Papers, ICURA and the Knowledge base

Ingo has finished and submitted the paper on the 3D Virtual World development methodology to the i-Society conference. According to the web site the notification of acceptance will be sent out by the end of February and hopefully the paper will be accepted as this is an important step for Ingo towards finishing his PhD.

In the meantime he has been working on another paper project, which is equally important. It is the paper on the usability evaluation of Itchy Feet which will be submitted to the Journal of Human Computer Studies. He worked in all comments that he had received on the previous version of this paper and also restructured some parts. The first half of the Background Section has been rewritten and a 3D Virtual World terminology has been established. The goals of the evaluation have also been restructured and formulated more clearly in five concrete goals. The Results Section has been restructured according to these five goals and some further results have been added. The revised version has been sent out for another review round.

Ingo also had the chance to test Icura for the first time and it already worked quite well. Some issues have been identified and together with Josef he has been working on the optimization of Icura such that it can be used in a user study. Josef spent most of the time with fixing logical issues to keep the game playable,  to avoid frustration and to keep it as linear as possible. He added additional GUI elements to display the goals and the controls during the game. Here’s the last screenshot from ICURA.

Another task has the goal of adding a mouse over functionality such that the mouse cursor changes when it is placed over an object with which the user can interact. Ingo was able to implement a first version that outputs the name of the object. As soon as this functionality is integrated the game, ICURA can be considered as finished (at least in a first version ) The next step will be to plan and conduct the evaluation.

Markus had to rebuild the complete ontology and knowledgebase because the inferencer used by SwiftOWLIM was producing some awkward data. The reason was that the RDFProperties which were defined as functional in the original harmonise ontology (that was transformed from RDF to OWL) were somehow misinterpreted and in the end e.g. a hotel instance was also declared as skiing. Therefore, all functional properties were corrected, the ELMO classes were rebuild and the knowledgebase was repopulated. For some reason the newly generated ELMO classes need considerable more time to generate RDF triples than the original did. This needs to be fixed in near future. Furthermore, Markus started to work on a Gold Standard Set, which is a very tedious task. First fetched webpages that hold valuable information need to be separated from pure webpages that list just hotels or cities without any additional information. Subsequent a knowledgebase needs to be generated that holds the entities from the EC database that are mentioned within these webpages. Finally the documents need to be annotated by hand referencing to the entities (concepts & instances of concepts) in the knowledgebase and indexed by the systems indexer. Markus will start working on the book chapter today for the next 4 weeks therefore little progress in terms of the Semantic Search System is expected.