Semantic Profiles of Internal Corporate Communications Analysis for Subject-Oriented Business Process Modeling

<p align="center"><strong>Abstract of the Master Thesis &ldquo;Semantic Profiles of Internal Corporate Communications Analysis for Subject-Oriented Business Process Modeling&rdquo;</strong></p><p>In the thesis I investigate whether the corporate e-mail data can be used for the subject-oriented business process modeling. I suggest an algorithm that analyzes the internal corporate communications and detects business processes, implement this algorithm in software, and apply it to the e-mail data corpus of Enron. Corp, which is freely available for research.</p><p>The thesis has the following structure: the introduction, the related work section, the description of the available input data, the research methods and results, the conclusion, the list of references, and part of the program code.</p><p>In the introduction I set the research goals and show their relevance in the theoretical and practical contexts.</p><p>In the related work section I review the existing methods for the corporate email analysis and the main approaches to the Process mining. I also present the basics of the subject-oriented approach to the business process modeling.</p><p>The input data section consists of two parts. First, I give information about Enron Corporation, whose e-mail data corpus is analyzed.&nbsp; Then I describe the method of corporate e-mail data indexation that is used to get the semantic profiles of the Enron e-mail.</p><p>The research methods and results are described as follows. First, I present an algorithm for business process search, which is based on the statistical analysis of the indexed communication data. Then I describe an implementation of this algorithm in software. Finally, I explain how the algorithm was applied to the Enron data and demonstrate a few real business processes found by the algorithm. This proves the validity of the algorithm and confirms that the email data is indeed suitable for business process modeling.</p><p>In the conclusion I summarize the results and discuss their relevance for the future work. Finally, I give the list of references and provide the source code of the algorithm in the appendix.</p>