A tool supporting investigative procedures using automatic inference

Project coordinator:

• Professor Czeslaw Jedrzejek – Poznan University of Technology.

The institution performing the project:

• Poznan University of Technology, Department of Electrical Engineering, Institute of Control and Information Engineering.

Project objective:

The objective of this project is to develop a prototype AFAC (Analyzer of Facts and Connections) system supporting analysis of evidence.

Project description:

The project will deliver an expert system for police analysts. The system will use a knowledge base constituting an interface between a human and information. The knowledge base will present knowledge in various fields in a semantic way (i.e. by using ontology in a way similar to descriptions in a natural language used by humans) and will use facts, available information entered into the system, and rules.

Using rules, along with ontologies, should make it possible to discover knowledge that would be difficult to obtain if only queries to commonly used relational databases were utilized. An illustration of this idea would be a closet with our clothes (which corresponds to a relational database). In this example, a single clothing item corresponds to a piece of information. After the request, “Give me a blue, long sleeve shirt”, the system will check if such a shirt is in the closet and take one. A knowledge base is required to fulfill the request, “Give me clothes for a three day kayaking trip”. The system has to understand what clothing is required for this kind of a trip. The closet (similar to a database) must first be loaded, both at the conceptual stage, i.e. by deciding how things are to be arranged, using the tools of semantic description of the reality, and then a decision must be made on purchasing the specific items that are missing.

The procedures and the practice of using the system will be elaborated in cooperation with the police. The functionality of the system will comprise many scenarios of possible economic crimes, as well as a strategy how to look for additional data in systems operated by the police (as well as certain other systems), for the purpose of economically effective collection of evidence.

Project financed by The National Centre for Research and Development