Tomasz Xięski

Towards a practical approach to discover internal dependencies in rule-based knowledge bases

In this paper, we intend to introduce the conception of discovering the knowledge about rules saved in large rule-based knowledge bases, both generated automatically and acquired from human experts in the classical way. This paper presents a preliminary study of a new project in which we are going to join the two approaches: the hierarchical decomposition of large rule bases using cluster analysis and the decision units conception. Our goal is to discover useful, potentially implicit and directly unreadable information from large rule sets.

Inference in Expert Systems Using Natural Language Processing

The authors show the real life application of an expert system using queries submitted by the user using natural language. The system is based on polish language. The two stage process (involving data preparation and the inference itself) is proposed in order to complete the inference.

Density-Based Method for Clustering and Visualization of Complex Data

In this paper the topic of clustering and visualization of the data structure is discussed. Authors review currently found in literature algorithmic solutions ([3], [5]) that deal with clustering large volumes of data, focusing on their disadvantages and problems. What is more the authors introduce and analyze a density-based algorithm OPTICS (Ordering Points To Identify the Clustering Structure) as a method for clustering a real-world dataset about the functioning of transceivers of a cellular phone operator located in Poland. This algorithm is also presented as an relatively easy way for visualization of the data’s inner structure, relationships and hierarchies. The whole analysis is performed as a comparison to the well-known and described DBSCAN algorithm.

Efficiency of complex data clustering

This work is focused on the matter of clustering complex data using the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm and searching through such a structure. It presents related problems, focusing primarily on the aspect of choosing the initial parameters of the density based algorithm, as well as various ways of creating valid cluster representatives. What is more, the paper emphasizes the importance of the domain knowledge, as a factor which has a huge impact on the quality of the clustering. Carried out experiments allow to compare the efficiency of finding clusters relevant to the given question, depending on the way of how the cluster representatives were created.

Backward chaining inference as a database stored procedure — the experiments on real-world knowledge bases

In this work two approaches of backward chaining inference implementation were compared. The first approach uses a classical, goal driven inference running on the client device — the algorithm implemented within the KBExpertLib library was used. Inference was performed on a rule base buffered in memory structures. The second approach involves implementing inference as a stored procedure, run in the environment of the database server — an original, previously not published algorithm was introduced. Experiments were conducted on real-world knowledge bases with a relatively large number of rules. Experiments were prepared so that one could evaluate the pessimistic complexity of the inference algorithm.