“…The work on RL can be broadly classified into three categories: (i) effective RL, (ii) optimal selection of similarity measures, and (iii) efficient RL. The works in (i) [4,5,10,11,17,32,37,38] employ a broad range of machine learning techniques such as decision trees, SVM, logistic regression, correlation mining, and clustering. In (ii), the goal is to automatically select optimal similarity functions [7] for each attribute of an entity (e.g., using edit distance for the attribute phone and Jaccard distance for name) and determine similarity thresholds [31].…”