Fast and accurate annotation of short texts with Wikipedia pages

摘要

We address the problem of cross-referencing text fragments with Wikipediapages, in a way that synonymy and polysemy issues are resolved accurately andefficiently. We take inspiration from a recent flow of work [Cucerzan 2007,Mihalcea and Csomai 2007, Milne and Witten 2008, Chakrabarti et al 2009], andextend their scenario from the annotation of long documents to the annotationof short texts, such as snippets of search-engine results, tweets, news, blogs,etc.. These short and poorly composed texts pose new challenges in terms ofefficiency and effectiveness of the annotation process, that we address bydesigning and engineering TAGME, the first system that performs an accurate andon-the-fly annotation of these short textual fragments. A large set ofexperiments shows that TAGME outperforms state-of-the-art algorithms when theyare adapted to work on short texts and it results fast and competitive on longtexts.