Publication Details
Towards a Framework for Comparing Automatic Term Recognition Methods
Schmidt Marek, Ing. (DCGM FIT BUT)
Smrž Pavel, doc. RNDr., Ph.D. (DCGM FIT BUT)
Zdráhal Zdeněk (OPEN)
automatic term recognition, keyword extraction, evaluation
Automatic Term Recognition focuses on the extraction of words and multi-word expressions that are significant for a given domain. There is a considerable interest in using ATR for automatic metadata generation, creation of thesauri and terminological glossaries, keyword extraction, ontology building, etc. This paper introduces a new ATR evaluation framework that consists of the open-source referential implementation of advanced ATR algorithms, annotated datasets and tools for the task-specific evaluation. We demonstrate the advantages of the framework on an experimental study comparing the base ATR methods as well as their combinations under various conditions. The created platform is freely available and prepared for extensions by other researchers.
@INPROCEEDINGS{FITPUB9065, author = "Petr Knoth and Marek Schmidt and Pavel Smr\v{z} and Zden\v{e}k Zdr\'{a}hal", title = "Towards a Framework for Comparing Automatic Term Recognition Methods", pages = "83--94", booktitle = "Znalosti 2009", series = "8th Annual Conference, Brno", year = 2009, location = "Bratislava, SK", publisher = "Faculty of Informatics and Information Technology Slovak University of Technology in Bratislava", ISBN = "978-80-227-3015-0", language = "english", url = "https://www.fit.vut.cz/research/publication/9065" }