Analyzing sentiments on official online lending platform in Indonesia with a Combination of Naive Bayes and Lexicon Based Method

Abstract

The rapid development of social media makes people write their opinions on something, and make the data source for this research one of them from social media, Twitter, which will be analyzed in the form of sentiment analysis which is a process to understand and process data to get the information contained. In the opinion of the sentence. The trend of legal online loan/credit applications widely used by people in Indonesia is a hot topic discussed by the public on Twitter. If you want to know the tendency of public comments on online loan/credit applications, whether positive or negative, then sentiment analysis is carried out. The stages in conducting sentiment analysis in this study are data preprocessing, data processing, classification, and evaluation. The nave Bayes method is used because it has a high level of accuracy in classifying sentiments. Combining it with the lexicon can add precision to organizing emotions compared to other methods. The data used in this study were 4059 data with the distribution of 70% training data and 30% test data. Sentiment analysis obtained in this research shows that Twitter users in Indonesia give more negative comments. This study shows that by using both methods and testing using RapidMiner, the level of accuracy for classifying positive and negative sentiments achieves entirely satisfactory results, namely 82.06%, where the accuracy is higher than using only the lexicon-based saka method by 80%.