Question

I have a database which has very similar rows within the same table. Those rows are similar because they have nearly equal column values. I need to integrate those corresponding rows into one single row.

For example, those two users (u1 and u2) should be integrated:

 u1 = User(name = "William Henry Gates III",
           age = 55,
           nationality = "american",
           alma_mater = "Harvard Univesity")

 u2 = User(name: "William Henry 'Bill' Gates III",
           age: 55,
           nationality: "America",
           alma_mater: "Harvard U.")

I am thinking of using some edit distance and stemming techniques. Other algorithms and techniques suggestions? Any helpful libraries to use (preferably in Python or Java)?

Answer 1

Considered something like Refine?