HCLSIG/LODD/Meetings/2009-08-19 Conference Call
Conference Details
- Date of Call: Wednesday August 19, 2009
- Time of Call: 11:00am Eastern Daylight Time (EDT), 16:00 British Summer Time (BST), 17:00 Central European Time (CET)
- Dial-In #: +1.617.761.6200 (Cambridge, MA)
- Dial-In #: +33.4.89.06.34.99 (Nice, France)
- Dial-In #: +44.117.370.6152 (Bristol, UK)
- Participant Access Code: 4257 ("HCLS").
- IRC Channel: irc.w3.org port 6665 channel #HCLS (see W3C IRC page for details, or see Web IRC)
- Duration: ~1h
- Convener: Susie
Agenda
- iTriplification Challenge - Anja
- FDA related data set - Adrian TBD
- Web questions demo - Anja
- LinQuer compared to Silk - Anja
- Data update - Anja, Jun, Matthias
- Use case progress - Bo, Ted, Susie
- AOB
Minutes
Attendees: Jun, Matthias, Anja, Elgar, Bosse, Susie
<matthias_samwald1> --- TOPIC: Triplification Challenge Paper ---
<matthias_samwald1> Anja: we had much work to do in the last days, the page limit was a bit problematic
<matthias_samwald1> ... Oktie, Jun and others cleaned up the LODD wiki pages
<matthias_samwald1> ... now we have a clear roadmap, an interlinking section...
<matthias_samwald1> ... I hope we will be keeping it up-to-date
<matthias_samwald1> ... I will send a link to the PDF of the paper later on
<matthias_samwald1> ... we also submitted the source code (with the exception of Oktie)
<matthias_samwald1> ... there will be 5-10 "nominated" papers
<matthias_samwald1> ... those have to give a talk
<matthias_samwald1> ... Chris Bizer is already a keynote speaker, so he cannot be a presenter for the challenge.
<matthias_samwald1> ... maybe I could attend
<matthias_samwald1> Susie: I put together a guideline for authorship and author ordering for HCLS publications
<matthias_samwald1> ... it is important to get fair recognition.
<matthias_samwald1> Anja: about the prices, I have a proposal that worked in previous occasions.
<matthias_samwald1> ... spending the money on a meeting and catering.
<matthias_samwald1> Susie: we should decide before we (potentially) win. Spending it on a meeting sounds good.
<matthias_samwald1> Anja: the paper is at http://www.anjajentzsch.de/lodd/2009%20Triplification%20Challenge%20-%20LODD.pdf
<matthias_samwald1> Susie: are there other opportunites for submitting papers?
<matthias_samwald1> Anja: I have a 6-pages version of this paper, we should look out for deadlines of other conferences.
<matthias_samwald1> ACTION ITEM: Look out for other opportunities.
<AnjaJentzsch> http://www4.wiwiss.fu-berlin.de/lodd/topquestions/
<matthias_samwald1> anja: http://www4.wiwiss.fu-berlin.de/lodd/topquestions/ can be used to answer example questions
<matthias_samwald1> susie: thomson reuters is working on extracting drug company information from news articles
<matthias_samwald1> anja: we also have to search for duplicates, maybe with SILK
<matthias_samwald1> ... in Dailymed we have "marketed" drugs, without information about MOA, so i looked up Drugbank to get MOA information.
<matthias_samwald1> ... another way was via LinkedCT (information about ongoing trials), but there was no information about the chemical compounds.
<matthias_samwald1> susie: how can we improve the visualisationß
<matthias_samwald1> jun: better labels that make relationships clearer
<matthias_samwald1> anja: also, how can we tackle more complex question (such as "what lifestyle changes should we recommend?")
<matthias_samwald1> susie: do you have thoughts on how to do this?
<AnjaJentzsch> http://esw.w3.org/topic/HCLSIG/LODD/Interlinking
<matthias_samwald1> anja: i have some variants of such question, maybe we can discuss them on the wiki, and we can try to find similar questions that can be answered.
<matthias_samwald1> bosse: if you type in "Astra Zeneca" you get some results, but there could be some things to change and add
<matthias_samwald1> anja: this is not for getting ALL the information, but for getting quick overview and links
<matthias_samwald1> anja: please send suggestions for improvement to me via mail.
<matthias_samwald1> anja: we could also add a form to the page where visitors could suggests other questions of interest
<matthias_samwald1> TOPIC: SILK and LINQ
<matthias_samwald1> anja: in SILK you can define rules and find matches between two SPARQL endpoints.
<matthias_samwald1> ... LINQ does similar things, but not for SPARQL endpoints. it is more complex.
<matthias_samwald1> ... Oktie believes it is very good, especially the metrics.
<matthias_samwald1> ... Oktie will be away for a few months, though.
<matthias_samwald1> ... i cannot set it up on my own
<matthias_samwald1> susie: i guess that waiting until Oktie comes back is the thing to do.
<matthias_samwald1> TOPIC: TCM data
<matthias_samwald1> jun: the TCM data is quite stable now.
<matthias_samwald1> ... now, i want to help anja with the question answering service.
<matthias_samwald1> ... another topic would be high-quality pathway datasets
<matthias_samwald1> ... but i prefer to orient on the top questions on the wiki
<matthias_samwald1> we migrate all the datasets to the HCLS KB at DERI, the perfomance was quite reasonable
<matthias_samwald1> susie: maybe you could provide a demo of your application in one of the following calls?
<matthias_samwald1> TOPIC: various things
<matthias_samwald1> anja: we cleaned up the STITCH dataset
<matthias_samwald1> ... also got some feedback (we don't have alternative drugs for Aspirin, for example)
<matthias_samwald1> TOPIC: Use-case progress
<matthias_samwald1> susie: i think there is not to report at the moment. providing feedback to the top questions seems useful.
<matthias_samwald1> susie: the translational medicine ontology group is quite busy with defining classes and relationships.
<matthias_samwald1> .. beyond that, it might make sense to have a joint call.