businessman, 20 years. Major task to work with AT public data (customers AT). Work with AT chamber of commerce, to find their need. We have been dealing with PSI since 2003. Data has been a thing even before that
- G: responsible for national oGD portal, sets standards. Initiator of ODG portal
- W: represents pan EU open data portal, latest DG Connect initiative (collects national metadata). Setting up, referencing metadata, support network of member states. Monitor reuse and value. Give over to the commision after project
- P: is it ok to charge for data?
- A: researching for 6 years, officals are extremely sophisticated to giving explanation why not to do (even if they have official order). We are stuck, they will not turn to angels, we should nudge them. Most data is publishable (not private, national security etc).
- T: 50 shades of not to open. Fully support that, all across Europe chamber of commerces has monetized this situation
- D: my experience from my clients is that everything is possible, even the ministries are goodwilling, but departments are not interested (have heard, but not discuss)
- G: definition is important, researched, but not reinvented AT special definition. Wanted to give the viewers that it is really easy to use (free, only one licence)
- W: if metadata is bad, we are not cleaning it. Changing/definig DCAT. We are not willing to pay to the government, but for private apps
- Audience (Richard): wahy pay again, because it has been paid already? Political decision should obey, but they protect their data
- Audience (Noel): paying is not the issue, far worse is quality of data. Eg DE is charging for geodata, they use less money than they ask. In FL open data platform paid data is not allowed
- T: nobody wants another label, it is a top-down approach
- A: in 2011 published a article, which proved that most US agencies is doing minimum. Australia have small willing to exchange data between agencies. They set up public sector company, where agencies are data suppliers and owners. Taxpayers have already paid, but more money for good quality or compensation is needed
- W: we allow everything, that is useful information, could create peer pressure to improve
- Audience (Yuri): open data is really interesting for them, so making it business wise is not the only solution
- G: we did not invent OGD, only laid out some principles, recommended format, need for licence before upload etc. PSI has allowed paid data
- D: I am very radical fundamentalist, don't pay twice. My task is to make easier, so if it makes easier then pay. We have data sets harvested and collected by federal law (have high legal quality), some has not, but everything is useful to put to market
- Audience (Georg): we have got the regulation, that's what we have to implement. We should not open the door now about marginal cost
- T: marginal cost is very bad, because we have to go to court to understand what that definition means. Government is the gatekeeper, create raw data, everything else is others' problem/task. We will take care of the data, if you open it
- D: Government should give people the right to use the data. It is very nice if the federal gov has registry which is available and which is not available (when)
- W: there is the issue about the definition on vastutus. Gov is under the sun, which has enabled that all. I work a lot on big data, but businesses cannot understand open data. We have to make them understand, that it is not only bus timetable data etc
- G: in eGov projects the border is changing, who does what (gov distorts market vs does not act enough). In OGD the distinction is much clearer. Entities that charge for data need new business models, which is difficult to do
- A: main task for public sector is to release raw data, but they are never given enough money to create good quality metadata, have enough support from "market" and they cannot produce big data info
- P: we could use github for data quality improvement (fork, modify and merge)
- Audience (Noel): public task is everything that public sector does. If you look at metadata for CKAN (6-7 fields), that is extremely easy.
- P: gov acts sometimes anticompetitive, which stops the value adding thing
- Audience (Richard): value adding is only for private sector
- D: we can discuss, where should the public task stop. Raw data is ok, but metadata creation cost could be shared
- Audience (Noel): raw data is not good definition, source data is better. I've been telling gov not to do certain tasks, eg statistics
- T: I see a lot of good things happening, eg French gatekeepers, who ask money for PDFs (make 30 KEUR/month)
- G: how to keep the motivation to open and use data, need usage up
- A: agencies want use statistics, to give them credit if they opened data (eg open data for app that has 1 download)
- D: one of the most important thing is to be in contact between business, public officials and members of parliament
- W: we are getting there and moving fast. A lot of initiatives around PSI. we are facing paradigm shift, where gov should be doing something else than "own data"
Day 2
Plenum Talks 3
Swimming against the tide - turning data back into information; Peter Bainbridge-Clayton; CTO Kompany.com
Scribe: Valentina Janev
Attendants: 35-40 people
Peter introduced the startup – 3 years already, Raw data does not have value, Kompany provides heavily processed data
Goal: Easy accesses to company data, enhance with official/unofficial, scrape from the web
- Official information is taken, put together, trusted access,
- We don’t have easily access to some data
- Verify evaluate and validate
- Clearing house for business data
Swimming against the tide
- When merging information, there are challenges – highly contextual silos of information
- Misassumptions based on partial data, context is key (E.g. legally responsible person different in Germany and UK; e.g. temperature outside)
- Our mission is to take the data but preserve the context
What is context in this context ?
