Develpment of the VER (Visual Entity Reconciler) Tool

• Grant number: 08/10121-1
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: February 01, 2009
• End date: January 31, 2010
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Cristina Dutra de Aguiar
• Grantee: Mariângela Cardoso Miguel
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil

Abstract

In spite of the advantages achieved in data integration, the challenge of identify occurrences of different references to the same real-world entity and reconcile them is an open problem. State-of-art techniques focus on generating clusters of reconciled entities, so the entities of a cluster have a certain similarity degree and, therefore, have a high probability to be the same. However, the generated clusters face two main problems. The first one refers to the precision of the cluster, as its entities are determined using similarity functions. The second problem is related to the fact that the techniques do not generate, for each cluster, the integrated entity that best represents it. The objective of this project is to develop the VER (Visual Entity Reconciler) Tool. VER is aimed at helping users to improve the precision of the clusters of reconciled entities and to generate integrated entities. Regarding its first functionality, VER will allow users to move entities from one cluster to another, therefore grouping them in the most appropriate cluster. As for the second functionality, for each cluster, VER will allow users to generate an integrated and consistent entity from the original entities present in the cluster.