Outlier Mining in Rule-Based Knowledge Bases
The paper presents the problem of outlier detection in rule-based knowledge bases. Unusual (rare) rules, regarded here as deviation, should be the subject of experts’ and knowledge engineers’ analysis because they allow influencing on the efficiency of inference in decision support systems. A different approaches to find outliers and the results of the experiments are presented.
Keywordsoutliers cluster analysis rules knowledge bases inference process efficiency
Unable to display preview. Download preview PDF.
- 2.Koronacki, J., Cwik, J.: Statistical learning systems. WNT, Warszawa (2005) (in Polish)Google Scholar
- 3.Nowak, A.: Complex knowledge bases: the structure and the inference processes, PhD thesis, Silesian University, Katowice, Poland (2009) (in Polish)Google Scholar
- 4.Nowak-Brzeziska, A., Wakulicz-Deja, A.: The choice of similarity measure and the efficiency of clustering rules in complex knowledge bases. Studia Informatica 31(2A(89)), 189–202 (2010) (in Polish)Google Scholar
- 5.Nowak-Brzeziska, A.: Mining knowledge and the effectiveness of decision support systems. Studia Informatica 32(2A(96)), 403–416 (2011) (in Polish)Google Scholar
- 6.Pearson Ronald, K.: Mining imperfect data - dealing with contamination and incomplete records, pp. I–X, 1–305. SIAM (2005)Google Scholar
- 7.Seo, S.: A Review and Comparison of Methods for Detecting Outliers in Univariate Data Sets. University of Pittsburgh (2006)Google Scholar
- 8.Cherednichenko, S.: Outlier Detection in Clustering, University of Joensuu, Department of Computer Science, Master’s Thesis (2005)Google Scholar
- 9.Hawkins, D.: Identification of Outliers. Chapman and Hall (1980)Google Scholar
- 10.Pawlak, Z., Wiktor, M.: Information storage and retrieval system - mathematical foundations. Computation Center Polish Academy of Sciences (CC PAS), Warsaw (1974)Google Scholar
- 11.Breunig, et al: LOF: Identifying Density-Based Local Outliers. In: KDD (2000)Google Scholar