Paper: Enhancing Linguistically Oriented Automatic Keyword Extraction

ACL ID N04-4005
Title Enhancing Linguistically Oriented Automatic Keyword Extraction
Venue Human Language Technologies
Session Short Paper
Year 2004
Authors

This paper presents experiments on how the performance of automatic keyword extraction can be improved, as measured by keywords previously assigned by professional indexers. The keyword extraction algorithm consists of three prediction models that are combined to decide what words or sequences of words in the documents are suitable as keywords. The models, in turn, are built using different defi- nitions of what constitutes a term in a written document.