Press Agency ANSA chooses Expert System Semantic Search for ANSAFoto Image Archive
5 Nov 2013
Expert System today announced that ANSA, a leading global press agency, has chosen the Cogito® semantic technology to improve access in its ANSAFoto image database.
The rapid growth of information has forced media and publishers to rely on technologies that enable filtering to access only the content you’re looking for, and real time sharing, both quickly and easily. In response, ANSA has implemented the first image search engine that uses semantic technology for analysis of captions in English and Italian. Users may browse the agency’s collection of millions of photographs using advanced search capabilities that are more intuitive, and with an approach that mirrors the way users speak and think.
Compared to traditional analysis systems, Expert System’s semantic technology enables automatic processing with unmatched precision for text accompanying ANSA photos, and categorizes them according to a predefined taxonomy. Thanks to these unique features, the semantic search behind ANSAFoto ensures rapid recovery of desired images and simplifies access to the full image database.
“The semantic search engine developed for ANSAFoto is at the leading edge of our technology, which is generally used for the management of data, documents and web content,” said Bryan Bell, Executive Vice President, Expert System USA. “This is the first application of Cogito for the explicit use of search for images in a rich archive, allowing users and editors to quickly identify the right photo for every need.”
The ANSAFoto semantic search engine features:
- Meaning based search to enable immediate understanding of search intent.
- Natural language search style that is similar to how users think and speak.
- Multi-language search (search in English will also extract pictures with the description in Italian).
- Automatic suggestion of related searches to quickly lead users to desired results.
- New visual search feature allows search for similar photos based on the degree of similarity between images using their tags and captions.