Emotion Recognition for Phone Reviews in Non ...

Emotion Recognition for Phone Reviews in Non Query Based Systems

Author Name : Xavier Johanna Christy.L, Siva Sankari.E

ABSTRACT

Emotion recognition in text is a content based classification that deals with extraction and analysis of emotions. Extracting emotions from online reviews is pivotal to many applications in other domains. Finding emotions expressed in phone reviews help the company in providing some information in the decision‐making process to ensure business growth. The goal of this project is to conduct emotion analysis on Amazon phone reviews using various Natural Language Processing (NLP) techniques in Non Query based systems. The dataset that consists of general information from where the needed data has to collected and preprocessed. The phone reviews of customers are analyzed to find out the opinions about the product. The five major psychological emotions happy, sad, fear, surprise, anger are recognized from the reviews. The positive  emotions (happy, surprised) of the product help  the company to focus more on their production and increase their sales and  the negative emotion(sad, fear, anger) in the reviews are found so that the company can make necessary change , improve the quality of product ,acquire more customers and increase their revenue. The emotions from the text are detected using emotion packages which computes the emotion percentage from each review. The overall percentage of emotion from all the reviews are also found out and visualized to get a better understanding of customer opinion about the product.

Keywords: emotions, emotion recognition, phone reviews, Natural Language Processing