High Impact Factor : 4.396 icon | Submit Manuscript Online icon |

Cluster Based Text Mining

Author(s):

Nikita Rajan Andhrutkar , MET's Institute of Engineering, Nashik; Prashant M. Yawalkar, MET's Institute of Engineering, Nashik

Keywords:

Text Mining, Text Feature Extraction, Text Classification

Abstract

In text mining, discovery of relevance features (RFD) is the challenging task as term-based approach is used for it. A term based approach suffered from the problems of polysemy and synonymy. Pattern based method is also introduced in previous systems which perform better than term based approach. In proposed system, relevance feature discovery model is introduced which discovers positive and negative patterns from given dataset of text documents. Pattern taxonomy mining i.e. PTM along with n-gram is proposed for pattern discovery. The proposed model describes relevance feature in three groups such as, positive, negative and general. F-Clustering algorithm describes the feature clustering approach in which set of positive documents DP+ and set of negative documents DP- get sorted. Term based classification is given as a contribution of proposed work.

Other Details

Paper ID: IJSRDV5I50336
Published in: Volume : 5, Issue : 5
Publication Date: 01/08/2017
Page(s): 483-488

Article Preview

Download Article