Army posts NLP funding opp
On November 16, the U.S. Army posted a request for proposals for natural language processing under the DoD SBIR program. Proposals are due no later than 12:00 p.m. Eastern on January 4, 2022.
The objective of this topic is to develop Artificial Intelligence / Machine Learning models to augment Natural Language Processing (NLP) capabilities in 2 main challenge areas: relationship detection and aggregation – automatically detecting relationships that exist between the entities that were extracted from the data. Some extracting attributes of entities include: hair color, nationality, model of tank, armor of tank. Pattern recognition, analysis, and exploitation – automatically recognizing patterns such as indications and warnings and courses of actions and analyzing them – is an integral part of this topic.
The purpose of this topic is to demonstrate how novel approaches and techniques can address these challenge areas and to develop prototypes that can be transitioned into PM IS&A’s products. PM IS&A products will empower the intel analyst by providing them with access to critical data/information and advanced analytics.
NLP is a critical capability for PM IS&A products that is necessary to make sense of large amounts of unstructured data from multiple unclassified and classified sources. Unfortunately, today’s technology struggles at adequately addressing this problem. This should be done by reducing the cognitive load on the intel analyst, improving situational awareness and situational understanding for the war fighter and intel analyst, addressing gaps in the existing NLP technologies, and improving data fusion and pattern recognition and analysis tools that are used for intelligence applications. Currently, information is manually extracted by people.
Today, intel analysts spend half of their time looking for the golden nugget in massive amounts of data sets by using a mixture of techniques and tools. Today’s NLP technologies excel at very specific tasks such as identifying attributes and extracting attributes from entities (not correlating back to either other). We propose developing AI/ML models to help augment the NLP technologies. Based on Industry Engagements over the last 12 months, small portion of Industry Partners have started to introduce AI/ML modeling to improve NLP performance. This provides us confidence in the approach to leverage AI/ML modeling to augment NLP technology.
IC News brings you business opportunities like this one each week. If you find value in our work, please consider supporting IC News with a subscription.