Natural Language Processing (NLP) - a subfield of artificial intelligence and linguistics - studies the problems of automated generation and understanding of natural human languages. Natural language generation systems convert information from computer databases into normal-sounding human language. Natural language understanding systems convert samples of human language into formal representations that are easier for computer programs to manipulate.

A typical NLP problem in life sciences research is the time consuming and tedious compilation of information on interactions between biological entities like proteins and small molecules. This information is scattered across different sources of scientific literature. Manual curation of literature is expensive due to its size and continuous growth. A practical solution is to design a fully automated NLP system that can extract interactions from literature. Stratagene's software tool PathwayArchitect^TM - powered by Strand's avadis^TM - is indeed such a solution.

Strand's NLP technology tries to maximize control over different aspects of the sentence as well as accuracy without compromising recall. Extraction of interactions is done in four main phases illustrated below. They are named entity recognition, syntax analysis, semantic analysis and semantic inferences. It works sentence by sentence and extracts only interactions which occur completely within a single sentence. Each sentence is put through the four phase pipeline illustrated in the schematic alongside.

Click here to know more about our NLP technology jointly developed with Stratagene.