Open Access
Movie Sentiment Analysis u sing Feature Dictionary and Multiview Light Semi Supervised Convolution Neural Network
Author(s) -
Chaitra Kulkarni,
R Suchithra
Publication year - 2020
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.l7981.1091220
Subject(s) - computer science , artificial intelligence , sentence , classifier (uml) , artificial neural network , feature (linguistics) , sentiment analysis , convolutional neural network , convolution (computer science) , point (geometry) , machine learning , domain (mathematical analysis) , natural language processing , pattern recognition (psychology) , mathematics , linguistics , philosophy , mathematical analysis , geometry
Emotional information in film commentary is very important for emotional analysis. An emotional analysis that focuses on classifying opinions into positive and negative classes according to an emotional glossary is a study. Most existing research focuses on word synthesis and user evaluation, while users' attitudes toward feedback are ignored. To consider this point, this paper uses an emotional analysis and in-depth learning approach to examine the relationship between online film reviews, and this point is used for movie box revenue efficiency. In this paper, this work present a 11 different types of Feature Dictionary. It is modeled with information from sentences (i.e., reviews) and aspects simultaneously. First, Feature Dictionary is created with all aspects of the sentence. After obtaining the aspects, it utilize all data in the source domain and the target domain for training Multiview Light Semi Supervised Convolution Neural Network (MLSSCNN) classifier. To understand the predictive performance of this approach several performance metrics are used. The experimental result shows that the MLSSCNN provides a superior predictive effect than other classifier.