Article

Latent Dirichlet Allocation and Sentiment Analysis of Global FinTech Discourse: Mapping Innovation Narratives in the Digital Economy

---

Abstract

The rapid expansion of Financial Technology (FinTech) has reshaped the global digital economy, yet empirical understanding of how FinTech innovation is narratively constructed in global discourse remains limited. This study applies an integrated Latent Dirichlet Allocation (LDA) and sentiment analysis framework to systematically map dominant innovation narratives and their affective orientations across large-scale FinTech-related textual data. Using a curated global corpus comprising industry reports, policy documents, and financial media content, the analysis identifies eight dominant thematic clusters, including digital payments, financial inclusion, open banking, artificial intelligence–driven credit analytics, blockchain, cybersecurity, regulatory technology, and central bank digital currencies. The results demonstrate that consumer-facing innovations, particularly digital payments and financial inclusion, account for the largest share of discourse and exhibit consistently positive sentiment, indicating strong narrative legitimacy and adoption confidence. In contrast, governance- and risk-oriented domains, such as cybersecurity and RegTech, display lower prevalence and persistently negative or mixed sentiment, reflecting problem-centered framing and institutional caution. Temporal analysis further reveals that market-driven narratives are relatively stable over time, whereas policy- and security-related narratives are highly volatile and event-contingent. Topic–sentiment interaction analysis confirms the absence of a linear relationship between visibility and sentiment, highlighting systematic narrative polarization within FinTech discourse. These findings provide empirical evidence that FinTech innovation is shaped not only by technological and economic factors, but also by discursive dynamics that influence legitimacy, trust, and governance priorities in the digital economy. By offering a comprehensive topic–sentiment map of global FinTech narratives, this study contributes a novel analytical perspective for researchers, policymakers, and industry stakeholders seeking to understand and manage the socio-technical evolution of digital finance.

Keywords: Financial Technology, Digital Economy, Topic Modeling, Latent Dirichlet Allocation, Sentiment Analysis, Innovation Narratives, FinTech Governance

How to Cite:

Eugie, L. L. & Beatrice, L. M., (2025) “Latent Dirichlet Allocation and Sentiment Analysis of Global FinTech Discourse: Mapping Innovation Narratives in the Digital Economy”, FinTech Innovation Journal 1(4), 388-407. doi: https://doi.org/10.63913/ftij.v1i4.87