| Sol No.: |
Navy SBIR FY2012.2 |
| Topic No.: |
N122-136 |
| Topic Title: |
Cloud-Based Tell-Me-About Q&A System |
| Proposal No.: |
N122-136-0816 |
| Firm: |
Vcrsoft LLC
2310 Bamboo Drive STE J303
Arlington, Texas 76006-5952 |
| Contact: |
VC Ramesh |
| Phone: |
(817) 213-6184 |
| Abstract: |
The essence of the problem is to allow military intelligence and operations staff to pose "tell me about" questions concerning an entity, and obtain a finished information product. There are 4 main technical challenges / sub-problems: a) Question content modeling: The question needs to be semantically mapped to an ontology. b)Semantic enrichment of unstructured text: Unstructured data (text, images etc.) needs to be marked-up with semantic tags that relate the content to the ontology. c)Sufficiency tracking of discovered content: Relevant content needs to be discovered and its sufficiency evaluated as an adequate response to the question. d)Presentation of finished product: The answer needs to be presented as a report in a familiar format such as MS Office or Adobe PDF. We present a cloud-based approach to address all these issues. |
| Benefits: |
The DoD intelligence community is the first anticipated beneficiary. Other DoD agencies and other government agencies including NASA analysts are the next anticipated beneficiaries. Law enforcement agencies and private-sector entities such as healthcare providers are also commercialization targets. |