Tuesday, August 29, 2017

Charles River Analytics Inc., developer of intelligent systems solutions, has announced a contract to build a mission readiness tool for the US Navy. The Grammars for Graph-based Assessment of Mission Readiness, or GGRAMR, effort builds on Charles River’s expertise applying semantic reasoning technologies to difficult problems. Semantic reasoning is branch of artificial intelligence concerned with representing and reasoning about knowledge. The two-year contract, awarded by the Office of Naval Research, is valued close to $750,000, with an option, if extended for an additional $1.25 million.

Commanders have access to a wealth of information the Navy collects and stores in large knowledge graphs, which they can use to determine a unit’s readiness for a particular task. However, it is difficult to sift through the data and retrieve portions of a graph relevant to an assessment. To simplify this process, Charles River is developing GGRAMR with its teammates at the University of Maryland, Baltimore County (UMBC) and Stardog Union. GGRAMR is a framework commanders can use to access and analyze mission relevant data and predict mission measures of performance.

Calculating unit readiness and predicting measures of performance illustrate the challenges of service-wide graph databases. The required data may be scattered throughout the database, it may use different terminology than the readiness and performance calculations, and data elements may need to be combined to produce the values needed by readiness and performance calculations. The research goal is to develop a method that simplifies the process of transforming graph data for specific applications. Rather than flattening graph data into tables, graph grammars provide a way to work with graph data directly—maintaining the flexibility, connectivity, and visual advantages of graphs.

“As the military moves to schema-less, service-wide graph databases, a key question is how to efficiently process the information they contain,” said Dr. Terry Patten, the Principal Investigator on the GGRAMR effort. “Working with UMBC and Stardog Union, we are showing how information from huge graph databases can be used to efficiently calculate the readiness of Marine units and then to accurately predict how quickly or how well Marine units can perform mission-essential tasks. This information will be extremely valuable to commanders.”

Another project at Charles River Analytics that involves semantic information retrieval is the Content Retrieval and Extraction for Advanced Tutoring Environments, or CREATE system, which helps authors of intelligent tutoring systems quickly find the source material they need.


This material is based upon work supported by the Office of Naval Research under Contract No. N68335-17-C-0151. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the Office of Naval Research.

Close