Mining Enterprise Models for Knowledgeable Decision Making

Knowledge is stored in an enterprise in various forms ranging from unstructured operational data, legal documents to structured information like programs, as well as relational data stored in databases to semi-structured information stored in xml files. All these information if viewed from a holistic standpoint can help an enterprise to understand and reflect upon itself and thereby make knowledgeable decisions whenever required. In order to satisfy this objective of holistic knowledge representation and decision making, we begin with mining unstructured information present in an enterprise. In particular, in this paper, we intend to mine a document intensive business processes and extract information as a knowledge repository that captures the various stakeholders along with their intentions and the tasks they perform. The goal is to automate the validation of such business processes by eliminating any manual verification, which is time consuming and error prone. We believe this is the first step towards realizing our broader objective of collective modeling of enterprise knowledge that will involve mining of information available in unstructured, structured, relational as well as semi-structured form present in an enterprise.