创建一个解决方案，使用 python 将地址自动拆分为单独的组件

Question

I am trying to find a solution for being able to automatically split address into their separate components using python. below is some sample data

Full Address	Street Number	Street	City	State	Zip Code
661 Camel Back Road Tulsa Oklahoma 74120	661	Camel Back Road	Tulsa		Oklahoma
68 Gnatty Creek Road Roslyn New York 11576	68	Gnatty Creek Road	Roslyn		New York
1 Raccoon Run Seattle Washington 98119	1	Raccoon Run		Seattle	Washington
616 Friendship Lane Santa Clara California 95054	616	Friendship Lane	Santa Clara	California	95054
3878 Grand Avenue Maitland Florida 32751	3878	Grand Avenue	Maitland	Florida	32751

The above data is a representation of what I am trying to achieve. on the left is my input address, and on the right is the result after having being split out automatically. The problem here, as cannot be seen in this over simplified example, is that the input addresses don't come in the same order, and will include components such as names of buildings etc.

My options so far are the following:

1. REGEX
2. MACHINE LEARNING MODEL

The REGEX option is familiar, but it will still be largely inaccurate. I need this solution to be as accurate as possible.

The MACHINE LEARNING MODEL option is more difficult in that I am not aware of any model or framework capable of classifying multiple categories as once. Can anyone help?

so far I haven't really started the REGEX in anticipation of major gaps in capturing groups.

Answer 1

I think the only way to do this and get a fairly accurate result is to get the list of zip codes, for instance from here: https://www.zipcode.com.ng/2022/06/list-of-5-digit-zip-codes-united-states.html?m=1 and a list of US cities.

Then you can match the zip code, state and city to the lists.