Multiword Extractor

This is a lexical acquisition web-service for the automatic extraction of multiword expressions from large corpora. The service takes in input a POS-tagged corpus in CoNLL-X format plus a pair of POS-tags for the first and last word of a MWE, and outputs a list of extracted (candidate) multiword expressions with a set of linguistic and statistical information. The output can then be post-processed through filters that will refine and improve the accuracy of the extraction, and finally converted to an LMF-compliant XML lexical resource. The tool code is available open-source at https://github.com/francescafrontini/MWExtractor. Further details can be found in: Quochi Valeria & Frontini Francesca & Rubino Francesco. 2012. A MWE Acquisition and Lexicon Builder Web Service. In Proceedings of the 24th International Conference on Computational Linguistics (COLING 2012), December 10-14 2012, IIT Bombay, Mumbai, India. Frontini Francesca & Rubino Francesco & Quochi Valeria. 2012. Automatic Creation of quality multi-word Lexica from noisy text data. In Proceedings of the Sixth Workshop on Analytics for Noisy Unstructured Text Data (AND2012). December 9, 2012, IIT Bombay, Mumbai, India (Co-located with COLING2012).