Question Answering for Dutch: Simple does it

by Arjen Hoekstra, Djoerd Hiemstra, Paul van der Vet and Theo Huibers

When people pose questions in natural language to search for information on the web, the role of question answering (QA) systems becomes important. In this paper the QAsystem simpleQA, capable of answering Dutch questions on which the answer is a person or a location, is described. The system’s algorithm does not use a lot of complex NLP-techniques, but instead uses the magnitude of and redundancy on the World Wide Web to its advantage. The system has been evaluated on the DISEQuA corpus and performed quite well: MRR near 0.5. For further improvements it can easily be extended by adding more rewrite rules and applying more sophisticated filtering and tiling.

[download pdf]