Discovery’s strength lies in its ability to handle both key word queries and general concept-based queries. Using a combination of advanced natural language processing, concept based classification and neural networks technology, Discovery extracts concepts in a query, relates them to concepts in stored documents, and then returns an accurate list of automatically categorized results, making it easy for users to locate the exact piece of information searched for.

Automatic Categorization

The Discovery search engine creates categories of search results on the fly. When a list of related results is being returned to a user, Discovery will identify how many conceptually different categories are present and divide the results up appropriately. The robust technology behind the search engine is able to discern and sort similar information that is different, and groups the most similar results together. Discovery then labels each category by locating the dominant concept in each group and calls it by that name.

For example, a search on health care in America might return results on the history of health care, accidental deaths, cost of health care, children’s health issues and government policy on health care. Discovery will break up the complete list of results into these categories, place the relevant document within each, and name the group appropriately. This powerful functionality is unique to Discovery and allows users to perform more inclusive and comprehensive searches that are also extremely accurate.