Cross-document named entity coreference resolution for Dutch as a pre-process for named entity based text mining

People create massive amounts of texts, finding information within these texts usually involves reading. It is possible to extract information from texts automatically but most techniques are farm from perfect. One example is text mining based on names: extraction of names from texts and discovery of information using these names. However, names often have a lot of variances, the same name can refer to different things and different names can refer to the same thing. Finding out which names in a large set of documents refer to which ‘entities’ in the world is the focus of this graduation project. This research studied methods to perform cross-document named entity coreference resolution for Dutch and the impact of this resolution on name based text mining.