SV_MEETING_TITLE -- 08 Sep 2008

<kei> Eric, am I on the right track so far?

<ericP> yup, great

eric, i will be scribe.

How do I capture multiple lines for one speaker without re-typing the name for each line?

I think there was a trick.

<ericP> matthiassamwald: I'd loved to scribe

<ericP> ... but how do i capture multiple lines

<ericP> ericP: like that ^^^

understood.

<kei> zakim who is here?

<kei> matthias, did you call in?

technical diff

<cgolbrei> christine

is the passcode HCLS or BIORDF?

<holger_stenzhorn> BIORDF

<Huajun> the pass code is biordf

<kei> the passcode is BIORDF

give me one minute.

Zakim ??P25 is matthiassamwald

<AdrianPa> I can not see the link to the slides on the webpage

<AdrianPa> me neither

<holger_stenzhorn> http://esw.w3.org/topic/HCLSIG_BioRDF_Subgroup/Meetings/2008-09-08

Kei: Olivier sent his regrets.
... He will hold his presentation in the next BioRDF call, in two weeks.
... First, let me introduce myself
... I have been involved in Semantic Web developments for more than 3 years, before the HCLSIG started.
... My main interests are bioinformatics and database integration
... Let's start with the agenda. Please open the slides that are attached to http://esw.w3.org/topic/HCLSIG_BioRDF_Subgroup/Meetings/2008-09-08
... First I would like to thank Susie for coordinating the last charter of the task force. She did a very good job.
... I will talk about the near-term goals of the task force.

<ericP> slides

Kei: I listed four near-term goals
... "Enhance the HCLS KB. As part of the enhancement, we would like to incorporate new data sets that help the KB span from bench to bedside."
... "Broaden the community of users who have skills to add data to the KB."
... "Increase the value and use of the HCLS KB by identifying scientific use cases"
... "Work on a human-friendly UI, explore

mechanisms for updating the data sources, and

adding new data sets.

scribe: "

<ericP> HCLS KB sources

scribe: a list of datasets in the Neurocommons KB can be found at http://www.w3.org/TR/hcls-kb/
... The datasets were available in a variety of formats, some of them already were provided as RDF/OWL
... But in most cases we needed to convert to RDF/OWL
... During this process, we also created semantic links between datasets.
... Once we have these ontologies loaded into these triplestores, applications can be developed.
... The datasets we include will be driven by the needs of the scientists.

EricP: A federation approach would be better than an aggregation approach, which is not THAT different from warehousing.

Kei: (agrees)

Adrian: Would it be a static or a dynamic integration?

Kei: In the warehouse approach, the datasources would be converted periodically and loaded into the triplestore from time to time. This might be associated with better performance for query.
... In the federated approach, queries must be mapped to distributed sub-queries.

<Huajun> can u hear me?

<ericP> ack ??P16

Huajun: As more datasets will be loaded, will there be a performance issue?

Kei: I cannot answer that accurately now. It depends on the queries.
... On with the agenda... Olivier will talk about the conversion of UMLS into RDF in the next telecon.
... The reason for inlcuding the work on UMLS was motivated by the interest of many participants, e.g., interest in SNOMED, interest from the COI group...
... UMLS is also important for bioinformatics, e.g., for network analysis and data mining.
... Of course, there is an issue with licensing.
... It is open-access within the US if you sign an agreement with NLM.
... It might be a problem.
... What are the licensing issues with SNOMED?

Holger: We are currently working with SNOMED, it is a problem here in Germany.
... John Madden wanted to ask the people at SNOMED. We can give you an update when we find out more.

EricP: It might make it easier if we just require parts of SNOMED for certain projects.

Kei: Exploring those licensing issues could be an agenda item for someone in this group.

(Who is talking?)

Kei: Let's talk about hosting of the KB
... The first was at Science Commons, the other at DERI. They were not exact mirrors, but close.
... How do we migrate the data from the current to the new KB?

<holger_stenzhorn> Matthais: finished my PhD on SW and biomedical research and neurological research

<holger_stenzhorn> M.. now working for DERI Ireleand

http://hcls.deri.ie

<holger_stenzhorn> ....DERI agreed on hosting the knowledgebase at their site

<holger_stenzhorn> ...DERI still has still this commitment

<holger_stenzhorn> ...and to start projects in the HCLS area

<holger_stenzhorn> ...cureently, quite good situation: we can probably use the DERI server for the next 2 years in our developments

<holger_stenzhorn> Kei: what support can be given by DERI for the hosting?

<holger_stenzhorn> Matthias: inofficial agreement, that we can use the DERI servers for the next 2 years

<holger_stenzhorn> ...but it needs to get official agreement for that

<holger_stenzhorn> ...but everybody gives me thumbs up

<holger_stenzhorn> ...I have personally set up the current HCLS server at DERI and would continue that

<holger_stenzhorn> ...the machine that the Virtuoso server is running on is quite good, a quad core with large memory

<holger_stenzhorn> ...should be sufficient for supporting the data over the next 2 years

<holger_stenzhorn> ...so that server can be the main hub for the data over the next 2 years

<holger_stenzhorn> Kei: could relay that to the request to the HCLS chairs and get back to you

<holger_stenzhorn> Matthias: that makes sense

<holger_stenzhorn> Kei: so I will do that request

thanks, holger. will scribe again.

Kei: Adrian, could you describe your situation now.

Adrian: I will move to Berlin soon and be part of a new working group.
... I keep on working on Rule Responder and Semantic Web for e-science.
... So this would still be of interest.
... But I am still negotiating with the institution in Berlin.
... So I cannot commit anything at the moment.
... I will try to get more details soon.

<ericP> matthiassamwald: re KB development, how will the HCLS KB sync with the Science Commons KB

<ericP> ?

Kei: Unfortunately, I still don't know the official status of Science Commons in this charter.
... We need to talk to Neurocommons and the HCLS chairs to build some consensus.
... Last agenda item -- ideas for future work.
... There is a wiki page about project ideas.
... We should also define how we can interact with other task forces (e.g. clinical).

<kei> http://esw.w3.org/topic/HCLSIG/Project_Ideas/

eric: i am most interested in projects that build on what we currently have.
... and also 'bench to bedside' work.
... we have some gaps in the KB that hinder interesting use cases.

kei: huajun proposed an integration of traditional chinese medicine data with western medicine.
... i proposed 'wiki neuron'.

matthias: i plan to focus on semantic web technologies for (neuro-)cognitive science.

Kei: the next call will be in two weeks.

<AdrianPa> bye

<cgolbrei> bye

<kei> eric, are you still there?

<ericP> on irc, yes

eric, you take care about rrsagent, right?

- DRAFT -

SV_MEETING_TITLE

08 Sep 2008

Attendees

Contents

Summary of Action Items

Scribe.perl diagnostic output