Using LDA for audit risk assessment of the Indonesian BOS fund: Insights from news analysis

Abstract

This study explores the implementation of text mining in audit risk assessment. We use the latent Dirichlet allocation (LDA) algorithm to reveal hidden topics representing risks in the management of the Indonesian School Operational Assistance Fund (BOS Fund). Using 1,460 news data points from a leading Indonesian news portal, this study proves that using text mining with the LDA algorithm effectively identifies the risks of an audit object. This study makes two important contributions to the information systems and audit literature. First, it provides evidence from online news archives to facilitate a more reliable, current, and comprehensive selection of potential audit areas by encompassing evolving social realities and facts. In the contemporary era, the accelerated and precise dissemination of information via the Internet renders the LDA approach feasible and prudent. Second, it provides a practical and applicable framework for audit risk assessment using nonfinancial sources from independent parties, which can be used as a guide for the development of audit models in the public and private sectors.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 1 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.