- Technical definition – data with metadata we keep context and add what you expect that information to be
- Overarching problem that we solve, we take retrieve -> analyze -> transform -> store -> enhance -> playout
- Retrieve Problems – different technical standards – so we build extensive interface for interfaces
- Analyse problems – Germany and Austria different structures but meaning of entities is similar. How we fix – human analysis - the skill of automatic curating / understanding in infancy
- Transform – what we have to do is on the storage layer to handle unknown data/unknown format – we defined an ontology extensible – entity has data -> put in standard format
- Store – RDF – graph model for putting data into structure – ongoing compatibility with SW applications - mySQL behind, e.g. Entity is company, and address, and Managing Director - so we come to a graph structure, Confidence level linked to relation, Source is also important
- Enhance - we use different sources e.g. Facebook; Artificial intelligence (evolutionary algorithms), Neural network applications
- Playout (GUI) - Challenge multilinguality
Officially licensed
- Questions that have to be answered around PSI – ownership, copyright, reuse, origin, privacy
- Interlinked problems to be addressed on EU level
Data banks - Data as an asset under the control of owner/custodian; Joseph Azzopardi; Malta Information Technology Agency MITA
Scribe: Valentina Janev
Attendants: 35-40 people
The goal of the presentation is to propose a concept for repository of document and datasets for governments, taking into account the long lasting principle of the Banks (where we store/keep money). Statements:
- Government do not use their own data, should start from inside, sharing internal data between Gov bodies and then make them freely available
- Charging mechanisms (Art. 6, paragraph 2 - charging mechanism inhibit the opening of data; base registers are foundational layer, on top of that other data can be published with the aim to allow economic growth)
- Open Data Dilemma (we need proper foundation before openning the data i.e. master data that is small )
- Open Data Dilemma (if we talk about open data, then exists closed that that is sensitive, but that data can be exchanged as well between the government bodies)
- Million dollar questions (What to publish - I would start with what is in the law, From paper to PDF next to machine readable format; How to publish – if we have money we will do in machine readable format)
- What to publish ? First Legally mandated registers - Suggestion – before we work on foundational level – give legal basis for person and address maybe through open data strategy
- How to publish ? First in public administration, then through services in private sector
- National Data Strategy / Paradigm Shift
- 4 pillar approach – give identity to entities (vehicles, people, organizations, addresses) Based on the URIs
- Establish register of registers
- Foundation layer will reduce the costs – no multiple copies, Cost is to upload and store
- in the proposed paradigm notary is giving signature to the document with URI
- Third party – Data Bank / Digital Post Box
- Early adopters of the Digital post box – Austria, Denmark, Switzerland and Norway
- We have to reduce the silos - now each government body acts as an own government
- Digital post box could be extended to the fully fledged Data Bank that is the main subject of this presentation
Parallel Sessions 2
FINODEX. Open data, open for business; Miguel Garcia;
Scribe: José Luis Roda
Attendees: 15
The session was divided in:
- Introduction to Finodex
- 3 Success stories
- Some Statistics
- Presentation of the new call for proposals
1) Introduction of FINODEX:
European Project co-funded by the European Commission for the creation of ICT innovative services and products
To participate in FINODEX there are two requirements: Use the FIWARE Techonologies and use of Open data
There are 16 accelerators promoting FIWARE technologies and selecting companies in Europe.
FINODEX, presented by Miguel is the only open data focused accelerator.
The set of 16 accelerators, known as FIWARE Accelerate are currently creating a good community around this technology, as over 500 companies all over Europe are currently using FIWARE technologies.
FIWARE is a platform with different open source technologies open to anyone willing to use them, covering a wide range of technical chapters (big data, cloud, IoT, etc.).
FIWARE has been developed by a set of EU Research projects led by top ICT sector companies such as telefonica, thales, atos, enginering, orange..
FIWARE technologies are open source and free to download. The sustainability in the future is guaranteed in the coming months by the opening of commercial FIWARE as a service platforms by these major companies.
FINODEX accelerator is divided in 4 different phases: a)proposal: open to anyone, b)design: 51 companies c) development: 30 companies, d) tuning: 10 companies, and 3 prizes.
There is a set of 49 incubated projects which will meet in the first networking event of the accelerator during the following month of June
Between both open calls in FINODEX they will accelerate 100 projects.
2) Success stories:
3 companies are here in the session.
OPEN MOVE from Italy:
- transport ticketing made easy
- In the area of Trentino.
- open data and open apis
- you pay for all transportation around in one ticket through the mobile
- local merchants as a third variable to get value
- they have won different awards
- network, coaching, technology, and money
- techonology....reduce costs for implementations using fiware
- standard transportation ,,,, all around europe
- we get a lot information where people is moving.... reports on that. recomended system.
- we dont share data, only agregate data on mobility
- we need agrement with the local ransportation
DIGMAP from Croatia:
- open data maps disseminations.
- geospatial open standards
- based on INSPIRE
- enables easy and standarized way to disseminate spatial and non-spatial data
- it uses open source and open data
- they provide a DigMap tutorial and a DigMap lab.
- get a map in the office , make queues , pay... no... get a self service web site
- you get a map in pdf from different sources. and some more information
- the business is to produce components for free freemiun...consultation, training....
OPEN CHECK from Belgium:
- he shows the documentos of the proposal to show people how it works
- they use 4 open data sets... very important that is really open.
- the funds were in our accounts for the first phase was in 2 days.
- He explained all the problems he had in the first phases.
- provide the business registry data for free
- full text search...
- usefull for journalists, citizens,...
- people register and we can send subcriptions
--- they scrapped public bodies and provide this information
- we have all the belgium companies, millions of documents and we structured them and give it for free.
--- create value from this new open data.... a new model.
3) Statistics
196 submited proposals in the first call.
