A Comparative Study on different Keyword Extraction Algorithms

Abstract

Growth in the number of research documents getting published is increasing. Finding a research document under interested domain by referring the whole paper has become a tedious task. Keywords, Keyphrases gives the summary of the text. Keywords and keyphrases help in understanding the information described in the research document. The domain of a research document can be determined based on the keywords and keyphrases extracted. Extracting keywords and keyphrases manually is a tedious task. Automatic keyphrase extraction techniques help in overcoming this challenging task. This paper is a comparative study of unsupervised keyphrase extraction algorithms without using corpus. It compares the performance of PositionRank which considers the position of the all words occurrences in the document with TextRank and RAKE (Rapid Automatic Keyword Extraction).