NEW YORK--(BUSINESS WIRE)--Today, NTENT announced a breakthrough in its quest to transform information discovery with the advancement of its Enhanced Semantic Ranking and Knowledge Base technologies. The adjustments and improvements to these core capabilities allow NTENT to more efficiently decipher user intent, making it faster and easier for people to find information from across the web. By applying advanced semantic ranking algorithms across a vast lexicon and custom ontology using machine learning and natural language understanding, NTENT’s Enhanced Semantic Ranking disambiguates complex queries to accurately detect user intention and deliver relevant results.
Using this proprietary approach, NTENT’s disambiguation technology incorporates three stages of semantic understanding referred to as Concept Ranking, Intention Ranking and Intent Answer Ranking. Each stage works symbiotically with the other to identify concepts, derive user intentions and provide answers.
Concept Ranking infers the meaning of words according to the context of a query. It applies various deduction methods to distinguish between connotations for homonyms and integrates other factors, such as user location, to deduce interpretations and produce a more accurate semantic connection. Intention Ranking correlates the inferred contextual meaning of words to the most probable user motives behind them. Once defined and understood, Intent Answer Ranking interfaces with third-party data sources to provide precise answers within a specific domain such as sports or weather.
“For example, in the query ‘Italian Friday Manhattan’ our system interprets ‘Italian’ as a type of food rather than a person who comes from Italy,” said Chief Executive Officer for NTENT Dan Stickel. “Intention Ranking characterizes the semantic relationship as a user who wants to eat Italian food on Friday in Manhattan, and then uses Intent Answer Ranking to deliver a series of pertinent answers and other useful information prioritized according to relevance.”
Language Independence Provides Multi-Lingual Support
Thanks to a vast ontology that includes numerous concepts, connections and idioms as part of NTENT’s Knowledge Base, a significant characteristic of NTENT’s Enhanced Semantic Ranking is its ability to differentiate concepts with the same level of accuracy, across any language. NTENT’s Knowledge Base is able to interpret words with the same meaning but spelled differently, words spelled the same with different meanings, and words that sound and mean something completely different than how they read.
“Once the concepts behind the words become clear,” said NTENT CTO, Dr. Ricardo Baeza-Yates, “our system processes the parallel intention in a language-independent manner, comparing and contrasting user context with potential outcomes to produce a standard interpretation of the query that can be used by other subsystems. With all complicated Natural Language Understanding work done from a central place, it can be leveraged by all services in our platform.”
NTENT’s technology recognizes the English phrase “museums near me” as equivalent to the French phrase “Musées près de moi.” Context is determined using factors from circumstantial and tangible data including things like categorical references to geographical landmarks or movies. This universal, conceptual understanding means a more simplified, streamlined approach to fast, relevant global search, transferable to all cultures and languages.
NTENT made headlines last June when they announced the expansion of their Natural Language technologies to support the Russian language.
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.