49 projects passed to second phase
30 passed to third and current phase
Presented in slides:
- Where proposers are coming from
- Where data comes from. Special remark on OpenStreetMap as the main source of open data for all
of the projects and remarkable participation of data coming from different EU dependant institutions
(publications office, EC, European Environmental Agency, etc.)
- What geo scope does the data have. Special remark on the international scope of data (data that is not related to a specific location but to several ones at the same time (a map) or to none (images related to health diseases)
- Project domain vs data domains: some patterns might be seen here.
4) New call for proposals (2nd call).
Explained the new call for proposals.
It finishes june 17th
Requirements for submission: three documents 1) the proposal divided in four sections and a maximum of 10 pages 2)Declaration of honour and 3) admin form. Everything submitted online through F6S.com/fiware-finodex before the 17th of June at 17:00CET
The proposal itself has four sections: technical excellence (including open datasets reused) fiware usage, impact and experience of the team
Beneficiaries are SMEs individually or groups of individuals (from 2 to 4 people) legally established in any EU-28 or FP7 associated country.
Questions.
Q:how important are ties with universities or other centres to submit a proposal?
A:Not forgetting about the business model and monetization of the projects proposed and also this is not designed to fund research but to develop a product
Q:how many projects have customers before submission?
A:around a 20%.. you must do trials and try your product in phase 3
A:its important to look deeply inside fiware platform to know what they have and how to link it.
Publishing and improving the quality of open data with Open Data Certificates; Amanda Smith & Sumika Sakanishi; ODI
Scribe:
Sumika: Myth busters. ODCs don't assess the quality of data. They assess the process around the data publishing.
Certificates do not replace existing standards. It looks at the service. ODCs don't tell you which standards to use.
ODCs are not UK-only. Localised in 9 countries, including local legal framework as well as language.
ODCs are not trying to push everyone to the top level of Expert. It depends on the value of the data and ROI. National Information Infrastructure stuff should aim for Expert because it's so important. But a niche dataset, or test, don't got for Expert.
It doesn't take long because ODI does the certification for you. Working with partners like data.gov.uk. Once you have a certificate you might then be able to generate certificates automatically.
Slides are informative - showing the benefits to different stakeholders.
Discussion
- Is this approach useful?
- What are the tools?
AZ - you said the ODCs can be used as a framework for assessing quality. How?
SS - Assesses things like whether you have documentation, a contact point where you can ask questions and send corrections. by going through the questionnaire, publishers learn what good practice is like. It shows what open actually means.
NvH - So it's not quality wrt the dataset, it's about process. Things surrounding this include legal aspects which vary from country to country.
SS - No, it's not about quality of the data. It's - have you thought about X, Y and Z. It's more about can someone use your data. There is a tech aspect, but we don't emphasise it.
Frederique - how do ensure that users understand your certificates?
SS - We have a lot of users that we know at ODI. But we know that if no one uses them then they're not useful.
F - is there an icon?
SS - Yes, data.gov.uk shows the badge next to datasets with certs.
SS - back to localisation. Close to finalising US version (no database rights etc.). Canada, Germany, NL etc. working on this (volunteer help).
AS - Three most recent certs right now - CZ, SK and UK.
NvH - do you have a trustee in each country?
SS - We have nodes in many countries but volunteers come to use as well. DE and NL done by volunteers, not nodes.
Sebastian - can we see an example?
SS - Jeni Tennison was the brain behind this. How does a potential user make a choice? Open Knowledge Handbook does a great job already. But there's no guarantee you've done it. ODCs give you a self-assement tool.
University Business Accelerators on Open Data: Activities, Challenges and Best Practices; Yannis Charalabidis; University of the Aegean
Scribe: Johann
See resulting Best Practice File:BP StartupAccelerator.pdf
Parallel Sessions 3
Italian National Guidelines for the Valorization of the Public Sector Information; Gabriele Ciasullo, Giorgia Lodi, Antonio Maccioni, Francesco Tortorelli; Agency for Digital Italy
Linked Data Business Cube – Modelling Semantic Web business models; Tassilo Pellegrini; FH St. Pölten, Christian Dirschl & Katja Eck; Wolters Kluwer
OpenMove: How Trentino opened public transportation data and benefitted of a mobile ticketing solution for free; Lorenzo Modena; CEO OpenMove
Parallel Sessions 4
re3data.org - Making research data visible and discoverable; Robert Ulrich & Hans-Jürgen Goebelbecker, Karlsruher Institute of Technology; Michael Witt, Purdue University;,Heinz Pampel, German Research Center for Geosciences
Scribe: Yury Glikman Presented by Robert Ulrich
Research data are valuable. You need to validate the research results and reuse them in future research
The discussion on how to access research data is still going on.
Putting them online is not enough. They should be kept in a sustainable way.
Data journals. Data repositories with data attached to it.
Research data: Observations, technical descriptions, findings
There is a legal aspect in publishing research data.
The current trend is that researchers are putting data online.
Currently we are focussing on disciplinary research repositories.
Repositories are heterogineos as data. There are different kind of Research data repositories (RDR)
Two RDR examples: ftp server with data, where it can be downloaded
Our mission is to built global registry for the RDRs from all academic disciplines
Main asset is the schema describing the repository covering aspects: - General inf - Responsibilities - Legal aspects - Tech standards
There are 140 properties. Special graph. icons are developed to aggregate them.
