The 95% Typed Challenge

Freebase contributor spencermountain made a great suggestion the other day on our developer mailing list: that our next big goal should be to have 100% typed topics in Freebase, and have a big party when we get there.

100% is hard — really hard — because there will always be an odd selection of topics that we don’t have types for, or that are new and not yet typed, or whatever. But 95% could be achievable, with time and dedication. And so, the 95% Typed Challege was born.

Here’s the deal: I’ll be blogging regularly about the status of typed topics in Freebase. Typed topics will be defined as those that have at least one Commons type other than /common/topic. I’ve got a public google spreadsheet that shows recent progress and current status. We’re at about 80% typed at the moment, which means that one in five Freebase topics is untyped.

typed

To help us get more typed topics, we’ll run data mobs to work on particular areas of untyped topics. Let’s start with three today!

  • Association (5072 instances) — most of these look like Organizations.
  • Roman Catholic Diocese (1699 instances) — most of these are Religious Jurisdictions
  • New Zealand (1348 instances) — these are a range of things, from governmental bodies to railways to locations

There are a few tools available to help you type these. The first is tfmorris‘s Untyped Topics application. This simply lists 100 untyped topics that match any string, and you can then do what you like with them, like open each in a tab and use the Freebase.com website to “Add type”.

If you know that most of the topics will end up being of the same type (as in the diocese and association examples above), you might prefer spencermountain‘s Keyword Typer. This allows you to specify a string to match on as well as a likely type. You’ll then be asked to confirm or skip the typing for each topic.

If you’ve got a few minutes to spare, please go type a few topics! As each of these areas is completed, we’ll post a new one.

Tags: ,

Comments are closed.

About

Freebase is a free database of the world's information. This is the official Freebase blog.