Uncovering Hidden Sentiments and Topics in Online Lending Application Reviews with the Valence Aware Dictionary and sEntiment Reasoner (VADER) and Latent Dirichlet Allocation (LDA) Approaches

Abstract

Online lending (pinjol) has become an important part of the digital transformation of
the financial sector, offering people easy access to funds. However, the increasing reliance on
user reviews as a decision-making factor raises concerns about their authenticity and credibility.
This research aims to analyze the sentiments and topics that appear in the reviews of Akulaku,
Kredivo, and EasyCash lending apps on the Google Play Store. Using text mining techniques,
VADER sentiment analysis, and LDA topic modeling, this research reveals dominant positive
sentiments related to ease of use, service speed, and customer support. However, there were also
negative reviews regarding loan application difficulties, technical issues, and bad experiences
with billing and payments. This research provides valuable insights into the preferences and
concerns of pinjol users, which can serve as a reference for service providers to improve the
quality of their products and services.