There is a number of requirement to RDRs: run by legal entity, clear access conditions, previously the interface had to be in English (now it is not a strict requirement anymore), focus on research data
A workflow for publishing a repository in the catalogue has been established.
Researchers use the service. Over 1200 repositories are harvested.
We established collaboration with many initiatives (see the slide). Databib is a similar project in UK with which we collaborate. We merged with Databib in spring 2015. We receive now not only German but also US funding. Technical infrastructure: own proprietory XML schema!!!, Customers request: Excel files, zip files. Many (most) of developers do not understand RDF and do not want to learn.
Keep infrastructure small and running
Provide convenient multiple data formats
Raise awareness by scientists. Teach students and scientists, but not during lectures
Responsibility of users have to clarified
Idea from Phil: search page optimised and customised for a customer
Q: is it a Pan-European research data catalogue
A: we act worldwide
Q: do you include license information
A: we include license information
Q: Are you harvesting the information about data catalogues or about sets of data on those portals
A: we are a catalogues of catalogues only
Scribe: Yury Glikman
Data publishers expect services to be created by third parties. The services are often not sustainable and rely on limited number of datasets
We looked on the process of the service creation. We modelled the service design process consisting from 6 steps (see the slides)
3 main :
- Idea generation
- Idea maturation
- Concept evaluation
What the data in the services:
- Service is based on data. Availability of data is used for the service ideation process. :
- Service provides data visualisation
- Service gives new meaning
- Services with data as resources
- Services validated or enriched with data
Conclusion: Open Data sets can play other roles than just be a basis for visualisations
Question: Do you think that we should promote the idea that data can be used for data traceability
A: Yes
Q: Do you think that data is the solution looking for problems, which people did not think yet
A: Yes
Extracting Structured Data from Unstructured Open Data; Uldis Bojars, Renars Liepins;
Scribe: Yury Glikman
The goal is the discussion on the public information extraction technics
In Share-PSI we focus on structured data (datasets). What about the unstructured or semi-structured data.
They are available on public web sites and have some structure, but require web scraping. Usually such activities are performed by volunteers
Many public catalogues require scraping. Usually they are html pages with more or less structure. There are many technical problems: - Template is changing - Company names are different. Companies IDs could help, but because of copy/paste there are many mistakes
Issues (see the slide) - Publish as structured data - Select datasets, for which people have demand
Unstructured data
What can be done with text: - Basic Metadata extraction - Entity extraction - Relation extraction
Practical use case: people profile extraction from news. Custom solution for Latvian language has been developed. Keywords based relation identification. 26 types of relations have been defined. Finally we got a graph showing how people are connected.
Example: Cultural heritage data. From comments to old photos relations have been identified. Is such information a part of PSI?
Question: If you look on the public web site is it a structured data?
A: no, not enough structured
Q: tools required taking as input structured data and presenting them nicely on the wqeb site
A: there is no enough structured data published.
C: Scraping of scientific and legal papers and finding relationships and making these relationships a part of metadata is very valuable.
C: Coming up with interesting question can help to motivate people to publish structured data.
C: Scientific abstracts are written in the same way. They could be a good subject for the scraping
Q: Where is the line between structured and semi-structured data? Is the amount of structured data is enough
A: All governmental data, which can be opened, should be opened.
Q How do you measure the quality of your tools? Do you have practical ways to measure it
A: There is a sample set of articles, for which the results are evaluated. The correctness is about 50%
Q: When I analyse text of articles I do it differently depending from the research question. It is difficult to understand how you can do it automatically.
A: We provide base layer facts
Bar Camp Sessions
Open Science & Technology
- chair: Benedikt Kämpgen
- scribe: Martin Alvarez
- 15 attendees
Benedikt: There is a huge gap between research and technology nowadays
... During these days we have heard that this gap is 25-50 years
... What is now, what is the ideal situation, what are the barriers, and what the possible implementations
Laslo: Do we talk about research results or data?
Benedikt: You tell me.
Laslo: Up to date papers contain data, but just as an advertisement.
Benedikt: Research results are mainly published as PDF.
Robert: In my presentation before, I presented research catalogs to make more reuse of research.
... Using the data, scientists can repeat the studies, validate the research, and expose the results again.
Hannes: I don't want to validate the research, I trust on scientists.
... I would like to have the results themselves, I don't care about the validations.
Phil: Researches work on assumptions that maybe not all people understand. They have an advanced baseline.
Benedikt: We have two different audiences: (1)Those who wants just the results, and (2) researchers who wants all the information.
Michiel and Phil: Public understanding is important, but scientists use other language.
Hannes: We should look for a broader audience: the public in general, not only researchers.
Uldis and Renars: In order to communicate the results to a broader audience there is a need of an intermediate step.
Phil: One barrier: tradition.
... A good example of misuse of technology and barrier is that for the WWW Conference, all the papers must be submitted in PDF. HTML is not allowed.
Hannes: We need to achieve an efficient research process
Sarven: We don't use ICT properly to express the research results. We can use a larger version (more explained) for citizens and more concise for other researchers.
... HTML is a good layer to include RDFa for data, also CSS+JS for better representation
Hannes: You shouldn't open your data before the publication is issued. Scientists are afraid of exposing the data.
Robert: If you force to publish the information early they are reluctant. So they should receive incentives for that.
