Mining Organizational Structures from Email Logs: an NLP based approach

Exchanged emails of both personal and business contexts are among the most information sources that are particularly useful in Business Process Management (BPM). Thus, the information flux given by exchanged emails represents an essential part of multi-actor cooperation to achieve any required process. Email contents may be extracted and processed to understand the interactions between workflow actors. So, we aim to take advantage of the exchanged emails to meet the challenge of Organizational Structures (OS) mining in a workflow. By OS, we mean the social structures which define the interactions' semantic of an actor's group involved in the workflow (namely federation, coalition, market, or hierarchy) for activity distribution driving. In this paper, we propose an agent-oriented approach to characterize each interaction between two actors thanks to the social abilities of agents. Then, we show an analytical framework for business-oriented information harvesting and classifying extracted from email bodies to support the OS mining. The feasibility of our approach is proved by experimentation based on the open Enron email dataset. (C) 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://crativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of KES International.