So Google bought MetaWeb a couple of weeks ago. That’s old news in the fast-moving world of tech company acquisitions, but very few commentators have really understood what that means for the future of the web so I wanted to write this post to further the discussion.
What Is FreeBase?
Freebase is MetaWeb’s flagship product and is the central reason for Google’s purchase.
It is often touted as a database of things AKA an
entity database and grew out of a project to add semantic data to Wikipedia articles. The result is a beautifully curated database of companies, people and events.
Freebase does provide web pages for its topics, but the real strength of the database is that it provides an RDF representation for each of its topics. This is hugely important for people building linked-data where subjects and objects are links to RDF documents.
FreeBase Has Authority
There’s much more to Freebase than just things - a big part of its database is concepts. Basic concepts like North, Aluminium, House, Kitten etc are also present. These RDF documents are the very foundations of the semantic web - an enormous number of third parties use them to describe their own entities.
So if I want to create some linked data stating that my shoes are white, I would link to Freebase’s representation of white, rather than creating my own.
Similarly, if I wanted to find a set of people who have white shoes, I would start at Freebase’s white node and traverse the link graph searching for white shoes and their owners.
All this means that FreeBase is the Wikipedia of the semantic web:
- It has lots of inbound links
- It does not link out
- It has age and human curated data
- It has authority!
What Does Freebase Mean For Google?
Google just bought a big chunk of the semantic web (relatively cheaply) with only one real competitor - DBpedia. OWL’s
sameas method of mapping entity equality pretty much takes care of any competition from DBpedia (from an indexing and linking point-of-view rather than a commercial perspective).
Freebase gives Google an instant foothold into the web of linked data and you better believe Google knows a lot about links!
As more and more web documents get enhanced with semantic markup, Google will be indexing and ranking that data. It is a good bet that the search engine is going to enhance its results using that data. I would put money on a new onebox appearing for select queries and displaying factual data much like Wolfram|Alpha.
I’m also hoping that Google will provide some new APIs that provide very fast graph traversal for all this data.
What Does It Mean For Web Publishers?
The semantic web is heating up and with all this investment from some big players I think we’ll see consumer applications emerge soon. When that happens, the linked data graph will become another SEO battleground.
Web site owners should prepare for that future by publishing linked data about their company, products and services right now (I’ve been advocating semantic search optimisation for a while now).
Build authority for your data by:
- Capitalising on your current domain authority
- Publishing accurate and timely data
- Build links to yourv entities
Some SEOs complain about Google’s love for Wikipedia but unless they start paying attention to the linked data web it will happen again.
More From Us On The Semantic Web
Creative Commons licensed photo by Eric M Martin.