Graph based Automatic Odia Keyword Extraction from Text Document

Abstract

Keywords are the words in a document that create a center of attention to readers for better understanding and comprehension about the subject. The applications of keywords are enormous. Now a days, there is a vast availability of Odia text and not much research work has been published for keyword extraction from documents written in odia script. This paper proposes a novel unsupervised undirected weighted graph based approach for extracting keywords from odia text. Importance of extracted keywords is analyzed by computing the weights of the nodes in the graph generated from the text. The performance of the proposed technique has been reported in terms of precision, recall and F-measure. It is observed from the experimental result that the proposed graph based technique can effectively extract keywords from Odia text with minimum computational complexity because of its implementation simplicity.