Ανάλυση της διαχείρισης σύνθετων συμβάντων (complex event processing) στη λήψη αποφάσεων και εφαρμογή της σε χαρακτηριστικές μελέτες περίπτωσης
Τσικερδής, Δημήτριος Σ.
Nowadays, companies need to develop methodologies and technologies that will enable them to manage their vast amount of data moving across operational procedures and that come from many sources, in order to make immediate decisions. Event Processing, which is a new area of Information Technology, seeks to solve the problem of managing information from heterogeneous sources and contribute to decision making. Τhis thesis aims to analyse this technology and whether it is an effective response to these specific needs of a company. We also investigate how this technology can be combined with Expert Systems and whether it can eventually be applied to typical problems of decision making. The structure of the thesis can be divided into two main sections. The first section presents the Event Processing and how it can be implemented through Event Driven Architecture (EDA). We then present key features of Expert Systems in order to result in a hybrid architecture that combines the EDA architecture with that of Expert Systems. In the second section we analyse how an event processing technology can be implemented through Drools, which is an open source platform from Jboss. We also present two typical case studies relating to fraud detection and managing stocks.