Data Mining of Electronic Mail and Auditing: A Research Agenda

Abstract

ABSTRACT A potentially important form of audit evidence is corporate email, which includes the background, intent, and outcome of managerial actions. Emails are semi-structured data, with known fields for sender, recipient, subject, and date, as well as the email body and attachments. Data mining of emails (DME) combines understanding of social networks, textual analysis using natural language processing and other techniques, and domain knowledge. Email data mining research received a considerable boost by the availability of a large archive of emails from the now-defunct Enron Corporation. The paper provides an overview of DME techniques, and the potential expanded applications of DME by auditors. The paper also discusses a set of research areas and research opportunities.