Sentiment Analysis on Online Transportation Service Products Using K-Nearest Neighbor Method

Abstract

In 2018, the everyday client of online transportation administrations is up to 8 million clients/day in the Southeast Asian locale. Indonesia is the biggest country in utilizing the administrations contrasted with other Southeast Asian nations. The clients present their reactions to the administrations given by online transportation specialist co-ops through different media through the Instagram remarks segment. The responses submitted have additionally differed, so they likewise contained opinions that communicated their sentiments about specific administrations. In this manner, an opinion investigation framework was made utilizing the K-Nearest Neighbor strategy to decide clients' reactions to the administrations given, the degree of fulfillment, and assist clients with picking the best administrations. This examination shows that the proposed method can characterize client assessment with 94.4% accuracy, and for recall and precision, the F1 Score has the same result, 94.4%. This infers that from assessment results, the proposed calculation performs well to dissect opinion consequently.