Peter Bainbridge-Clayton: Raw data is sometimes a mess. Publishing big raw data may not help to understand the research.
... If I ask for a business research I may only want the result in PDF.
Phil: Lack of recognition of data. [Mentioning UK Open Research Forum that sets some principles]
Hannes: There is few data about ongoing research
Phil: Social media tools can help as a research virtual environment.
Hannes: One goal is to make research process more efficient
Peter: So we should focus on scientists.
Robert: Data and software are together when we refer to research data.
... SW is only one expression of the process and methodology to manage the data.
Tassilo: How we can compare if two artifacts are similar (equal)? This is another topic.
Sarven: Do you refer to quality?
Tassilo: Not necessarily, maybe statistics are produced using different mechanisms. So how we compare them.
Phil: Another barrier is asking specialists doing non-specialistic tasks.
... Uploading software to the data is also something that can be done if your discipline is related to data.
Phil: Tooling is one solution to solve these barriers.
Hannes: Social paradigm (process) is more important than technology. Technology is there.
Tassilo: I'm a data scientist but for me SW is just a black box.
Hannes: As a scientist you don't need to control the whole process of your research. You don't need to understand all details, just apply your expertise in the concrete field.
Robert: You cannot discuss about the process without technology, which is the enabler of the research.
Hannes: Achieving is another barrier. It's costly.
Sarven: You have the Web, this is your library. How do we post the results? Is it a blog enough? We need to enable researchers.
Tassilo: Reputation system of science is really important. How to build the authorised entities.
Phil: Part of the portal's jobs would be enhancing SEO for publications. This is needed to have the work findable on the Web.
Robert: If my paper is good ranked on Google is more reputed or what?
Phil: Nowadays is relevant, of course, if your work is in the first page will be more referred.
Sarven: Anything relevant should be identifiable.
Benedikt: So, breaking down papers in relevant sections referenceable.
... So we will be able to identify sections (and references) by URIs.
Tassilo: The document is easy to consume and all the process is centered on the document (paper).
Peter: Papers are really useful (static, you can read them on the move, etc.). If we eliminate papers, the structure of conferences is wrong?
Sarven: I think is a good idea a well to read it on my hands, but published on the Web as well.
Laslo: The research process will be more transparent if data is published. Researchers are afraid to show the dirtiness of their work.
... If you open the window, people from outside (from your group or community) could enter into your research and this is not desirable in this field.
Sarven: It depends on the theme, sometimes is really interesting receiving inputs. Just use Wordpress, it's just a few minutes.
Tassilo: We should make Open Science attractive to researchers. How? Enhancing the reputation system.
Sarven: Discoverability and accessibility is really important. Be discoverable on the Web (of course).
Peter: There are online journals, abstracts, etc. So people is moving into the Web. There are in their job. Scientists doesn't care about this.
... If some data is published, many researchers will be interested in simplify results, re-use them, etc.
Hannes: I see the same excuses as in the public administrations. So maybe Scientific institutions should be included explicitly under the scope of the PSI Directive.
Benedikt: This session was not intended to get a best practice. Do you think we can develop a best practice from this conversation.
... This best practice is relevant for the PSI Directive?
Tassilo: Some UK universities use data.gov.uk to publish data. The curious thing is that none of them uses an open licence.
Benedikt: It would be nice for the national initiatives to understand the challenges of science research and learn form each other.
Templates and Classifications for Best Practices
Yannis describes a proposal for categorizing the BPs with a specific example based on the template for capturing BPs used on the Timisoara’s template
Opendata Best Practices Classification
1. Title
2. Author
3. Email
4. Countries (multiple sel)
5. Application Level (multiple select)
6. Management level (higher/lower)
Nancy: Do we need such a category, where is the value in this information?
Peter and Yannis: it is important to now the level of managemet in a public body or a company or other thatis inc=volved
7. Administration Sectors
8. Best Practice Subject Category: Management and Policy Issues
9. Best Practice Subject Category: Infrastructure and tools
10. Best Practice Subject Category: Utilisation and value creation
11. Best Practice Subject Category: Semantics and Interoperability
12. Number of Citizens actively engaged
Noel: this will not make sense in most BPs
Amanda: it is good to have metrics but this will not be easy to fill in
Yiannis: It is optional, it will not be filled in if not applicable
Peter: preferable to ask for a description of the impact
Noel and Amanda: agree with Peter’s suggestion
Jan: Estimate of a potential impact and refer also to boundaries involved in the BP
Noel: the idea in general is to show that the BPs will be searchable, we go further than the EC requires to add value to our work, not because we are forced to do so.
Amanda: add an option –“unknown”
13. Number of enterprises actively engaged
14. Best Practice Target Group (we ADDED this category during the discussion)
Discuss how we can apply the categorization to the BPs that are now in DOC format.
Makx suggested and everyone decided to take out 13 and 14
Noel: How is this searchable?
Yiannis: this produces an XLS file with all the responses
Yiannis shared in google docs the agreed categorization form with Peter who will distribute it when finalized.
