Search with the Monkey

Last week I went to the Semantic Conference in San Jose. Before I went there I was pretty new to the whole Semantic Web, so the conference gave me a chance to discover some interesting things.

One of the people I met at the conference is Peter Mika from Yahoo! Research. He is working on SearchMonkey, which is a service offered by Yahoo! that lets developers and site owners represent search results in a more appealing way. Allowing site owners to create more enhanced results can give a boost to the relevant traffic to the website.

A Searchmonkey enabled search result

A SearchMonkey application consists of structured data and a PHP script. The data may be either cached or loaded at runtime. A number of ontologies are available to be used in these applications. One of those is the GoodRelations ontology by Martin Hepp (discussed by Stijn in a previous blog post). Although the entire Semantic Web movement has goals that go far beyond this kind of functionality, it is something quite helpful, and I am curious to see how it evolves.

Google uses something they call Rich Snippets. It is similar in the sense that site owners need to provide structured data but the visualization is controlled by Google.

Mircosoft’s Bing search engine doesn’t seem to be using structured data in this way. Structured data provided by site owners is not yet used by Bing. It uses intelligent data extraction to populate the document preview caption and the website’s result after a search.

Yahoo! leaves the customization up to the site owners themselves. With Google you can choose the data and Bing has still some work to do in this area.

Unfortunately it is still hard to find some nice examples that make use of SearchMonkey. Have you seen any? Created any apps?

What whould you like to use? A system that lets you create the apps or does it for you?