| Abstract |
One topic of this presentation is my past (master project) research into an affect analysis of Dutch social online media. Crowdsourcing (mturk.com) is used to annotate sentences with emotions, which are then used by a classifier to determine the affect in a text.
The other topic is my current and future work for the Data2Semantics project. This project aims to provide a semantic infrastructure for e-Science. The goal of my research is to design/develop/evaluate techniques for ranking of query results of linked data. The focus of my research is twofold:
- extend reasoning to take top-k ranking into account: current sparql top-k queries calculate the complete result set, and apply the sorting afterwards. The performance might increase by developing a way to incorporate the top-k ranking into the reasoning.
- improve/implement semantic distance measures, which will be used to rank query results on |