Google Base works around the lack of a Semantic Web

by Dan Zambonini

Related link:

Most developer blogs will probably be mentioning Google Base today. If the screenshots are to be believed, it looks like a behemoth roll-your-own web database service, but without the low-level worries of normalisation, data types, or anything else. The GUI looks simple enough that anyone with an idea for listing and searching structured data (apartments, classifieds, parties, businesses, etc.) can get up and running without any technical knowledge.

Could this be another blow for the development of the Semantic Web? One of the powerful side-effects from aggregating Semantic Web data is the ability to map/relate information, no matter what type of information it is, or where it comes from. Google appears to be bypassing the need for users to individually publish their data (in RDF), by asking them to create their data records directly into the Google database. Data within the Google Base can therefore be explicitly related (if desired), negating the potentially laborious normalising/mapping of distributed semantic web data.

Is this likely to evolve into a Semantic Web application? Are Google recording this data in an RDF data store? (I doubt it) Or will they expose a Semantic Web friendly API (e.g. SPARQL), or publish the data as RDF? It's unlikely that the data will be easily exposed as full recordsets - the value of the data is key to the application, to share it could be dangerous.

So, is Google Base another blow for the Semantic Web, or will it highlight the potential power that a real Semantic Web could offer?

What are your thoughts on Google Base?


2005-10-26 08:51:48
centralized vs. decentralized
From what I gather, Google Base sounds like a centralizing force. It comes at a time when decentralizing technologies are in vogue, e.g.: blogging, RSS, VoIP, P2P and tagging. Vint Cerf suggests that the revolution is on the edges of the 'net (, and I agree.