SDWIG - Stats on the Web BP
22 November 2017
billroberts: 2 1/2 years Spatial data on the web working group
… working group suggested specific follow on work
… statistics on the web: target 'best practice'
… timings and aims are in the charter
<billroberts> https://www.w3.org/2017/sdwig/
<billroberts> https://www.w3.org/2017/sdwig/charter.html
billroberts: aim to complete half way through 2019
billroberts: my interest arose from sdwwg work. day to day work with statistical data, using rdf data cube
… work for swirl, a UK company working often with government to publish statistical data
… would like to: work with all of you; find good ways to tackle problems, document best practice
AndreaPerego: Joint Research Centre of the European Commission, involved in setting up European infrastructure for environmental data (INSPIRE), including statistical data
… geospatial and statistical data availability
… focus on interoperability for spatial and statistical data
… contributed to spatial data on the web best practices
ChrisLittle: Met Office, sit on SVCG working group in OGC and founded metOcean OGC group and work on OGC architecture board
… interest in graphics, I have interesting use cases for statistics
Dennis: statistics netherlands architect. open data service started 3 years ago. lacking data set interoperability.
… interest in how we can provide more interoperability
… not sure how to do this with data cubes now
josephabhayaratna: Australian company (PSMA Australia), have stong relationship with Australian Bureau of statistics, publishing linked data
… australian statistics geographic standard. we are helping them to publish
… exploring linked data to use more generally
… keen to learn how other people have done things
kalampokis: involved in european union projects: linked open statistical data, using rdf data cubes
… interested in exploring how end users can exploit linked data
… collaborated with SME and public administration
… applying rdf data cube standard. need to create best practices to accomodate all these standards
kalampokis: member of semantic statistics community group
kalampokis: i am a joint point of contact for these groups
<AndreaPerego> Nope.
<marqh> Microphone issues :$
<marqh> I'll type a quick intro for you
marqh: work at met office, colleague of chris Little
… interested in how to encode statistical post processing data sets
… their metadata gets complicated
… enough :)
MichaelGordon: work at Ordnance survey UK, work with public sector. combining geospatial data for policy outcomes
… interested in use cases. worked with office of national statistics UK
billroberts: do you know if Claire is intested in participating in this work
MichaelGordon: I think she is
billroberts: suggestions for how to organise our work. this is largely copying the way we worked on sdwwg
<billroberts> 1. Use Cases
<billroberts> 2. Examples of good practice
<billroberts> 3. Requirement
<billroberts> 4. Best practices
<billroberts> https://www.w3.org/TR/sdw-bp/ and https://www.w3.org/TR/dwbp/
billroberts: links are to best practice document lists from recent groups
<billroberts> https://www.w3.org/TR/sdw-ucr/
<billroberts> https://github.com/w3c/sdw/tree/gh-pages/stats-bp
billroberts: I suggest we follow a similar style. I would like people to make a start on use cases and requirements
… we may need help with permissions set up
if you weren't in sdwwg, you may not have rights to write
<billroberts> fd@w3.org
billroberts: in terms of mechanics, I am happy for people to edit directly into these documents
… happy to pick things up from mailing list
billroberts: over the next couple of weeks, if everyone can start volunteering use cases
… good and interesting ways people have gone about sharing statistical data through the web
… don't be shy about posting your own work
ChrisLittle: I would like to highlight problem areas: interested in scope
… many years ago, we used controlled vocabularies: pressure, humidity; lots of statistics for these parameters
… want property language, or attribute language or statistical language
… have parameter names, and statistics list; avoid combinatorial explosion
… vocabulary, ontology? of major statistical parameters. Do you think that is in scope?
… deliverable for OGC Met Ocean domain working group. we have statisticains on board now
<Dennis> i think its in scope
billroberts: I think this is interesting, it sounds like it is in scope
ChrisLittle: domain over which statistics are calculated, such as time. When RDF data Cube was defined, dimensions and slices over complex time periods is hard; has not been well tackled
… complex language, tied up with calendars and coordinate reference systems. is this pushing the boundaries of scope to far
billroberts: to some extent, if participants think something is in scope it is in scope
… things may need to be prioritised later
<Zakim> AndreaPerego, you wanted to point at the UCR template of DXWG as a possible alternative to the SDW one
<AndreaPerego> DXWG UCR (working draft): https://w3c.github.io/dxwg/ucr/
AndreaPerego: other group is active, don't know if you are aware of this group (link)
<AndreaPerego> An example: https://w3c.github.io/dxwg/ucr/#ID20
AndreaPerego: structure, including description. has use cases, which could be mapped. in github ther is a template for capturing issues, within the issue tracker
… we are looking at the use of DCAT, this could be relevant
… this group should be aware of the work going on in other related activities
<AndreaPerego> DXWG charter: https://www.w3.org/2017/dxwg/charter
billroberts: is that's the stats dcat application profile group?
AndreaPerego: no, this is on their github site
billroberts: it would be good to liaise with those people
… we don't want to be reinventing things
… On the use case template: this would add some structure, without looking like it would be difficult to use
… I'd encourage people to add those section if people have the information available
AndreaPerego: the group likes to have evidence about the requirements and use cases
billroberts: are there more comments on chris's questions about scope? or other questions about use cases?
… initially, writing things down will help. It would be good to add a use case and an example
… each
billroberts: roba welcome. have done round of introductions
roba: I am here on behalf of OGC, working on next geos project
… systems mostly have some geostatistical issues
… sometimes aggregations, or integral
… worked on sdwwg on qb4st and sdmx, for underlying metamodel
… this W3 note, I am eager to support evolution of this, if it turns out to be useful
billroberts: this is an interesting one: extends statistics into aspects of the coverage domain
… I am encouraging people to write down initial use cases, and note down examples of good practice
… links in meeting notes
roba: a lot of sdwwg work was on coverages. sdmx is more about statistical areas. some of principles of QB4ST set out
… need to explore how it links to dggs and register of spatial standards
… more work to be done to look at practical realisation
… getting to the stage of implementation experience
<kalampokis> https://islab-uom.github.io/qbBestPractices/
kalampokis: to warm things up, here is link to challenges already defined
… we would like to discuss these tings in this group
… keen to extend the work already done
billroberts: it would be good to have some time in the next meeting to go into more details on this
<roba> (belatedly :-()
billroberts: final contributions?
billroberts: please everyone make a start on use cases and examples
billroberts: for fixing on next meeting: is this a good time for Europeans, Australians, Americans. are there objections to sticking with 2000 UTC time slot
<roba> +1
<AndreaPerego> +1 from me.
<josephabhayaratna> +1 from me
<Dennis> +1 from me
<MichaelGordon> +1
<kalampokis> +1
billroberts: let's stick with this slot
chrisLittle: OGC meeting in Palmerston North, NZ at OGC; could have face to face meeting
<AndreaPerego> I won't be there, but I can try and join remotely
marqh: Is the aim for this group to meet fortnightly?
billroberts: suggest we meet at this time in 2 weeks, then get a report in a month's time
roba: I would like to announce this group to the OGC Geosemantics working group
billroberts: marqh: yes aim for fortnightly meetings
josephabhayaratna: geosemantics sent me an email this morning. they don't have a ton on their agenda, could we have this meeting within the geosemantics DWG
… at same time as this meeting is scheduled
josephabhayaratna: OGC meeting may connect via gotomeeting, not webex
… would need a good plan for the meeting, present scope, invite comments
… could someone put together a slide pack with who we are and scope?
josephabhayaratna: i'll send template, we can put together information via call or email
<billroberts> bye all
<AndreaPerego> Thanks, and bye!