Customer product reviews, collected from manye-commerce sites like Amazon, Flipkart have been wildly used to study the customer purchasing habits, purchase motivation, recommendation systems etc.Sentiment analysisis the term given to the analysis of purchase behavior from reviews. Opinion Mining is one of the major branches of Natural Language Processing (NLP). One of the tasks of Opinion miningis to predict positivity or negativity or neutrality out of users’ opinions throughonline reviews. Aspect level opinion mining is the concept of analyzing the customer reviews at aspect levels, i.e., opinion mining will be performed forproducts’ features analyzation.This paper proposes a two stage model for opinion mining of customer reviews on laptops extracted from review dataset from SemEval-16. In the first stage data preprocessing and identification of product aspectwhere implemented. Later opinion mining at the aspect level using convolutional neural networks is performed as the second stage. Finally results were presented.
Volume 12 | Issue 2
Pages: 2617-2623
DOI: 10.5373/JARDCS/V12I2/S20201312