Today, Yahoo posted on their developer network blog an announcement of a new content analysis API. Its aim is to rank content by overall relevance, point to particular Wikipedia pages and annotate the results with extensive meta-data. The service is available as a Yahoo Query Language (YQL) table and more information can be found here. You can try out a sample query request and see the XML code that is returned in response, as well as documentation for the particular fields that are part of the interface.
(Sample results from a typical query shown above)
The API is something that Yahoo internally developed to analyze content for its own searching purposes, and this announcement opens up that process to any developer who knows their way around YQL. It replaces the older Term Extraction service and extends it in new and interesting ways, such as the ability to extract key terms from the content stream and rank the results according to various metrics. Mapping content to the Yahoo taxonomy is also a plus. Right now US English and Chinese languages are initially supported.
The Content Analysis service is limited to 5,000 queries per IP address per day and to noncommercial use. The Term Extraction service will be phased out by the end of March 2012.