Title :
Application of Modified Spectral Bisection for Segmenting Malayalam Documents
Author :
Dhanya, P.M. ; Jathavedan, M.
Author_Institution :
Dept. of Comput. Applic., Cusat, Kochi, India
Abstract :
Topic Segmentation has been an area of interest for Text summarizers. Many of the Malayalam documents confirm to a main topic as well as many subtopics. This paper presents the idea of using the repeated spectral bisection method to find the subtopics in the document. This method is a widely used for partitioning graphs based on the properties of the similarity matrix created.
Keywords :
document handling; natural language processing; modified spectral bisection application; segmenting Malayalam documents; text summarizers; topic segmentation; Buildings; Dynamic programming; Partitioning algorithms; Silicon; Vectors; Web sites; Bisection; Segmentation; partitioning;
Conference_Titel :
Advances in Computing and Communications (ICACC), 2013 Third International Conference on
Conference_Location :
Cochin
DOI :
10.1109/ICACC.2013.13
