Whitepaper
In the whitepaper we introduce the TextWise Trainable Semantic Vectors (TSV) technology and its use in similarity search. TSV based similarity search, provided through the SemanticHacker API, enables developers to build scalable semantic matching applications to associate text of similar topics WITHOUT requiring specific keyword matching, WITHOUT requiring marking up CRC triples in RDF formats, and WITHOUT having to manually maintain intensive linguistic rules and knowledge bases.
TextWise TSV technology enables you to map text to meaningful relevant weighted concepts, which substantially increases clarity about the subject matter in the text in a cost effective manner, and to use it to find text with similar weighted concepts. A similarity score is provided that indicates “How much something relates to something else.”
Examples of semantic matching applications to associate text of similar topics:
- Matching ads to web pages
- Matching similar blogs, news or other web content
- Matching similar patents
- Matching similar enterprise documents
- Matching similar products to one another (shopping applications)
- Mapping keyword queries to documents