Application of Latent Semantic Analysis in Accounting Research

Abstract

ABSTRACT The purpose of this study is to review a text topic modeling methodology, latent semantic analysis (LSA), and provide researchers with the requisite knowledge to allow them to learn and implement their own accounting research study using LSA. The authors first provide a brief literature review of prior business and accounting research studies that have utilized the LSA methodology. Using a provided dataset, the authors present details of how to employ LSA in a research study by replicating the mechanics used in an LSA study conducted by Hutchison, Plummer, and George (2018b). Their intent is to present thorough guidance on data selection, the analysis platform, and the necessary steps needed to conduct LSA research. This article also briefly compares LSA with other topic modeling methodologies, presents several accounting research opportunities where LSA could be utilized, and outlines LSA’s limitations. Data Availability: Data are available from the public sources cited in the text; sample dataset is available for download, see footnote 5.

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 — 3 citations exist upstream.

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