NEW YORK--(BUSINESS WIRE)--NTENT announced today a significant step forward in its effort to understand human intent. Through a powerful, proprietary Knowledge Base developed using cognitive, semantic and behavioral perspectives, NTENT is able to classify, interpret and compare data using the same thought patterns as the human brain, independent of, and applicable to any language.
Many words have multiple meanings that change according to culture and language, however the concepts between the words remain ubiquitous. From a cognitive perspective, NTENT’s Knowledge Base provides a strong estimation of general, categorical and factual data that’s classified by levels of abstraction and linked to relevant labels. This core model operates independent of language, making it easier to sustain and employ multilingual functions.
From a semantic perspective, the Knowledge Base acts as a universal language for NTENT’s Natural Language Understanding technology that conveys the meanings of words or phrases as ontological concepts. Behaviorally speaking, the Knowledge Base supports practical reasoning tasks and implements a unique disambiguation module to aide in determining lexical equivocation.
Three Layers to Deciphering Intent in Any Language
The Knowledge Base is built on three fundamental layers. First, a language-independent ontology enables top-tier segmentation across multiple categories following an entity-event-attribute-relations pathway, across all domains. In the second layer, a vast lexicon consisting of language-specific phrases is linked to language-independent concepts, and used to decode words with multiple meanings by detecting the correlation between them. The third layer is a sophisticated Onomasticon, housing lists of specialized terms relating to various subjects collected through third-party data sources on the web.
“By nature, the human brain stores information, reasons on it and takes action to put it to use,” said Dan Stickel, Chief Executive Officer for NTENT. “Our Knowledge Base works the same way in that it collects information, assigns a probability of intent connecting the data, and applies the most relevant solution. In a query for the word ‘mint,’ the system would recognize it as a personal finance software, an herb with a distinctive flavor, a Linux distribution, a place where coins are produced, a type of hard candy and a music venue in Los Angeles. Our technology weighs the possible relationships between all words in the query as well as the context of it, to determine the most accurate interpretation(s) of the query and the predicted intent(s) behind it,” said Stickel.
Last December, NTENT announced a breakthrough component of their Knowledge Base that helped disambiguate queries called Enhanced Semantic Ranking.
About NTENT: NTENT™ sits at the crossroads of semantic search and natural language processing technologies. Our patented, proprietary technology powers our comprehensive platform that transforms structured and unstructured data into relevant and actionable insights. This level of intelligence enables us to predict and deliver relevant information based on user intention. Learn more about NTENT at http://www.ntent.com.