Analyzing User Comments on Fintech E-wallet Applications in Indonesia Using the LDA Algorithm

The main objective of this article is to compare user comments from 4 Fintech applications in Indonesia using the Latent Dirichlet Allocation (LDA) algorithm. This research is word processing research carried out using text mining. The statistical population is comments from Fintech users Ovo, Dana, LinkAja, and Doku, with data taken from May 2022-May 2023. LDA and the Python programming language are applied to analyze the data and implement text mining algorithms from topic modeling. Each application group has different topic coherence, each application group is analyzed and 4 large categories are found, namely system convenience, system problems, customer service (CS) problems, and service ease. The results of text mining and processing using Coherence and Latent Dirichlet allocation can be used to identify hidden topics and be used as input for system improvements