DocumentCode :
478693
Title :
Evaluation of records similarity in a duplication search engine using neural network
Author :
Paskalev, Plamen ; Nikolov, Ventsislav
Author_Institution :
EuroRisk Syst. Ltd., Varna
Volume :
2
fYear :
2008
fDate :
6-8 Sept. 2008
Firstpage :
21855
Lastpage :
23316
Abstract :
The article discusses the fine tuning and adjustment of a module for detection of duplications, which has been developed as a part of a larger online system. The problems and the current solution are discussed. In this paper a methodology to determine the influence of field similarities on the record pair similarity using a neural network is proposed.
Keywords :
neural nets; search engines; duplication search engine; neural network; online system; record pair similarity; records similarity; Artificial intelligence; Artificial neural networks; Automobiles; Couplings; Databases; Intelligent networks; Neural networks; Search engines; Telephony; User interfaces; artificial intelligence; duplications detection; intelligent user interfaces; neural networks; record linkage;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Systems, 2008. IS '08. 4th International IEEE Conference
Conference_Location :
Varna
Print_ISBN :
978-1-4244-1739-1
Electronic_ISBN :
978-1-4244-1740-7
Type :
conf
DOI :
10.1109/IS.2008.4670510
Filename :
4670510
Link To Document :
بازگشت