Natural Language Processing (NLP) is a field of artificial intelligence and computational linguistics. It investigates the problems around the automatic generation and understanding of the human language as opposed to simple keyword searching.
Cognition’s NLP employs a unique mix of linguistics and mathematical algorithms which enables the computer to understand meanings of and associations of words rather than just being able to match the requested search keywords. It also understands relationships between words such as ‘finger’ and ‘digit’ as well as the taxonomy of words e.g.: a ‘finger’ is part of a ‘hand’; a ‘cow’ is a ‘bovine’ which is a ‘mammal’.
Unlike many other text searching programs in use today, that use pattern-matching technologies, to find certain words or phrases, Cognition’s Semantic NLP understands not only the meaning of a word but also the meaning of the word in context to the search phrase supplied. What this means from an end-product point of view is that systems implementing this technology become smarter and more accurate, in turn meaning search queries return more accurate and relevant results, resulting in quicker response times and less load on systems.
Cognition’s NLP understands the origins of words, a word’s context within a particular phrase especially so with ambiguous words, synonyms of words as well as the ontological relationships of words (or the hierarchy/family of words within the English language). Their technology also understands the various ways words can be spelt, or misspelled. Their unique technology also finds appropriate content based on synonyms of words from the original search query e.g.: if a user searched for ‘fatal fumes in the workplace’ it would also find documents with terms like ‘gas’, ‘steam’ and ‘vapour’ since they relate to ‘fatal fumes’ in the search query.
Cognition have been building a semantic map of the English language over the past 23 years thus ensuring that Semantic NLP is complete and unique in enabling other technologies that require word searches and contextual-based searches to become more productive. With this in mind they are able to return search results with over 90% accuracy in comparison to search engines like Google’s which use pattern-matching.
To help one understand the power of their Semantic NLP they have created a search comparison tool to show you the differences between what their search queries return in comparison to Google’s search engine. Here’s an example of a simple phrase search such as ‘strike a match’.
They have a number of applications that their technology can be used with – click here to see the list.