Output details

Building a high-quality sense inventory for improved abbreviation disambiguation

<22> Variability and ambiguity of terms pose challenges for indexing and search of document collections, acronyms being a prolific source of variability and ambiguity. This work is significant for its novel method to automatically disambiguate extracted acronyms and their expanded forms, by clustering the expanded forms into sense inventories. This method outperforms all other approaches to acronym disambiguation. The method is currently applied to enhance search of Europe PubMed Central, is a service offered by the National Centre for Text Mining and has been used by the National Library of Medicine-US.