Fraud Detection in Financial Statements Using Classification Algorithm
Vasudha Sarda, Prasham Sakaria, Dharmeshkumar Mistry, Devanshi Sanghvi · 2014
There have been many high profile companies whose fraudulent financial statement is being telecasted widely. Data mining techniques have been used to detect such fraud in the financial statements. These extensive techniques have mostly considered only the quantative part of the financial statement like the financial rarios but there has been very less usage of the qualitative information present for classifying the financial statement as fraudulent. There is very little research on the analysis of text such as auditor's comments or notes present in published reports. Mining of the textual data can be used to detect fraud using text mining techniques. Support Vector Machine algorithm is used to analysis whether internal audit reporting structure and internal audit sourcing arrangement affect the ability of the internal audit function to prevent financial statement fraud or detect if the statements of a company are fraudulent in the financial statement reporting process.