Compare structured and unstructured postal addresses [closed]

Question

I have a list of unstructured postal address strings and structured postal address strings. What should I use to compare these?

Example addresses:

Unstructured: john appartments 7 koramangala bangalore india 560066
structured:   7, john appartments, koramangala, bangalore-560066, india

Answer 1

If you are limited to finding the likelihood of these strings being similar then you need to look into techniques mentioned here Finding groups of similar strings in a large set of strings

Another approach - if you have access to maps/dictionaries then you can "structurize" any address (by finding the name of the country, postal code, street name, etc.) and then compare.

Good luck