Semantic Web Interest Group IRC Scratchpad

Welcome to the Semantic Web Interest Group scratchpad generated automatically from discussions on IRC at irc.freenode.net port 6667 channel #swig by the chump bot, instructions in the chump user manual. Please use UTF-8 charset on IRC and pastebin for code or data more than 10 lines long.

Nearby: IRC logs (Latest) | semantic-web list | W3C Wiki (Recent changes) | delicious swigbot

last updated at 2002-03-13 22:23
e.g. let's get the web/cwm to realize danbri, danc, and aaron will be at WWW2002 in hawaii in May
 
who, what, where, and when?
DanC: e.g. let's get the web/cwm to realize danbri, danc, and aaron will be at WWW2002 in hawaii in May
DanC: to scrape: dan's event log,
DanC: speaker bio page of AALL conference
DanC: "where can I find you" section of Aaron's home page:
DanC: w3c talks and upcoming apperances
DanC: later: internal W3C travel schedules, perhaps
 
danbri: I find this sort of thing crops up a lot. THere's a long and a short way to say something. The short way avoids blank nodes, but involves extra vocab.
danbri: Compare: ownerWillAttendEventWithHomePage( mailto:danbri@w3.org, http://www.www2002.org/) with the longwinded version that mentions the mailbox owner and event more explicitly.
danbri: We can write cwm/n3 rules to make mapping between short form and long form more explicit.
danbri: Can we do anything else to support these conventions? (eg. a bit of vocab that makes triangle structures like this visible to indexing tools that don't grok full N3/rules?)
DanC: I made a composedWith thingy one time... where did I put that...
 
danbri: Idea for cute hack: Dump a list of channels, nicks, and isSubscribedTo relations from irc.openprojects.net. Pump it into the path finding tool Damian and Libby wrote for co-depiction (I've already run it on companies and board members) to find connections amongst the communities who use this server.
danbri: Issues: (1) is it OK to harvest the data? (2) Without knowing more about the people whose foaf:nick's these are, it might not be an interesting dataset (3) a study over time would be interesting... (4) what about privacy? is there a P3P policy for openprojects? does it extend to chat?
 
danbri: Neat! Next challenge is to write the metadata that tells us that there is a full size version, so clipart tools can extract images of chaals and wendy in full detail.
danbri: ...and Daniel, and everyone else, I mean :)
danbri: Ideally, my thumbnail script that generates the other images on my site would write this, and we'd take note of things like the black border around the image, since that might offset the figures slightly.
 
 
 
chaals: example output based on (very rigidly structured for compatibility) an example file
 
danbri: "Presenting tailored resource descriptions: Will XSLT do the job?", Alison Cawsey (www9 paper)
 
danbri: Am looking for better sample images to markup
danbri: nearby: html imagemap experiment and Jim's SVG imagemap editor.
chaals: RDFWeb: Codepiction in SVG seems to be the only documentation of a sort that has many connected words in it
danbri: See this rough writeup for a bit more detail.
 
danbri: It stops before I get to ontology stuff. I might plunder this for other writeups at some point; chumped here so I can find it again.
gk1: This note casts RDF schema as a kind of constraint language, in some sense limiting the things that can be said. I see it as a kind of opposite to that: RDF schema allows one to extract more information from what is said - an RDF schema allows additional information to be inferred from a given graph.
danbri: Interesting reading of it! I guess I agree with you, and talked about constraints mostly in terms of RDF (not just RDFS) being a set of constraints on ways of writing wellformed XML. This would be worth clarifying; thanks!
 
jhendler: thoughts about an RDF editor/markup tool hooked to an ontology library
jhendler: code coming along later
gk1: I saw this earlier -- I look forward to the download being available
chaals: mailto:chaals@w3.org suspects that the note should mention the times that XML infrastructure under RDF is handy - like when a graph is normalised for a particular query that can be done by really dumb syntax-processors.
 
danbri: With the SOAP and Squish stuff, I've almost got a use case for hacking on this. Squish/SOAP interface to Annotea for annotations, ditto for database of W3C tech reports. Task is to take a query that asks for annotations of W3C docs that (say) are in Last Call, and decompose it, do the queries, and combine the results.
gk: Peeked at the paper, got thinking about query partitioning and service descriptions
gk: could the _original query_ be used as a kind of service query, with a database returning an indication of the query parts that it can fill in? Otherwise, we seem to be getting into a complex area of service description and, as the paper says, query decomposition.
danbri: Service description and query certainly has a role here. There are so many different things you might want to know about services and the databases behind them, if you're going to do distibuted query efficiently. It creates a nice marketplace for service description factoids.
 
tbl: Example of paper in eth area from google
 
gk: Follow-up to an exchange with DanC a while ago... my attempt to try and put some flesh on the idea of using N3-style contexts without quoting the URIs. It involves a small extension to RDF MT, inspired by something in Guha's thesis.
gk: The formalism isn't fully worked out with proofs, etc, but I'm hoping there's enough here to allow exploration of the ideas.
 
danbri_: See SOAP in Mozilla writeup for info on what they've already got. SOAP Encoding as an RDF datasource might be very attractive for the Mozilla folk, since lots of the GUI's work (via XUL/XBL) over the RDF abstraction
danbri_: SOAP's built in to 0.9.9 now. Wow, a prime number detection service, whatever next, stocktickers?. news headlines?!, aren't we supposed to use RSS for that? note: most of these eamples come from scaped demo services set up at [http://xmethods.com/|xmethods
 
danbri_: "A lot of structured data is stored in relational databases. ... REVERSE allows to connect to an existing relational database using JDBC, and to semi-automatically define mappings from the database tables to the ontology, and finally to extract the data [...]"
 
DanC_jam: scratching the "we need issue tracking tools" itch
 
DanC_jam: Dave Beckett (Fri, Mar 08 2002)
DanC_jam: hm.. acutally, he changed his mind... more recent proposal
dajobe: I'm wary of grabbing extra punctuation for ntriples that n3 might want to use - hence backing off (). The space before "." thing is something I'd like to sort out. If it remains optional, then we need to reduce the allowable chars in the language component of "chat"-en etc.
dajobe: BTW I've implemented xml"<a>foo</a>"-en and variants
 
DanC_jam: By CHRIS SPRIGMAN
DanC_jam: Tuesday, Mar. 05, 2002
DanC_jam: found it via fork
 
 
AaronSw: I have a Bacon number of 2
AaronSw: Eric Miller has an Aaron Swartz number of 2, BTW
AaronSw: Whether he's Eric(II) or Eric(V)
AaronSw: aka oracleofbacon.org
 
Created by the Daily Chump bot. Hosted by PlanetRDF.