S3T conference: Invited Talk by Allan Ramsay

Topic of the talk by Allan Ramsay was on exploring what logic / expressiveness is needed to formalize various natural language questions. Second part of the talk was on Satchmo, a higher order logic theorem prover implemented in Prolog.

S3T conference: The Share.TEC project

Several talks at the S3T conference illustrated aspects of the Share.TEC project. One talk gave an introduction into the adaptation approach – a mixture of implicit and explicit method. Only metadata and links to documents (probably in some file system or the Web) are maintained. Modular storage of document (components) or querying of a document repository is not of interest. However, the project members share the believe that the writing of small, atomic content units is better then the writing of large, complex ones (see little theory approach). Nevertheless, I don’t understand whether and how these granular units could be handled. Probably this will require to add a repository for the document content eventually.

Another shared the feedback of teachers to the metadata schema and their working behaviors with the system. Interesting for me is to see that the main criteria for evaluating content are first the annotations, then the rating, and only on third position the name of the author. It was explained to me that the participants of the workshop did not know the authors and first had to look at the annotations and then assess the quality by themselves. It would be interesting to explore if over time trust in certain authors established and the “name of author” then becomes the most important quality criteria.

S3T conference: Interoperability of Systems

I talked with Ricardo Jardmin Goncalves yesterday (publication list). One of his research interest is on applying semantic technology to support the interoperability of systems, e.g., to solve issues like “Americans calculate in miles and Europeans in km”. They are interested in handling changes, e.g., from one Airbus model to the next version. Thus they provide some kind of change/consistency management. Would be very interesting to learn more!

S3T Conference: Very nice demonstration on ellipse

by Pavel Boytchev.

S3T Conference: Workshop on Semantic Technology by Ontotext

I attended a workshop by the Ontotext company today. It gave an introduction into semantic technologies, semantic repositories (in terms of RDF Triple stores) and, particularly, focused on web mining. This was new and very interesting since I so far simple assumed that all this tedious markup is given. Now I learned about approaches how to at least divide these efforts between machine and humans. The key issues (like providing the type of relations) stay for the human user.

Some aspects I learned about: Most technologies by ontotext are based on GATE, a framework for web mining. Ontotext is well established, with several partners, e.g., the DFKI Saarbrücken, DERI and AIFB in Karlsruhe. Named graphs where mentioned as an approach to model different views on concepts, e.g., different visions of “web mining”.