Evaporation Parameter-Ant Colony Optimization (EP-ACO) Algorithm With Elam Neural Network (ENN) for Sentiment Analysis

In recent times, online shopping is major stream that refers to users for purchasing and also consume along development of internet technology. The user satisfaction is maximized efficiently through performing Sentiment Analysis (SA) of huge amount of user reviews on e-commerce platforms. That is still difficult for predicting a correct sentiment polarities of user reviews due to modifications in sequence length, order of textual with difficult logic. In this research, proposed an Evaporation Parameter-Ant Colony Optimization (EP-ACO) algorithm with Elam Neural Network (ENN) for sentiment analysis of amazon product reviews. The reviews from amazon platform are taken for sentiment analysis and then the pre-processing and word embedding process are performed. The features are selected by using EP-ACO algorithm which selects the relevant features for sentiment analysis. Then the sentiment analysis id performed by ENN network, that classifies the reviews as good or bad. Proposed technique obtained 98.26% accuracy, 97.73% precision, 97.03% recall and 97.41% f1-score that is better than previous methods like Convolutional Neural Network (CNN) – Long Short-Term Memory (LSTM).