Peter describes the Template for describing a BP
1. Tile of the Best Practice
2. Outline of the BP should be one sentence, very focused. Satisfy two main user groups.
A. Managerial and strategic role to help for monitoring processes
B. Informative for implementers
3. Challenge
4. Solution. Very tense bullet points
5. Identification. In more detail we explain the BP
6. Link to the PSI directive
7. Why is there a need for this Best Practice?
8. What do you need for this BP?
More detailed recipes or give a general idea of what is needed to be in place in order to achieve this e.g. I need to aplly these changes to my orgnaisation to do this
9. Applicability by other member states
Amanda: potential or what has been done, Noel: it should be a Y/N answer and describe if this we think that could be applicable in another country
10. Contact Info
Jan: Evidence, do we really need it? Should we add it?
Amanda: it would be helpful to add this information
Makx: It is good to have evidence such as Diagrams, images etc
Peter: Each good practice could be revisited and come up
People who are involved in the editing and reviewing process can take a BP and produce 2 or that target different challenges
Amanda: who will do this?
Makx: Us, the volunteers!
Nancy: We should have a numbering system to do the BP description and then apply the categorization
Numbering will be applied, we work on DOC form and categorization will be applied in Google docs
Johann: I suggest that we do not separate the tasks of describing the BP (in doc) and the categorization (in google docs). All information should be in one place
Peter will examine if we can import the information from DOC/WIKI BP description documents in new fields that will be added in the Web Form that Yiannis created.
Peter and Makx will assign to the team members the BPs that each one should work on.
We need to have these finished by September
Peter shows in the Wiki the analysis that has been applied on the existing Best Practices with analytics tools, showing graphs etc
Joseph: why do we need voting if we do not eliminate any BPs?
Makx: We describe each Bp, if there are comments and the BP should be reconsidered then we keep it aside and we decide at the end of the process if we keep it or not.
Noel: we should not vote if a practice is a BEST practice, all should be accepted there is no need to have this process
Makx: We decide if this makes sense to all of us
Nancy: also if there are BPs that should be merged because they describe the same thing
Nancy: before getting to Berlin we will have described and categorized all BPs, we will have the PPT template for the presentation of each BP. In Berlin we spend 3 minutes to present each BP and we reach consensus on which will
Makx: we have a skype call every two weeks.
Decided -->
Wednesdays, every two weeks, @10 CET starting from the 3rd of June.
--
NR
Project Meeting Friday
These notes complement those made in the agenda
Share PSI 2.0
1. reflectins on WS 4 Formal presentations
More plenums, lles discussions ... more submissions, had to sqeeze ...
best practices / papers ... ca. 20 submitted
Quality of presentations very good, maybe even better tan ad hoc interactive interactions
Would ike to get info from privaate sector - t is rarely heard n ublic
was very helpfull and succesful in getting businesses
at review last time we said we needed to get more involved with businesses
it was good to have so many businesses in the room
as the balance better this te - did we get more knowledge
I'd like to see more figures - about businesses
how much money is really made
discussions took place a lot in dialogues and during caffee breaks - had lot of talk ... for me it was caffee breaks where discussions were
would not change ratio of presentations / workshops
it was very good
Presentations were too long ... 10-15 min presentations plus longer discussions would be better
presentations werer too long / academic
15 + 10 pres + discussion
Have you ever been to conf where there was enough time for discussion
when you call companies and say they will speak for 15 minutes and need to spend 2-3 hours to get there
when you are dealing with constituence like businesses it is not good to give them format that hey are not familiar with
what is their point of view abot how to get the data in a format that is usefull
cannot just dump data and have them sort it out
individual data fields that are totally wrong ...
what the problems are is you put a whole mass of PSI ... instead an API that helps you select pieces that are useful
what is needed to make business work succesfully
business to work succesfully with dynamic data ...
precisely suggestion fro PSI 3.0 ...
I liked Nancy session historical on Greece ..
maybe in Berlin we could have similar session
to have overview what is going on
think it is good idea
maybe to have something very structured
to have same questions and pose then to all countries
comment ... this view is specially because we are looking at big picture
in Nancies you see a holistic view instead of having to get your bits and pieces
one could imagine 10 shorter versions from different countries
we have one workshop left and we are not aloud to make changes
some people were waiting to talk about this for Berlin
it could be a workshop ... 2 hours ... how many people could you have? maybe 5?
Nancy - my presentation was long, but it could be shorter ... just structured to
Valentina ... prepare structure and invite colleagues and tell us what you do and run the session
wha about collection of best practices
we an give key examples of best practices in each country ... just very short
just to open eyes and see what happens
we need to talk about open data policy
there are too many topics
we should focus on open data policy
there are couple of elements in PSI directive everybody is strugling with
e.g. marginal costs, time frame ... tehre are 6 to 7 things that are a challenge to everybody
what do you do with marginal costs ... for each challenge one slide ...
to have organization marked
nancy - liked workshops with business taking part
this was more open data in action
i would like to see a list of all good examples that we have presented in 4 workshops ... how businesses used data
in berlin i would like to see a list of all examples
summarizing all these stories
amanda should do it ... we have a section called stories ...we have KPI to have four big stories
i would also like to have in our deliverables a section like this
we would like to persuade our partners in public sector ... we could use these stories
there is a demand for these stories best practices ... sometimes they ask, and it is hard to provide
we had a meeting in austria and had a lot of austrian examples
my believe that i cannot prove we would have local examples there
we have reviewers
feedback from berlin ... it feels ok
put somethng in CFP ...
