Mining Email Transaction Logs Using Multi-Digraphs
Vatcharaporn Esichaikul, Sumanta Guha · INTERNATIONAL JOURNAL ON Advances in Information Sciences and Service Sciences · 2011
With the tremendous volume of email messages, it is difficult to determine significant messages and users in an email system without knowing the message contents. This paper presents a simple and intuitive model to mine an email transactions log for significant messages and users. No use is made of NLP or semantic analysis. The model is based only on scoring messages and users from a graph-theoretic analysis of the communication pattern represented in the transaction log. Practical experiments indicate the potential of the model.