Emotion Recognition-oriented Mobile Phone Users Purchase Evaluation Analysis

Abstract

Sentiment analysis refers to analyzing people's emotional tendencies expressed in subjective texts. Emotional dictionary and machine learning-based analysis methods are the main research methods. Appraisal theory, developed by linguist J. R. Martin based on systematic linguistics, is mainly divided into three subsystems: engagement, attitude, and graduation. Based on the attitude system of appraisal theory, this paper analyzes 158 overseas purchase evaluations of the Xiaomi mobile phone using the emotion dictionary. 158 corpora were evaluated based on the selection of adjectives, adverbs, and negative words as categories of the Emotional Dictionary from 500 online shopping reviews and their ratings as criteria for emotional attitudes. The results prove that consumers generally have a positive attitude towards Xiaomi mobile phones, which are embodied in battery performance, mobile phone fluency, and camera, while the network compatibility and overall performance of mobile phones show a relatively negative emotional tendency. Therefore, this research can help consumers and businesses conduct sentiment analysis in a more comprehensive and efficient way, promote purchase decisions, and enhance product competitiveness. This study also provides enlightenment for the combination of appraisal theory and sentiment analysis.