Entity Cube

I just came across Entity Cube from Microsoft Asia. The researchers mention right up front that accuracy of the entity extraction and categorization still needs work, so I won’t even comment on that. (Listing “Al Roker” in the category “Academic” seemed odd, to say the least).

From the site:

EntityCube is a research prototype for exploring object-level search technologies, which automatically summarizes the Web for entities (such as people, locations and organizations) with a modest web presence…EntityCube generates summaries of Web entities from billions of public Web pages that contain information about people, locations, and organizations, and allows for exploration of their relationships. For example, users can use EntityCube to find an automatically generated biography page and social-network graph for a person, and use it to discover a relationship path between two people.

So it’s really metadata (the summaries) about metadata (the entities)–a coffee table but about coffee tables, so to speak. Looks like we have so much metadata from the web, that we need help understanding it before we can use it. And it’s all done automatically–by machine algorithms. No need for IAs or the like, right?

Well, not exactly.

Turns out there is a ton of good ol’ hand-created IA on Entity Cube, primarily in the form of faceted navigation, both on the home page as starting points into an entity as well as in results lists to group results. But the site doesn’t really do a job with the information interaction design. They could have a much better experience with the information, in general. After all, it’s not just about getting the metadata (through entity extraction) but also how it’s presented and used.

It’s interesting, though, that browsing seems to be in the foreground of the product. Right on the home page, below the search input field, there are lots of links to scan and browse. Just another trend I’ve noticed recently–browse it “in.” Is browse the new search?

There’s also a visualization tool that attempts to present a wealth of metadata in a visual display. It’s pretty good, but I suspect it would have limited use and appeal in a broader context. That is, I don’t see the average web use “getting” the visualization quickly.