摘要
Reducing inconsistency is the key problem to improve data quality during data integration. In this paper,we first present a weighted algorithm of similarity coefficient which is superior to traditional algorithms if the sourcedata have multiple characteristic items ,all of which have to be taken into account ,especially during the complex infor-mation integration. Secondly,we apply it to the experiment of telecommunication customers integrating ,the results ofdata clustering show it has high feasibility and precision performance.
Reducing inconsistency is the key problem to improve data quality during data integration. In this paper, we first present a weighted algorithm of similarity coefficient which is superior to traditional algorithms if the source data have multiple characteristic items, all of which have to be taken into account .especially during the complex information integration. Secondly,we apply it to the experiment of telecommunication customers integrating,the results of data clustering show it has high feasibility and precision performance.
出处
《计算机科学》
CSCD
北大核心
2003年第8期92-92,F004,共2页
Computer Science