This Small Business Innovation Research Phase I project is to design a generic framework for formalizing and automatically extracting domain-specific information from unstructured text for the purpose of automatic processing. The rate at which new information becomes available has increased to a point that it is impossible for people to identify the nature the information content as it is made accessible and even less feasible to absorb the actual information content. Current search technologies solve the problem of finding documents, but they do not address the fundamental problem of cognizance of the information contained in newly available documents. Being aware of the information content has now become the real challenge. Therefore it has become critical to automatically process information as it becomes available. The framework will handle domain specific areas where language variations are wide. The commercial applications of Teragram's proffered technology include alerts based on content, feature extraction for clustering and visualization of large information contents, and structuring data from documents into databases for numerous domains, including financial analysis, financial earnings releases, sport results, weather forecasts, terrorist events, election results, and product price comparisons
