Detecting unusual email communication

Deception theory suggests that deceptive writing is characterized by reduced frequency of first-person pronouns and exclusive words, and elevated frequency of negative emotion words and action verbs. We apply this model of deception to the Enron email dataset, and then apply singular value decomposition to elicit the correlation structure between emails. Those emails that have high scores using this approach include deceptive emails; other emails that score highly using these frequency counts also indicate organizational dysfunctions such as improper communication of information. Hence this approach can be used as a tool for both external investigation of an organization, and internal management and regulatory compliance.