EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Hybrid Retrieval-Augmented Generation with Hallucination Mitigation for Indonesian MSME Financial Advisory

Rafi Farizki, Rani Santika, Fhrizz S. De Jesus

Journal of Management Economic and Financial · 25.04.2026 · 4 (2) · 157-169

DOI: 10.59261/jmef.v4i2.203 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Management Economic and Financial, online erschienen am 25.04.2026. Verfasst von Rafi Farizki und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Micro, small, and medium enterprises (MSMEs) are the backbone of Indonesia's economy, yet many owners face persistent gaps in financial literacy and difficulty navigating complex financial and tax regulations. General-purpose large language models (LLMs) can make such knowledge more accessible, but they often produce fluent yet unsourced or hallucinated answers, which is unacceptable in a financial and regulatory context. This study designs and evaluates IndoRAG-FA, a hybrid Retrieval-Augmented Generation (RAG) framework with explicit hallucination mitigation for Bahasa Indonesia MSME financial advisory. The framework integrates coherence-based semantic chunking; hybrid retrieval combining lexical BM25 and dense multilingual embeddings fused through reciprocal rank fusion; cross-encoder re-ranking; and grounded generation that produces source-cited answers with a confidence-based abstention mechanism and a groundedness verifier. A curated knowledge base was built from authoritative Indonesian financial and tax sources, and an expert-validated benchmark of representative MSME questions, including out-of-scope items, was developed. The framework was compared with an LLM-only baseline and a vanilla RAG baseline using retrieval metrics (Recall@k, MRR, nDCG), faithfulness, answer relevance and correctness, hallucination rate, abstention precision and recall, and a human-expert rubric. Findings indicate that a retrieval-grounded, abstention-aware design offers a practical and trustworthy path toward deployable financial advisory for Indonesian MSMEs and provides a reusable evaluation protocol for domain-specific LLM systems.

Zitieren

APA

Farizki, R., Santika, R., & Jesus, F. S. D. (2026). Hybrid Retrieval-Augmented Generation with Hallucination Mitigation for Indonesian MSME Financial Advisory. Journal of Management Economic and Financial, 4(2), 157-169. https://doi.org/10.59261/jmef.v4i2.203

Vancouver

Farizki R, Santika R, Jesus FSD. Hybrid Retrieval-Augmented Generation with Hallucination Mitigation for Indonesian MSME Financial Advisory. Journal of Management Economic and Financial. 2026;4(2):157-169. doi:10.59261/jmef.v4i2.203

BibTeX
@article{Farizki2026,
  author = {Farizki, Rafi and Santika, Rani and Jesus, Fhrizz S. De},
  title = {Hybrid Retrieval-Augmented Generation with Hallucination Mitigation for Indonesian MSME Financial Advisory},
  journal = {Journal of Management Economic and Financial},
  year = {2026},
  volume = {4},
  number = {2},
  pages = {157--169},
  doi = {10.59261/jmef.v4i2.203}
}
Kostenlos starten →