Online Text Categorization System Using Support Vector Machine

A. K. JUMANI, M. H. MAHAR, F. H. KHOSO, M. A. MEMON

Abstract


Text Classification is a need of day, large text existing in the form of stories, news etc. Likewise, this system came into being along several techniques like, Support Vector Machine, Neural Networks and Decision Tree. Stories, newspapers are the page collection that belongs to text categorization. Various Sindhi newspapers are regularly published and Daily Kawish is one of them. People are facing difficulties during reading newspaper because there is no any specific option that will categorize particular news related to sports, technologies, crime, fashion and current affairs. For this purpose, a Text Categorization System (TCS) for Sindhi language is presented in this paper. Five classes are used and scanned each newspaper page inside a single class. It is too difficult to predict how many users will read newspaper simultaneously and for this, web performance is tested. Moreover, for the classification of the text from pages, precision, recall and f-measure are used to measure and achieved 67% of accuracy to classify the text from newspaper pages. It would be beneficial for those who want to save their precious time during reading newspaper.


Full Text:

PDF

Refbacks

  • There are currently no refbacks.


Copyright (c) 2018 Sindh University Research Journal - SURJ (Science Series)

 Copyright © University of Sindh, Jamshoro. 2017 All Rights Reserved.
Printing and Publication by: Sindh University Press.