W3C

LODD

13 May 2009

irc record

Attendees

Present
EricP, bosse, Anja, junzhao, oktie
Regrets
Chair
Anja
Scribe
ericP

Contents


bosse's report

anja: did bosse have time to answer the pharma ontology group questions
... matthias as loaded them into a SPARQL endpoint
... we're now translating the questions into SPARQL

bosse: didn't find the appropriate data for the question that i got from the last call
... will return to this question. some of the answers should be in dbpedia
... will work on questions that come this week from pharma ont

jun's report

anja: we're trying to link junzhao and kei's datasets

junzhao: i think my question can be rephrased "are their alternatives that interact with a particular set of genes"
... as it stands, we don't have sufficient context
... would be interesting to see chemical compounds shared with western drugs and natural alternatives
... that question is more complex (fuzzy) than the others

anja: queries taking 18hrs/failing
... trying locally

oktie's report

OktieH: these questions aren't in my area of expertise

anja: perhaps you find something related to Linked Clinical Trials

OktieH: concearned about the use cases

<junzhao> do you mean this page? http://esw.w3.org/topic/HCLSIG/LODD/Business

OktieH: most of them are not use cases for linked data, they are use cases for a web of documents
... the URLs are to HTML pages

<OktieH> Cloud: http://esw.w3.org/topic/HCLSIG/LODD/Business

anja: the uses cases in the paper were about finding at alternate drug names and consituant compounds from web pages

bosse: could you [OktieH] write down in the wiki how you'd like the use cases modified?

OktieH: for most of these, we already have linked data for many of these sources

ACTION: OktieH to ground use case (question) pages with links to existing linked data

triplification challenge

<OktieH> http://triplify.org/Challenge/2009

<OktieH> The deadline is changed to June 30th

anja: we need a use cases that leans on triples instead of HTML pages

OktieH: yes, that could distinquish us from the other submissions

anja: how important is the schema to the triplification challenge?

OktieH: the thing that helped us get the prize was the metadata
... but this is not new this year
... we should work on factoring our data to finding common schemas

anja: should we use the same schema, or annotate equivalent classes/properties

OktieH: i think it's nice if we use shared ontologies where possible

junzhao: i caution against working too hard on sharing schemas
... prefer to focus on use cases and SPARQL queries

anja: OktieH, please record questions that cross your mind while addressing the use case

OktieH: perhaps folks more familiar with the domain can help with the use cases
... perhaps we should focus on fixing links
... there are some folks working on user ingerfaces which benefit folks more familiar with the domain than linked data concepts

anja: is anyone trying to test their UI witcP> OktieH: doesn't look good for the project
... could use a temporary discussion page before placing in the use case page

ericP: i propose continuing to record the use cases in the wiki page, but annotating each with it's maturity level [ACCEPTED]

Summary of Action Items

[NEW] ACTION: OktieH to ground use case (question) pages with links to existing linked data
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2009/05/13 16:10:47 $