Automatic Documents Annotation by Keyphrase Extraction in Digital Libraries Using Taxonomy

Abstract

Keyphrases are useful for variety of purposes including: text clustering, classification, content-based retrieval, and automatic text summarization. A small amount of documents have author-assigned keyphrases. Manual assignment of the keyphrases to existing documents is a tedious task, therefore, automatic keyphrase extraction has been extensively used to organize documents. Existing automatic keyphrase extraction algorithms are limited in assigning semantically relevant keyphrases to documents. In this paper we have proposed a methodology to assign keyphrases to digital documents. Our approach exploits semantic relationships and hierarchical structure of the classification scheme to filter out irrelevant keyphrases suggested by Keyphrase Extraction Algorithm (KEA++). Experiments demonstrate that the refinement improves the precision of extracted keyphrases from 0.19% to 0.38% while maintains the same recall.