SENTIMENT ANALYSIS OF SMARTPHONE PRODUCT REVIEW USING SUPPORT VECTOR MACHINE ALGORITHM-BASED PARTICLE SWARM OPTIMIZATION

research
  • 15 Sep
  • 2016

SENTIMENT ANALYSIS OF SMARTPHONE PRODUCT REVIEW USING SUPPORT VECTOR MACHINE ALGORITHM-BASED PARTICLE SWARM OPTIMIZATION

Nowdays, social media gives the very large effect to the digital improvement in terms of global communications. It can be seen from the increasing of consumers opinion and review about smartphone product that they write on various social media. So that can be recognized various sentiments about the product either positive, negative or neutral. Sentiment analysis is a computational study of the opinions, behaviors and emotions of people toward the entity. The entity describes the individuals, events or topics. That topics generally could be the review of diverse datasets, one of it is a product review. www.gsmarena.com is one of the website that provides information of smartphone products review. Reexamination of smartphone product review by classifiying it into positive and negative class is the good way to find out the consumers response of the products quickly and properly. From some techniques of classifications, the most often used is Support Vector Machine (SVM). SVM are able to identify the sparated hyperplane which maximize margin two different classes. However SVM is lack of electing appropriate parameters or features. Election features and setting parameter at SVM significantly affecting the results of accuracy classifications. Therefore, in this research used the merger method election features, namely Particle Swarm Optimization in order to increase the classifications accuracy Support Vector Machine. This research produces classifications text in the positive or negative of smartphone products review. The evaluation was done by using 10 Fold Cross Validation. While the measurement accuracy is measured by Confusion Matrix and ROC curve. The result showed an increasing in accuracy SVM of
82.00% to 94.50%.

Unduhan

 

REFERENSI