Development of data reduction engine based on computational intelligence

The main aim of the project is an investigation of the most suitable scientific framework and development of software tools based on computational intelligence for multi-dimensional data redaction and their implementation in real problems. The problem of patent portfolio analysis will be considered as a primary example for the implementation of data reduction algorithms. The main objective of an automatic patent text analysis is to obtain fundamental understanding of company technological areas of activity, as well to gain an overall insight into trends and state of that particular market. The second test case for the implementation of devised methodologies and tools will be data reduction for video quality assessment where there is an emerging need to reduce multi-dimensional space of objective video quality measures to smaller space which is easier to use and interpret. The next test case for the implementation of adopted methods will be project for automatic node localization in complex wireless sensor networks. In this test case the possibility of sensors position estimation based on collected measurements is of the special importance, especially with minimal energy consumption and through utilization of low cost sensors. As main project outcome it is expected to develop software for intellectual property (IP) analysis which will also be tested on the system for video quality assessment and as a tool for node localization in large ad-hoc wireless sensor networks.