Trifeed

Trifeed's technology transforms text written by human, such as news articles, scientific documents, e-mails, forum messages and more, into structured data. Utilizing high level of linguistic analysis, pattern recognition and a large knowledge base, it extracts, automatically and accurately, information of many types out of the organization content.

Named Entity Extractor - Identifies names of people, companies, organizations, places, events, books, movies, dates, sums of money and many more types of entities. It also identifies the entity attributes, such as position for person, and normalizes the entity name.

Quotes Extractor - Identifies quotations and their speakers (see our Who-Said-What web site).

Relations Extractor - Extracts relation between two entities. For example: The phrase "John Smith, vice president Business Development of IBM" is extracted as a relation between an entity of the type person with the name of John Smith and an entity of the type company with the name IBM, and the relation name is "works for".

Concept extractor - Extracts main concepts within a group of documents.

Classifier - Classifies a document to the categories it belongs to based on a pre-trained taxonomy.

Summarizer - Creates a summary based on chosen sentences that brings out the essence of the document.

Similar Documents Locator - Links a leading document to a group of similar documents in the system.

Clusters Generator - Creates clusters of similar documents within a group of documents. A group of documents may be a search result. The Clusters Generator divides the documents retrieved into clusters of documents close to each other by content.