Sentiment Analysis on Choice Home warranty Product & Service Reviews

Abstract

'Choice Home Warranty Product & Service" has ernerged as one of the online platforms to provide online product seliing and warratty services to the customers. This is a project to implement a "Sentimsrit Analysis on Choice Home Warr*nty Product & Service Reviews". To classi$ the revisws of every customer ei*rer positive or negative is diffrcult and time consuming and this siruation can make obstruct against a corporate progress. This project is basicaliy analyzing of the customer revieltrs to determine the customer reactions or the basis of polarity and subjectivity. This polarity is then convefied to a scale of -1.000 to 1.000 (where 0.000 is average) and thus the rating of individual product or service is obtained. The main prnpose of "Sentiment Analysis on Choice Horne Warranty Product & Service Reviewso' is to improve the relationship *ith the customsrs and provide the better services to customers as per customers demand by anatyzing the reviews either positive or aegative. 'Sentim Anatysis oa Choice Home Warranty Product & Serrrice Reviews" in this proj*t is inctuded few mo&rles- The modules included are Data Scrapping, Data Anaiyzing, Nanral I*anguage Processing, Data Cleaaing, kmrnatization, Review Classification, Graph Generating. Data Scrapping to import the web data from "Choiee Home Warraaty Product & Seruice Reviews" web page and sort out the reviews. Data Analyzing module not€ out the minimum unit of reviews" Natural Language Process basically defines the numerical value from each word of reviews. Data Cleaning rnodule reduces the anomaly and special characters &om the reviews. Lemtratization reftrs to dcing thing properly with the use of vocabulary and morphclogieal analysis of words to remove the inflectional endings and to return the dictionary form of a word. Review Classification is the target module to classifu the reviews on basis of polarity and subjectivify. Finally, a bar chart is produced &om the polarity to make visualize the analysis resuh by the Graph Generating module. Tbe toots used to develop this "sentiment Analysis on Choice Home Warranty Product & s€rrte Reviews" included Jupyter Notebook, python packages: bs4, pandas, numpy, nLt co,rpus. textblob, matplotlib.pyplot, and raadom.