PSO optimization for analysis of online marketplace products on the SVM method

The presence of online marketplace nowadays makes people shop via cellphones or laptops effortlessly. Customers are able to make transaction process easier. However, many internet users are still hesitant about choosing a marketplace. Some sellers still carry out a sale system which leads to customers’ dissatisfaction. Therefore, researchers will make a selection of the product marketplace based on reviews and comments. Researchers carry out the stage of selecting marketplace products based on an opinion or public opinion in the comments column on the online marketplace being used. From the number of comments, samples were taken which were processed and grouped into datasets. The researcher classified the data in the form of text by applying the Support Vector Machine (SVM) algorithm which would be compared using Particle Swarm Optimization (PSO). After a applied to SVM method, it produces a pretty good value, however this value may still increase as compared with PSO optimization. It is concluded that by applying the SVM method with PSO optimization the accuracy value is more maximal in processing the dataset in the form of text classification in marketplace reviews. The results of testing these data; sellers are able to find out which words are related to the sentiments that often appear and have the highest weight.