Improved Metaheuristics with Deep Learning Enabled Movie Review Sentiment Analysis

Abstract

Sentiment Analysis (SA) of natural language text is not only a challenging process but also gains significance in various Natural Language Processing (NLP) applications. The SA is utilized in various applications, namely, education, to improve the learning and teaching processes, marketing strategies, customer trend predictions, and the stock market. Various researchers have applied lexicon-related approaches, Machine Learning (ML) techniques and so on to conduct the SA for multiple languages, for instance, English and Chinese. Due to the increased popularity of the Deep Learning models, the current study used diverse configuration settings of the Convolution Neural Network (CNN) model and conducted SA for Hindi movie reviews. The current study introduces an Effective Improved Metaheuristics with Deep Learning (DL)-Enabled Sentiment Analysis for Movie Reviews (IMDLSA-MR) model. The presented IMDLSA-MR technique initially applies different levels of pre-processing to convert the input data into a compatible format. Besides, the Term Frequency-Inverse Document Frequency (TF-IDF) model is exploited to generate the word vectors from the pre-processed data. The Deep Belief Network (DBN) model is utilized to analyse and classify the sentiments. Finally, the improved Jellyfish Search Optimization (IJSO) algorithm is utilized for optimal fine-tuning of the hyperparameters related to the DBN model, which shows the novelty of the work. Different experimental analyses were conducted to validate the better performance of the proposed IMDLSA-MR model. The comparative study outcomes highlighted the enhanced performance of the proposed IMDLSA-MR model over recent DL models with a maximum accuracy of 98.92%.