Abstract: To ensure the effectiveness of decision-support computer systems it is essential to solve such problems as identification, filtration, precision etc. of information streams, as well as modeling and simulation of decision-making problems which are based on them. When working with information streams of expert knowledge, as a complex systems, in parallel with classical approaches of their modeling, the most important matter is to assume fuzziness. All these is connected to the complicity of study of incomplete, abnormal and extreme processes in nature and society, which are caused by lack or shortage of objective information and when expert data streams are essential for constructing credible decisions. Such problems include solutions of business problems in extreme environments, analysis of management and investment risks, problems of conflictology, sociology, medical diagnosis, etc. With the growth of complexity of information our ability to make credible decision about process development reduces to some level, below which some characteristics such as accuracy and certainty become mutually conflicting. Our research is concerned with quantitative-fundamental analysis of this uncertainty and its use for precision of informational processes and decision modeling. Consequently one of main objects of our attention is the analysis of structures of expert data and measures of its uncertainty. The most important of such analysis methods are the theory of the body of evidence.
The precision of decisions first of all means improvement of representation of decision making factors by Dempster-Shafer data structures. Of course, there are many methods for knowledge representations and decision making, which use the Dempster-Shafer structures. The novelty of our research in this direction is the technology for precision of the structure of body of evidence, which we call the temporalization of body of evidence. Temporalization means the construction of inclusion relation on the bodies of evidence. This approach is completely novel in study of expert knowledge representations and structuring. It will cause many heurstic methods of decision- making based on the expert knowledge representation to be modified. All above listed means the following: 1. representation of data which is an input of considered methods using Dempster-Shafer structures, so called pessimistic-optimistic representations. This will better exhibit the knowledge and intellectual activities of an expert. 2. the possibilities of representing of expert information streams in triangular or trapezioform fuzzy numbers will be considered. 3. the cases where focal elements in Dempster-Shafer structure are represented by fuzzy sets, and focal probabilities are represented by triangular or trapezioform fuzzy numbers will be considered separately, 4. in methods’ decision-making criteria represented in knowledge base of decision support technologies of inaccuracy and uncertainty aggregations will be used such as: Choquet integral, Sugeno integral, Dempster upper and lower expected values, positive and negative discriminations, OWA operators, etc. 5. in selected methods these aggregations will give us new criteria supporting more precise decision. Thus existing heuristic methods will obtain fundamental basis, final purpose of which will be to model more precise decision in the cases of expert knowledge streams input. 6. The decision support system will obtain higher credibility, which can be measured in new modified methods using the informational measures, such as confusion and chaos constructed on more precise decisions, inaccuracy and non-specificity measures etc.
Finally the process of precising decisions will be demonstrated based on Discrimination method which is one of the popular methods of decision making using fuzzy set theory.

Brief Biography of the Speaker:
Dr. Gia Sirbiladze is a full professor at the Department of Computer Science of Faculty of Exact & Natural Sciences of Iv. Javakhishvili Tbilisi State University, Georgia. He received his Ph.D. degree in 1991 from the Computational Mathematics Institute of the Georgian Academy of Science. He received his D. Sci. degree from the same institute in 2005. His scientific interests include areas such as Intelligent Fuzzy Technologies and General Systems, Fuzzy Technologies in Decision-making Support Systems, Fuzzy Extremal Dynamic Systems - Control, Filtration and Identification, Fuzzy Discrete Optimization Problems and Modeling Decisions. Dr. Gia Sirbiladze has published 54 scientific papers on the above-listed topics. He is an author of one monograph on Decision Making Problems in General Environment. Dr.Gia Sirbiladze has participated in many scientific conferences, including plenary speeches on WSEAS conferences. Dr.Gia Sirbiladze is a member of the National Union of Mathematicians in Georgia. He serves as a reviewer for Mathematical Reviews. He has reviewed papers for more then 15 international and local journals and conferences. He serves as Information Technology expert for Georgian National Scientific Fund. Dr.Gia Sirbiladze has participated in several national and international research projects.