SmILE Semantic Information Systems and Language Engineering Group

With a couple of months of delay, I'm going to share with you the experience I had at Web Intelligence 2008, in December, in Sydney, Australia.

The Web Intelligence series of conferences is advertised as being a confluence of different Artificial Intelligence directions and trends with the goal of using them individually or combined to shape the next generation of Web-enabled products, systems or services. And to my pleasant surprise, this was really the case. It was extremely interesting to have the chance of attending talks in different sessions focused on web mining, web information filtering, social networks, and ontology engineering and semantic web, and all under the same umbrella. Needless to say that, if we consider WI's parallel event, i.e. the Intelligent Agent Technology conference, I would have liked to fork myself :) and be present in multiple sessions at the same time (especially since agents represent one of my old loves). Unfortunately, this is not possible (yet), thus I concentrated on attending the sessions about Web mining and farming, Web information filtering and retrieval and Semantics and Ontology Engineering. And since I've mentioned this last session, another pleasant surprise was to see the number of Semantic Web 'usual suspects' increasing from the previous year's edition, and therefore showing (to some extent) the community's interest in disseminating its results in other communities that might gain added value from them.

Three of the invited talks were the most interesting talks I've attended during the conference. The first one was by Edmund Durfee, one of the main planning gurus, who talked about 'Planning for Coordination and Coordination for Planning'. The talk represented in its main lines, a quite comprehensive view on the current status of the planning research domain. The second one was by Nigel Shadbolt, about 'The Emergence of Web Science' (since I didn't have the chance to see this talk beforehand in some other venue, I was pretty happy that I had this opportunity here). The last talk, which on a personal basis I thought it was the most interesting, was about 'Granular Computing on the Web' by Tsau Young Lin. The main application idea presented here was about a way of modeling n-grams as granules of a particular document and then using these granules, for example, for Named Entity Recognition.

Overall, I would recommend having a look at the upcoming Web Intelligence Conference (http://www.wi-iat09.disco.unimib.it/WI09/WIhome.htm) to be hosted by the University of Milano Bicocca, in September.