REVIEW PREDICTION USING SENTIMENTAL ANALYSIS FOR MULTIPLE E-SHOPPING WEBSITES

1A.Rifana Banu, B.ShanmugaSundari, R.Meenakshiammal, L.Paul Jasmine Rani

105 Views
32 Downloads
Abstract:

Digital reviews play a crucial role in enhancing global communications Therefore the (digital reviews) sellers are sometimes benefitted , but always the loss is for the customers (online purchasers) due to the fake data.. E-commerce acts like Amazon, Flipkart, etc. provide a platform to consumers to share their experience and provide real insights about the achievement of the product to future buyers. In order to extract valuable insights from an outsized set of reviews, classification of reviews into negative and positive sentiment is needed. Sentiment Analysis may be a computational study to excerpt subjective information from the text. In this work, Sentiment analysis or opinion mining is one of the primary tasks of NLP (Natural Language Processing). In recent years , to review any product or about web site Sentiment analysis play a vita role. In this paper, we focus to implements the matter of sentiment polarity categorization, which is surrounded by the elemental problems of sentiment analysis. polarity categorization is proposed in a general process for sentiment with detailed process descriptions. Usage of data in this study are online product reviews collected from various e-shopping websites like Amazon.com, Flipkart, epay and etc. Product reviews were pre-processed using text processing techniques. In pre-processing, the Product review files are generated as a flat-file. The flat file is tokenized into sentences and the keywords are listed after removing the stop words. We have identified the frequency of each word and extracts the topic which has the highest frequency count. Similar comments in each topic are clustered and then the clustered words are classified into positive or negative comments. The classified comments are generated as a chart for straightforward visualization.

Keywords:

review prediction using sentimental analysis For multiple e-shopping websites

Paper Details
Month2
Year2020
Volume24
IssueIssue 2
Pages4329-4338