Qualitative analysis of financial statements for fraud detection

Fraudulent financial reporting has become a global phenomenon which is adversely affecting the economic and social growth of an organization and the nation they belong to. The financial statements published periodically by the companies contain quantitative and qualitative information. The quantitative information includes numbers, metrices, ratios etc. and qualitative information comprises of comments/notes given by the auditors, disclosures by the management in textual format. A number of analytical models have already been applied for finding the possible solution for financial statement fraud by using quantitative information. A little or no work has been done for detecting fraud by analyzing the qualitative information present in financial statements. In this paper, we propose a text mining framework for detecting financial statement fraud by identifying and analyzing the linguistic data available in the financial reports.