Search Paper
  • Home
  • Login
  • Categories
  • Post URL
  • Academic Resources
  • Contact Us

 

THE IMPORTANCE OF NEUTRAL CLASS IN SENTIMENT ANALYSIS OF ARABIC TWEETS

google+
Views: 176                 

Author :  Hamed AL-Rubaiee

Affiliation :  University of Bedfordshire

Country :  Bedfordshire, United Kingdom

Category :  Computer Science & Information Technology

Volume, Issue, Month, Year :  Vol 8, No 2, April, 2016

Abstract :


Product reviews are becoming increasingly useful. In this paper, Twitter has been chosen as a platform for opinion mining in trading strategy with Mubasher products, which is a leading stock analysis software provider in the Gulf region. This experiment proposes a model for sentiment analysis of Saudi Arabic (standard and Arabian Gulf dialect) tweets to extract feedback from Mubasher products. A hybrid of natural language processing and machine learning approaches on building models are used to classify tweets according to their sentiment polarity into one of the classes positive, negative and neutral. In addition, Regarding to the comparison between SVM and Bayesian method, we have split the data into two independents subsets form different periods and the experiments were carried out for each subsets respectively in order to distinction between positive and negative examples by using neutral training examples in learning facilitates. Similar result has been given.

Keyword :  Sentiment analysis, Mubasher, Twitter, Saudi Arabia, machine learning, neutral class, Pre-Processing

Journal/ Proceedings Name :  Computer Science and Information Technology

URL :  http://aircconline.com/ijcsit/V8N2/8216ijcsit02.pdf

User Name : srav
Posted 10-07-2017 on 11:44:26 AEDT



Related Research Work

  • An Adaptive And Smart System For Parental Control On Digital Games
  • Intelligent System For Solving Problems Of Veterinary Medicine On The Example Of Dairy Farms
  • A Diet Control And Fitness Assistant Application Using Deep Learning-based Image Classification
  • 2d Image Features Detector And Descriptor Selection Expert System

About Us | Post Cfp | Share URL Main | Share URL category | Post URL
All Rights Reserved @ Call for Papers - Conference & Journals