Dear Lazyweb,

it seems that one of my next challenges is to solve a street address normalization problem. I found a few commercial solutions that seem to be overkill and server based (coupled with credit card verification usually), an open source one that has many shortcoming (including being written in PHP) but nothing that seems clean, simple and in Java. It doesn’t need to be open sourced but it would be nice.

They’d need to deal with things such as:

• 4905 Bonbon Lane 4 / 4905 Bonbon Lane #4 / 4905 Bonbon LN
• 409 Natlie Drive / 409 Natalie Drivee /409 Natalie Dr / 409 Natalie Drive
• 520 Seymour Ave Unit 2 /Unit 2 520 Seymour Road / Room 2 520 Seymour Garden / 520 Seymour Way Rm 2

And things of that sort. I realize it might be impossible, that it might rely on a US/CD address database but at this point, I’d like to figure out what all the solutions could be.

Any idea ?

3 people like this post.

Unlike