The AP Metadata API that has internally powered the Associated Press’ award-winning news operation is now available to third-party developers and publishers. The service is aimed at publishers looking to “add rich metadata to their content, enabling them to connect their customers with more relevant news and information.” The AP delivers the Metadata service in a cloud-based offering via a set of APIs associated with the organizations news content.
The service includes two components: AP News Taxonomy and AP Tagging Service. The News Taxonomy constitutes a “comprehensive classification system, including standardized subjects, people, organizations, geographic locations and more, all designed with news content in mind.” The AP Tagging Service “uses human-created semantic rules to understand your content and identify the most pertinent concepts and topics.” Amy Sweigert, AP Vice President of Information Management, asserts: “AP Metadata Services offers publishers tools to get customers the right content at the right time…[and] allows publishers to benefit from the ongoing investment AP is making in the service.”
Data can be returned in XML, JSON, HTML or TTL. The AP continuously updates News Taxonomy data and includes more than 4100 subjects, 2000 geographic locations, 1100 organizations, 87,000 people and 38,000 companies. The Tagging Service derives data from organizations that submit English-language content using a RESTful API and receive back metadata based on the AP News Taxonomy, including named entities, generic subjects and the subject hierarchy.
The Associated Press has utilized the services internally since 2007. Since its internal inception, the AP has used the service to focus attention on appropriate areas and transition resources where needed (e.g. NCAA microsite). Now, the AP can generate a new revenue stream by offering its services to other publishers.