Open AccessReview Analisis Produk Marketplace Online pada Algoritma Support Vector Machine
Author(s) -
Rizqi Agung Permana,
Sucitra Sahara
Publication year - 2021
Publication title -
jurnal ilmiah informatika
Language(s) - English
Resource type - Journals
eISSN - 2549-7480
pISSN - 2549-6301
DOI - 10.35316/jimi.v6i1.1227
Subject(s) - computer science , support vector machine , popularity , product (mathematics) , the internet , service (business) , value (mathematics) , key (lock) , sentiment analysis , selection (genetic algorithm) , process (computing) , public opinion , world wide web , artificial intelligence , machine learning , marketing , business , computer security , psychology , social psychology , geometry , mathematics , politics , political science , law , operating system
With the presence of a marketplace, internet users are able to shop by cellphones or laptops which ease them to process the transactions. This advantage is the key to the popularity of e-commerce and marketplaces in the 21st century. Considering that users prefer having e-commerce to going out and looking for products. Internet users are still hesitant in choosing a good and bad marketplace since there is a disappointing marketplace system and a deficient service. The Researcher will make a selection of the product marketplace based on opinions or review comments. The stage of selecting marketplace products is based on an opinion or public opinion who buys goods on the selected marketplace, the opinion is given into online opinions such as comments on the marketplace. The comments that are used as samples will be processed and grouped into data sets. In this case, the researcher processes the classification or grouping of data by applying the Support Vector Machine (SVM) method in accordance with the classification of processing data sets in the form of text. After being applied to the SVM method, the accuracy value is 75.92%, the SVM method produces a fairly good accuracy value on the dataset in the form of text classification in the marketplace review. From the results of testing the data, it can help readers in determining a good marketplace.