A study on implementation of automated deci-sion process into the information systems

Abstract

Many studies on software project failures have identified difficulties in capturing requirements, managing complexity and dynamic changes of the environment influenced by the changes of evolving social and competitive business environment around us. This makes static capture of requirements and long implementation according to the traditional software engineering leading to the problems that can be solved by using models. This is especially important for decision-making, which is such a dynamic sphere of business. Business people who are first of all directly responsible for the results of automated business processes are left isolated from directly impacting the rules in automated decision-making processes, which are usually hidden in the code. This can be tolerated in case of simple operational decisions; however, it becomes a real problem when trying to automate more abstract tactical and strategic decisions based on complicated and unstable business logic and rules. In the presented thesis we offer modernisation of information system development methods used for implementation of automated information-, rule-, knowledge- and model-based decision processes assisted by early separation and development of a business logic model and implementation of decision-making and knowledge discovery process models with further support to business people with suitable interfaces for modification of decision-making processes without involvement of software developers in later exploitation stages. In the analytical part of the dissertation, decision models and methods of intellectualised information systems are analysed. According to the results obtained during the analysis, investigation on the proposed framework and a decision model based method for decision-making process automation is carried out. In the final chapter several experiments are described in order to evaluate the proposed method, and the general conclusions complete the research.