Open AccessAN EXPERIMENTAL ANALYSIS ON ECOMMERCE REVIEWS,WITH SENTIMENT CLASSIFICATION USING OPINION MINING ON WEB
Author(s) -
S S Latha
Publication year - 2021
Publication title -
international journal of engineering applied science and technology
Language(s) - English
Resource type - Journals
ISSN - 2455-2143
DOI - 10.33564/ijeast.2021.v05i11.045
Subject(s) - sentiment analysis , computer science , product (mathematics) , feature (linguistics) , field (mathematics) , process (computing) , information retrieval , data mining , feature extraction , data science , artificial intelligence , mathematics , linguistics , philosophy , geometry , pure mathematics , operating system
Sentiment analysis is a big branch in the field ofnatural language processing. Sentiment analysis mainlytext based analysis, but there are some challenges thatmake it difficult as compared to traditional text basedanalysis. This paper empathizes on the need of an attemptto improve research process and progress of sentimentanalysis on the basis of investigation. Outcome of theanalysis are summarized in this paper.This paper analyze the reviews of products manuallyby collecting data in the form of a excel file. Then it willproduce and classify the reviews as positive or negativecomments to get the best product. Now it’s more relevantto automate reviews data it is growing exponentially. Thismethod works by web scrapping reviews from e-commercewebsite. Data cleaning is applied to remove the unwanteddata known as stop words. The features are identified. Thefeature can be camera, battery life etc. Obtain frequencyacross all the products and for all the reviews per feature.The intended work is to extract the features from thereviews and detecting the polarity for each aspect, thusresulting in feature extraction matrix (FEM). FEM matrixhas each row as an observation for a product and each ofthe columns represent the feature. List of Products basedon highest value of FEM for searched features and productrecommendations are generated based on the usersearched feature.