Title :
A classification algorithm based on multi-relation domain knowledge
Author :
Hu, Xue-Gang ; Hu, Xie-Fei ; Wang, De-Xing ; Zhang, Dong-Yan ; Hu, Chun-Ling
Author_Institution :
Sch. of Comput. & Inf., Hefei Univ. of Technol., China
Abstract :
Recently, integrating domain knowledge in knowledge discovery has engaged many experts´ attention. Various kinds of domain knowledge can take on a number of different forms and have a great diversity of effects in knowledge discovery in database. Concept lattice, which has a complete form in knowledge representation, can represent the complex relations among the objects in the world. In the paper, incorporating multi-relation domain knowledge in knowledge discovery is proposed and a classification algorithm CS_MRDK that using concept lattice for discovering the closest relationship among attributes is presented, which shows the good performance of using multi-relation domain knowledge in knowledge discovery.
Keywords :
data mining; knowledge representation; very large databases; classification algorithm; concept lattice; data mining; knowledge discovery; knowledge representation; multirelation domain knowledge; very large databases; Algorithm design and analysis; Classification algorithms; Data mining; Databases; Electronic mail; Explosives; Information technology; Knowledge representation; Lattices; Machine learning; Classification; Concept lattice; Domain knowledge; KDD;
Conference_Titel :
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on
Conference_Location :
Guangzhou, China
Print_ISBN :
0-7803-9091-1
DOI :
10.1109/ICMLC.2005.1527286
