A Hybrid Method Based on Particle Swarm Optimization for Restaurant Culinary Food Reviews

A review or opinion on culinary food restaurants carried out by consumers will produce information in the form of decision support for culinary food seekers who are looking for the best place to buy these foods. With the data in the form of review text obtained, the text mining based sentiment analysis model was obtained by using the best classification algorithm. The problem occurs while selecting attributes during preprocessing data. In this situation, the attributes that are generated are too many, and the best attributes with the best weight values need to be selected. In this paper, a hybrid model is proposed using Particle Swarm Optimization and Information Gain (PSO-IG). The proposed method is applied to 4 different methods, namely Support Vector Machine, Naïve Bayes, Decision Tree, and K-NN. Based on the results of experiments carried out on the proposed model, there was an increase in accuracy up to the highest accuracy level of 90.55%. The method using hybrid PSO-IG is a solution as an effort to increase the accuracy level of review classification of culinary food restaurants as decision support information.