6th ESAIR in Burlingame!

The sixth ESAIR (Exploiting Semantic Annotations for Information Retrieval) workshop was successfully executed on October 28 2013 in the somewhat less than vibrant but impressively scenic Burlingame, next to the San Francisco Airport, on the first day of the CIKM 2012 conference. A full report will be published shortly in SIGIR Forum: the main message as excerpted from the last hour of discussion appeared to be: “It’s all about the task” — no annotation model can be formulated, designed, implemented or evaluated separate from what might be expected of its utility and usage. The previously formulated truth from ESAIRs where semantic annotations was viewed as a linking scheme between some model and units of data was to some extent questioned: could one possibly formulate a model based on the data alone, and view annotations as links between units in the data – documents, entities, what not.