Enhanced Record Linkage Technique For Decision Making By Tree Construction |
Author(s): |
| Sathya.T , K.S.R. College of Engineering; V.Vennila, K.S.R. College of Engineering |
Keywords: |
| Data linkage, jaro-winkler, splitting criteria, pruning |
Abstract |
|
The current status in the field of record linkage is to match up the records from the various databases that do not have the common attribute. One-to-many data linkage is used for relating a record from the first database with a set of matching records from the other database which is implemented to classify the normal and the abnormal user. The proposed system implements two splitting criteria namely jaccard coefficient and jaro-winkler technique to select the best attribute at each node during the construction of the tree. A training table is trained with the two techniques and based on that test data is classified. And a pruning technique called post pruning is used to trim the tree. Thus, the proposed system improves the performance of the linkage process in terms of time and accuracy. |
Other Details |
|
Paper ID: IJSRDV2I12349 Published in: Volume : 2, Issue : 12 Publication Date: 01/03/2015 Page(s): 636-638 |
Article Preview |
|
|
|
|
