HeiNER - the Heidelberg Named Entity Resource

In this talk, we present HeiNER, the multilingual Heidelberg Named 
Entity Resource. HeiNER contains more than 1.5 million disambiguated
English Named Entities (NE) together with their translations and 
transliterations to 15 languages. Our work builds on the approach 
described in (Bunescu and Pasca, 2006), yet extends it to a multilingual 
dimension.

NEs contain crucial information about texts and thus are often used in 
Natural Language Processing applications. We will present how to mine 
Wikipedia to extract disambiguated NEs by simple a heuristic with a 
precision of 95% and exploit the crosslingual information contained in 
the online encyclopedia to carry out the translation and transliteration 
of NEs in various languages. The internal link structure can be used to 
gather linguistic contexts for every disambiguated NE which should make 
HeiNER a valuable resource not only for NE translation but also for NE 
Recognition and Disambiguation.