2. best practices arising from WS
3. review meeting Luxwmbourgh 18 Nov
4. Berlin prep (Yuri et al)
have slides
frauenhofer focus
difference from this WS ... everything will be in one building
topic : maxmizing interoperability
collocated with german open data cities
10 cities will join our meeting
scuss with ministry of interior to do something collocated with them still open
booked everything
time plan...
CFP ... 12. 6.
1.10. agenda published
24.11. project meetings
25.11. opening and parallel sessions
presentation of rooms and other facilities
all rooms are at the same floor
tegel airport is very close ... to know where to book
taxy below 50EUR , 10 minutes
also express bus
who do we invite from Berlin ... keynote or just some talk
we will invite people who are dealing with open data in Germany ..also
open knowledge ...should we do a meet-up with those
we work a lot with cities concerning interoperability ... businesses would like cities to harmonize ... it is less costly for cities as they can share solutions, but for busnisses it is important not to have too many different data formats
question how to engage businesses ... cities network ...
propose to plan ..
will they come???
if we plan I would know who to call. why wouldn't they come if the agenda is ... if they see value from meetings they will come
in finland we are working with cities it is god time to ???
does not government use a lot of ebXML .. my point is what we should be trying to do .. to be enthusiastic ... open data types ... and th
we should boorought into administrations and look at what they already have ... many are focused on UML modeled interactions ..
what we have been finding there are challenges to take what they already have and work from that
open data movement is diverting from what we have ... rganizatinally and legally imbedded systems ...
we have to find what are the challenges of translating what governemtns have into what can be used
germany XML standardization body in place ... 55 core components , model driven infrastructure ... generic code is important for all ... it is a challenge ... take tree structures and tables ...
support sebastian ICA programme is doing community of practices ... working on core vocabularies ... and there are different
i think inspire is something that people would like to hear ... what can they recommend us on interoperability theme - what to do also what are they doing next
inspire is a very good example of having interoperability it is not 100%, but it is a good start
i would like to hear from businesses ... if we have budget to get somebody from USA ... what is their view what government should do about interopeability ... we heard ... do not change it all the time. we can get through the mess, but please do not change it so often
tonio gabriela ... who will be on programme committee ... we will fransesco tortorelli
list of partners with info wether they had a paper/session and organization committe ... FOI in the green
there will be a lot of people
EDF is the previous week
16.-17.11. european data forum in luxembourgh
is anybody else going ... Share PSI on 18. there is project review ... hosts of events that happened during the review period... eparation for Berlin is in that ...
who is going to be in Luxembourgh anyway and would like to be there ...
morning is the time where most presentations are .. would be helpfull if we could stay and take part in review meeting ... it will be in commission building
review perid ends september .. a lot of reporting
one of the things that came up strongly was the positive stories we have .. we will be asking you
2 things .. progress report .. we did this
we need to show what each of partners have done relevant to the project
there is public version on the web site
i need reports from each partner for each WP ... relatively easy thing
the deeper thing is positive stories
they want to hear .. because I am in PSI I was able to do this .. they want to see an added value from your participation
if this network comes to something usefull .. we need to present to reviewers .. this was a positive outcome
they want to know about the best practices and the localized guides.
do you wan tme to do interoerability review ... period ands september we will have more time i will be prepared to do it for review
regarding wiki
need to register at w3c ... send e-mail you used to register ...
mailing list ...
wiki users have separate iling list ... for papers ...
programme committee memebers ... keeping recors of all papers ..peter
how to get a new password ... there is password renew, but it did not work ...send another mail ...
review meeting
project coordinator .. will be asking to fill in progress report
for each WP need to say I did that ...attending workshops, beeing on a committee etc...
rogrss report after the first review period with new versions following comments
than you to Chris for all photos ..appreciate the effort
share ...on line ..send link
ncy - review on meeting on tuesday ..yesterday bar camp on the topic ...
it was a working session ... did a lot of work
yesterday .. go through the process of collecting best practices in a structured way
structure of the best practices
good example of Peter ...paper
Timisoara workshop ... when we write best practice it is good to have examples
rtant thing is to be short, clear .. problem with material we have geathered ... thesre are topics that need to be split into three or two casaes instead of one ...
title
outline (1 or 2 sentances)
challenge
solution (bullet points ...how we solved it)
ID ... why is it best practice, link to PSI directive
why do you need it ?
applicability to other countries
contact info
that is a general structure we are working on
question f taxonomy ... cathegorization
Yannis gave a suggestion on how to work
untry could be a multiple selection
there is other regarding country ...
management level ... admin sector /e.g. health etc.) target group...gov, cit, bus, ngo
linking with PSI directive elements
there is a page on the project website ... list of elements of the directive ..13 ...
page is not yet y and link is not out
www.w3.org/share-psi/elements
idea to link also to other sources of examples
we may need to have discussion with actual provider of the practice
target is semptember ...
in Berlin we will reach consensus process .. we do not want to reject , but it is important to have an idea of what is the real outcome of the project .. importanzt to listen about each of practices.
proposers / reviewers will present
it will be possible to comment and than revisit practices that need to be corrected or improved
what are the days for updating best practices so that we can read themin advance
we will work during summer. in september we will have it ready
will have conference call on this
we are in 10 months review period
december to september period review
must include Lisbon - even though they were there.
final review will gro from octobeer to july 2016.
