(by Marieke van Erp)
This year, the 15th International Semantic Web Conference took place in Kobe Japan. The conference itself was 3 days with 3 parallel sessions as well as a 3-hour poster and demo session one evening. The two days prior to the main conference 5 tutorials and 16 workshops took place.
For NLP aficionados there was the the LD4IE (Linked Data for Information Extraction) workshop which I attended on Tuesday morning, the NLP&DBpedia workshop that I co-organised on Tuesday afternoon, the keynote by Kathleen McKeown (Columbia University) on Wednesday and the NLP session in the main conference on Friday. But there were other NLP papers dispersed along the conference programme.
For the CLARIAH community some of the work McKeown presented on computational analysis of novels is probably most relevant. It was also nice to see that more research is moving towards event extraction, for example in the work of Valentina Presutti and Aldo Gangemi (presented at the LD4IE workshop). They presented a new resource called Framester that links up all types of resources such as FrameNet, VerbNet and DBpedia to help describe events. New at the conference was the journal papers track, where I got to present our work on building Event-centric Knowledge Graphs [slides] [paper] to a pretty big room.
Sentiment analysis was also a hot topic, with several interesting papers such as On the Role of Semantics for Detecting pro-ISIS Stances on Social Media by Hassan Saif, Miriam Fernandez, Matthew Rowe and Harith Alani and A Replication Study of the Top Performing Systems in SemEval Twitter Sentiment Analysis by Efstratios Sygkounas, Giuseppe Rizzo, Raphaël Troncy. Incidentally, the last paper was only one of two replication papers in the conference.
There weren’t that many papers this year dealing with humanities research questions. Next year’s conference will take place in Vienna, perhaps CLARIAH can mitigate that?