The limitations in current search technologies inspired 8 years of R&D into new search tools and a latent semantic database. The result is a genuine innovation (not an upgrade of Boolean search). This neural system is self learning. It presents one of the first practical applications of artificial intelligence.

The gulf between this technology and traditional keyword search is as wide as between today’s anti-virus software compared with 1987’s...

The new technology is an intelligent software (Latent Semantic Analysis LSA) that understands the content of your search. Enter your natural language text query. The advanced linguistics engine will then convert your words into a concept, represented by a mathematical vector. The search engine matches and ranks your vector to similar vectors in the Semantic patent database. The new semantic database already has mapping of all patents present.

To ease up the search, the software allows for proximity searching, simultaneous keyword or Boolean search and filter by relevancy score. Full claims or invention disclosures can be inserted as the query (more than 4000 words). It understands that words have more than one meaning. This delivers dramatic benefits for your search results, which do not miss any patents, regardless of whether or not they contain keywords contained in the search query, and they are ranked in order of relevance. False positive hits are pushed to the bottom, saving considerable time over Boolean.

The tool discovers related and analogous art, patents purposely written to obscure discovery, and it obtains reliable search results when searching with misspelled words or improper grammar. It allows to search databases of unlimited size, saving valuable time and costs. Thanks to indexing of multiple languages, an English query will return relevant results from foreign documents.