Conference Details
- Date of Call: Friday June 19, 2009
- Time of Call: 11:00am Eastern Time
- 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: ~1 hour
- Convener: Tim Clark
- Scribe: Matthias Samwald
- Attendees: Tim Clark, Alexandre Passant, David Shotton, Tudor Groza, David Newman, Simon Buckingham-Shum, Paolo Ciccarese, Matthias Samwald
Agenda
- Simon Buckingham Shum, Open University / Knowledge Media Institute,
will present and discuss his work in scholarly discourse ontologies SBS Presentation
Minutes
Tim: Simon might be the person here that has been working on this topic for the longest.
---- PRESENTATION BY SIMON BUCKINGHAM-SHUM ----
simon: i don't know much about life science, but there are many domain - independent components to discourse representation
i am also interested in humanities etc.
while there are differences, there is also much in common.
simon: slides for the presentation: http://www.slideshare.net/sbs/supporting-sensemaking-by-modelling-discourse-as-hypermedia-networks
... slide 3 -- basic idea of a digital abstract
... i was working with enrico motta and john domingue
... i come from a human-computer interaction / hypertext background.
... as we had a bit of a debate at the HypER workshop, there is a lot of working going on around MACHINE annotation. however, in my opinion, we also need good human interfaces for browsing.
... back in 2001, we asked what will happen in 2010. will we still publish scientific results primarily as prose?
... let's start with IBIS, a simple, issue-based information system. (slide 6)
... issues, positions, arguments and relations between them.
... slide 7 -- the idea was to represent IBIS with icons, arrows etc.in interactive graphs. made with the "compendium" software (free, java-based).
... alisw 8 -- an example about the iraq war.
... slide 9 -- another example -- "mapping a nuclear power debate on a blog"
... slide 12 -- danny ayers developed an RDF schema for IBIS
... we are applying IBIS now in the ESSENCE project (deals with climate change)
... we want to facilitate interchange between debate analysis tools.
... slide 13 -- ScholOnto project, focused on annotation of scientific literature.
... slide 14 depicts the basic structure of citation in scientific publications
... the role that a particular concept plays is entirely contingent on the rhetoric move you want to make. this is quite distinct from most other ontologies.
... we wanted to avoid an explosion of different types, so we came up with a small set of basic types.
... slide 16 -- is there any empiric / data-driven way of coming up with these discourse relationships?
... but we also started with theoretical approaches, 'cognitive coherence relations' (slide 16)
... e.g., distinguish between semantic and pragmatic relationships.
... slide 17 -- there is an RDF schema for ScholOnto, Cohere supports RDF in the API and also accepts RDF uploads.
... slide 18 -- discourse annotation and search tools
... e.g., the ClaiMapper tool (slide 19), the ClaimSpotter annotation tool (slide 20, based on work by Teufel et al.)
... slide 21, Cohere (based on Java, we are now moving to Flex)
... slide 23: asking the question "what papers contrast with this paper"
... slide 24 -- "what is the lineage of this idea?"
---- DISCUSSION ----
tim: since the HypER meeting, has there been work on starting a collaboration with Tudor?
simon: i would be interested in whether IBIS is of use to you, how it compares to SWAN,for example.
... we focused on the user interface, not so much on the machine-readable part
tim: i have not really worked with cohere or compendium yet... what is most interesting to biologists is probably a subset of what you are currently doing.
... it would be interesting to look at what you model, and see if you are modelling this subset better that we currently are
... we should also see if there are intersection points between your ontologies and ours. putting the data on the web in RDF makes it possible to use a large variety of tools.
simon: have you dealt with different levels of discourse?
tim: with our use-cases, we mainly dealt with the 'lower' levels. so far
... we can model different kinds of arguments (e.g., 'slippery slope arguments' etc.)
... we can have a follow-up discussion about simon's work in the next call.
<AlexPassant> ttp://www.w3.org/2001/sw/hcls/notes/sioc/
---- TOPIC: Interest group notes ----
alexandre: i pasted a link to the note (http://www.w3.org/2001/sw/hcls/notes/sioc/ )
... does not describe main SIOC ontology (has already been published)
paolo: i did not want to go into details about each module in the note, so i put this information into an external web page
i invite people to read the notes, if you find something that is unclear, it is open to discussion.
paolo: we will start working on the joint note (SWAN-SIOC mapping)
... matthias started a wiki page listing ontologies for scientific discourse, at the moment only SWAN and aTags are listed, others (such as cohere) should be added.
<paolociccarese> http://esw.w3.org/topic/HCLSIG/SWANSIOC/Examples
tim: next meeting will be on 31 july, please send comments about the IG notes at lease one day before.
... send them to alex.
... thank you all for attending, thanks to simon.
