User Sentiment Analysis in the Fintech OVO Review Based on the Lexicon Method

Abstract

User reviews are important in the new approach to fintech services. To learn this information, a simple sentiment analysis can make the right observations to support the OVO fintech system in analyzing the success of the fintech system. The analysis has several stages, starting from how to extract comment data from the play store, extracting meaningful information from the play store platform, and extracting the data into valuable information. Moreover, accurate topic modeling and document representation is another challenging task in sentiment analysis. We propose a lexicon-based topic modeling in observing user sentiment simply by looking at the number of words that appear. The proposed system retrieves OVO fintech comment data from the Play Store, removes irrelevant content to extract meaningful information, and generates topics and features from the extracted data using NLTK. Data processing using google collab in Python language where data is used freely. Data analysis using the word cloud method, Exploratory Data Analysis (EDA), correlation analysis between words, ordering the number of words in sentences revealed that OVO comments in that period tended to be negative