we need to have something that was collected and prepared before september
having done that so that we can report on that ... show that during review period we made lot of progress
review last time 2 - that is not very good.i hope this review will be better
business of localized guides
people have already written localized guides
what we need to show in fina review ... either new or updated guides must be created
idea is in those documents you refer to best practices that we have put together ... need to show that these best practices have helped people create localized guides
when you signed you committed to be at all meetings, present a paper...and write and/or contribute to localized guides
format ... (pdf 8-), language ... local
requirement is that you point to best practices
flemish one is in english ... it is linked on the wiki
is that the same for the greek handbook
has anybody else written a handbook that was translated in english
... a handbook on open data ...
could anybody who has written anything link it to wiki
Check republic ... creating local guide ...on who does it, roles, standards, etc...processes
what if in lithuania we have a very fresh guide written by PwC ... we do not want to rewrite it ... it is already ordered by the government ... at least to link to existing document and shw that there is commonality
amanda
whaT IF YOU ALREADY PUBISHED A LOT OF GUIDELINES
WE HAVE 15 DIFFERENT GUIDES does that count?
i would say that
elements .. will become the most important think ..
if you have a guide that refers to an element of the PSI directive ... you can link it to the element page on wiki...
ideally you can link to an individual section of guides online ... from the elements wiki page ... al the guides that were done that will be a resource where you can find idelines ...
berlin is the last workshop ... we have time until june to sort it out
after Berlin this will be the focus ... writing, updating, linking hat exists
in proposal we did not omise that all countries will do it, but most ...should
we need to have
Zagreb and beyond
presentation FOI
when ... in March ... polishing guidelines local
what is the function of this meeting ...at wha stage?
ow far do we want to bee
middle ... more people ...
not towards the end
it is also worrying ... we get together ... there will be
regarding the PSI 3.0
we could also work on the new proposal
we do not know at the moment if there will be a call
we do not have it right now
tech park is important .. or not ... because 2/3 Zagreb ...
march ...2016
how long are we going to meet for?
afternoon - evening + full day + mornig ... two nights ... is that ok?
2 nights in Zagreb
...easter falls in march ...
friday 25. good friday ...
monday 14 march ...to 16 march
2 rooms for 50-60 people
check and get back
one birthday on 14th march 8-)
W3C is interested in doing Share PSI 3.0
ODPP
Open data public portal ... Wendy Carrara and ??? are working on
heading at opendata.gov.***
open data knowledge and open data support also working there Public Data RU
ODPP will take over those projects and they will close them up ...
they are rebuilding the portal, training, and reuse ... more evangelism, promotion ... a little bit of technical staff intersoft ... metadata management and harvest frauenhofer ...
Share PSI ... will stay on ...
it would be possible tha they will take the group over and carry on, but maybe not. what we might do .. or rather propose to commission?
significant fraction would like to carry on ... impact of the group will only increase towards the end
what can be done that does not replicate ODPP, and does something new to look at some of these high profile project s that some cuntries have done ... e.g. Slovenian superviser romania ... about taking multiple staff and doing things with then
this kind of the project can only be done by public administration ... official public sector decision
that is very important
can we take legislation and link to court records?
these things are hard to do, and expensive
are we talking about pan european or specific use cases in one country
do not know ... pick 5 to 6 high profile cases , one pan europena ... part of the proces is to select .. skills ... legak iinterop, tech skills etc.
that is a kind of project that we want to do
demonstrate impact of Share PSI, and
there is no such call
it would be expensive
need 3 mill Euro + just running the project
talking about actually implementing projects
Commission may say find matching funding
we received grant from austrian research institute 400K eur data quality link .. to enforce standards to automatically format ...
this is infrastructure that can be plugged in...
ransparency portal in Greece
Supervisor in Slovenia
lot of potential
nobody is offering this money ...
my suggestion is to write up a one pager and circulate and get comments on that
will send it to Martha ... and Christian and Lapsi...
than this is somethnig that we want to do next, it does not replicate ODPP ...
my only reason to do this is to use web properly ... if project ends by not using web properly, i do not want to take part
if you want to do this, and want us to coordinate this ... give feedback please
initial reaction ... if i was to propose this in open group ... we would be interested in integration of different systems ...
we are going to develop many best practices ... NGO ... dolores is taking proposals ... we are experimenting with pilot of linked data things ... releasing common vocabulary ... small pieces ... looking for opportunity to implement bigger things
it is important that we do not duplicate some other projects that are running open government initiative is doing some projects - would like to see more officials from government public sector
i would like to see more public admin
also cross border open data ... more focus we are currently focused on what our own country anbd businesses can do .. but pan auropean dsata and businesses ...
it would have to have a pan european aspect ...
if we had 3 mill Eur ... applicability ... imagine that the pilots were in Spain, Austria etc ... it would have to be applicable to other countries or else why would anybody from other countries be there
need to have transferability ...but at least one should be trans border
maybe we could invite the PSI group ... the official one that meets it would help to transfer
we should involve them directly
i will begin to write up ... and circulate ...
proposal ... to have e-mails of all participants ...hey are on the wiki ...
thank johhan ... i had nothing to do in organizing this meeting
johhan ... thank committee very high quality of papers ... reviewing and providing feedback
also ... thank to Georg who brought in bussinesses.
6. Close