Contextual Insight
Contextual Insight is next-generation search intelligence. It dynamically identifies relationships so that users can quickly find facts and answers to questions like, “When was the contract with Company A signed?”
Contextual Insight enables users to search for facts rather than documents. It detects concepts and entities, making associations to document structures (e.g., sentences and paragraphs) and drilling down into associated concepts and entities. As a result, users get new search experience with a deeper level of precision that allows them to quickly investigate related information through Contextual Navigation.
Contextual Insight is based on the following capabilities:
- Extracting semantic information and annotating it to text elements.
- The ability to search for associations between keywords and entities, as in, “Give me a paragraph of text that matches my keywords but also includes this date.”
- Greater exploration of entities associated with query results – for example, “I don’t know the person I am looking for, but I can navigate by name from the query result.”
- Expanded categorization of documents in the index based on similarities (with training sets), matching rules or programmatic rules.
- In Scope Search, the FAST ESP index maps the document structure into a hierarchical scope structure. The structure can be explicit from an XML document or automatically derived by FAST ESP. By searching, for example, at the sentence level, a user asks the search engine to return the sentences that match all conditions of the query. Documents are no longer the atomic retrieval unit, and unstructured content becomes structured.
Personalization that takes the understanding of the user context to the next level, combining the implicit information that the user provides with the explicit query data.
Source: FAST ESP Brochure 2007