Techniques for Duplicate Detection in Hierarchical Data

Suvarna Kale, Basha Vankudothu

Abstract: Duplicate detection is nothing but finding multiple representations of a same object and also object which are represented in a dataset. The duplicate detection is important to integration and data cleaning applications and it is studied for relational data in single table, but now data is stored in complex form. In this paper we improve the efficiency and effectiveness of duplicate detection by considering relationship between ancestors and descendants. We apply this strategy by implementing two algorithms RECONA and ADAMA. Recona re-examine an object if its induce neighbours is duplicates. This will reduce re-comparison of elements. Adama is efficient because it does not allow re-comparison

Keywords: Duplicate detection, XML data, hierarchical structure, candidate pair