Domain name classification API

All the systems that I am aware of manage a list, somewhat by hand.

Using a web-filtering proxies (e.g. WebSense) for inspiration, you could scan for keywords contained in the domain name, or in web content/meta tags at the specified location. However, there are always items that seem to match more than one category, or no category, and these need deeper analysis.

Eventually you end up building your own fairly complex logic, maintaining a list by hand, or buying a list from someone else.

SimilarWeb API does that.

It's really straight forward and returns a given domain's category from a URL.

If these are new domains or not used domains. There isn't any information on the internet yet. You can make use of a mechanical turk, like: https://www.mturk.com/ .

You could post an task with your list and possible categories. The downside is this will cost you money.

If these are domains that are already in use you can use a bookmark service as xmarks or delicious. Retrieve all public bookmarks from that domain and count the number of tags. The most used tags will indicate a category of the domain.