| Advanced Capabilities for Evidence Extraction
|
| Research, development, and large-scale testing and evaluation of the use of NLP for the extraction and interpretation of entities, events, and relations found in extensive textual collections of news sources. |
|
| Automatic Text Extraction Evaluation
|
| Automatic text extraction technology to provide an efficient methodology for the identification, extraction, and classification of information regarding terrorist activities. Using a behaviorally based pattern classification model, the Antecedent Behavior Consequence (ABC) Model, algorithms to automatically code antecedents were developed. Antecedents are defined as “any event or situation occurring prior to the behavior that is logically related to that behavior.” |
|
| Breaking the Metadata Generation Bottleneck
|
| Research and development of new NLP and Machine Learning technologies to automate current manual assignment of metatags to educational resources in math and science.
|
|
| eQuery
|
| Automatic knowledge extraction from massive data sources for the purpose of discovery of relations between people and organizations. |
|
| Feasibility Study of NLP Representation of Field Service Tickets
|
| A feasibility study designed to enable ConEdison to have improved access to and insight into their Emergency Control System’s Field Service Tickets. The goal of this exploratory text analysis is the discovery of emerging concepts, trends and relationships inherent in the data through improved processing and representation of the Remarks texts. Text-mining algorithms can then be run on this improved representation in order to detect patterns and trends. |
|
| Generic Event Recognition and Extraction
|
| Focus on the most general level of events expressed in text which will later be specialized for the domain of interest for specific applications. |
|
| Information Retrieval & Information Extraction
|
| Knowledge-base development for the travel and tourism industry, based on automatic information extractions specialized for the industry, for use in text based question-answering system, and intended for later incorporation in a speech-based system. |
|
| MetaTest: Evaluating the Quality and Utility of Metadata
|
| A two-year, spiral evaluation of the quality and utility of automatically assigned 23 metadata elements to learning resources as compared to manually assigned metadata, with teachers, professors of education, and metadata experts as subjects. |
|
| Modeling Public Health Interventions for Improved Access to the Grey Literature
|
| Research and development of NLP technology to interpret and represent the components of information from Public Health (PH) intervention reports that were deemed vital by focus groups of PH professionals as a clear summary of the problem, intervention, subjects, outcomes, etc. for information access systems.
|
|
| Question-Answering: Phase I & Phase II
|
| Research, design and development of a commercial Customer Relations Question-Answering system in collaboration with Cycorp, Inc. Includes development of tools for customers to specialize the application to their own needs and domain of interest. |
|
| Scouting Study of Natural Language Processing
|
| Analysis and data mining of consumer focus group transcriptions; English-Chinese cross-language retrieval on patents; information extraction for competitive intelligence; and feasibility study of interactive consumer dialogues. |
|
| StandardConnection: Mapping Educational Objects to Content Standards
|
| Development and testing of a semi-automatic middleware NLP tool to automatically assign to educational resources the national educational content standards which the resource supports, for use by teachers in selecting relevant lesson plans, activities, etc. |
|
