Technology Overview
All of Lymba's products use state-of-the-art computational linguistics technology to enable software to intelligently understand, interpret, and analyze a body of text. These processes can be applied to virtually any kind of data, and even input from the user, to allow for smooth and natural communication between the user and the computer. You might say we're teaching our software to understand English.
If that sounds like a challenging proposition, that's because it is. But our tools in natural language processing are the product of more than ten years of research and development. The soundness of the technology has been acknowledged by the academic and the government community, and more recently has been adopted by commercial customers. Each tool performs a specific and critical function for thorough automatic natural language understanding.
Did you know?
Semantic means "of or pertaining to meaning, especially in language." So semantically equivalent simply means that the meaning of the sentence is the same!
One common challenge in natural language processing is determining whether one sentence is semantically equivalent to another sentence when the words used to describe the content are different. We have addressed this challenge by building applications that utilize our inference engine, COGEX, which takes as input logic forms, lexical chains to create axioms, eXtended WordNet Knowledge Base to provide a source of knowledge for axioms, Jaguar as a source of domain specific knowledge, coreference chains to link multiple expressions that refer to the same entity, and word sense disambiguation in order to identify the meaning of the words based on the context of the sentences. For each of these tasks syntactic parsing and part of speech tagging are crucial inputs to these tools.