We propose an incremental technique for discovering duplicates in large databases of textual sequences, i.e., syntactically different tuples, that refer to the same real-world entity. The problem is approached from a clustering perspective: given a …