Domain ontology driven data mining: a medical case study

ABSTRACT

This paper reports a domain ontology-driven approach to data mining on a medical database containing clinical data on patients undergoing treatment for chronic kidney disease. Each record within the dataset is comprised of a large number (up to 96) of quantitative and qualitative metrics which represent the physiological state of a particular patient on a particular day of treatment. One of the challenges of mining such a dataset is that the meaning of many of the metrics/attributes is not easily understood by someone who is not familiar with the domain of kidney disease and treatment, and it is not clear which of the attributes are useful in data mining.

This paper explores the possibility of utilizing a medical domain ontology as a source of domain knowledge to aid in both extracting knowledge and expressing the extracted knowledge in a useful format. We describe an approach in which the domain ontology is used to categorize attributes in preparation for mining 'association rules' in the data; the mined rules were then reviewed by comparison to domain knowledge derived from a domain expert in order to gauge their 'usefulness'. We conclude that domain ontology driven data mining can obtain more meaningful results than naïve mining.

REFERENCES

• 1: H. Cespivova, V. Svatek, and M. Tomeckova. Roles of medical ontology in association mining CRISP-DM cycle. In Knowledge Discovery and Ontologies (KDO) at ECML/PKDD, 2004.
• 2: Y.-L. Chen and S. yi Wu. Mining temporal patterns from sequence database of interval-based events. In L. Wang, L. Jiao, G. Shi, X. Li, and J. Liu, editors, FSKD, volume 4223 of Lecture Notes in Computer Science, pages 586--595. Springer, 2006.
• 3: P. Domingos. Toward knowledge-rich data mining. Data Mining and Knowledge Discovery, published online, April 2007.
• 4: Szymon Jaroszewicz , Dan A. Simovici, Interestingness of frequent itemsets using Bayesian networks as background knowledge, Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining, August 22-25, 2004, Seattle, WA, USA [doi>10.1145/1014052.1014074]
• 5: D. Kudenko. Ontology-based constructive induction. In International Conference on Machine Intelligence, ICMI, 2000.
• 6: Joseph Phillips , Bruce G. Buchanan, Ontology-guided knowledge discovery in databases, Proceedings of the 1st international conference on Knowledge capture, October 22-23, 2001, Victoria, British Columbia, Canada [doi>10.1145/500737.500758]
• 7: J. Rauch. Interesting association rules and multi-relational association rules. Communications of Institute of Information and Computing Machinery, 5(2):77--82, 2002.
• 8: E. Winarko and J. F. Roddick. Discovering richer temporal association rules from interval-based data. In A. M. Tjoa and J. Trujillo, editors, DaWaK, volume 3589 of Lecture Notes in Computer Science, pages 315--325. Springer, 2005.
• 9: Dong Xin , Xuehua Shen , Qiaozhu Mei , Jiawei Han, Discovering interesting patterns through user's interactive feedback, Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining, August 20-23, 2006, Philadelphia, PA, USA [doi>10.1145/1150402.1150502]