Defense Acquisition University (DAU) Chooses Finch Computing for Custom Text Classification Project
Reston, VA – Finch Computing, a leading provider of real-time natural language processing technology to federal government and private sector customers, today announced that it has been selected by the Defense Acquisition University (DAU) for an important text classification project enabling better search on the agency’s massive trove of online instructional materials for the acquisition workforce.
“We are excited about this partnership and about its transformative potential for DAU,” Finch Computing Chief Technology Officer Scott Lightner said. “Once this work is complete, the defense acquisition workforce will be better able to find more relevant content – and more of it – on the DAU intranet, enabling them to ultimately become better stewards of taxpayer resources.”
DAU is the Department of Defense’s acquisition training arm. Its mission is to provide a global learning environment to develop qualified acquisition, requirements and contingency professionals who will then deliver and sustain effective and affordable warfighting capabilities for the Department of Defense.
Finch Computing was selected for this work because of its expertise in understanding unstructured text and helping customers find value in their textual assets. Its proprietary Finch for Text® software leverage natural language processing, machine learning and artificial intelligence to turn text into actionable insights for decisionmakers.
“We know that our customers need their content classified into domain-specific topics or ontologies,” Lightner continued. “We use state-of-the-art techniques to build domain-specific classifiers, and we combine data-centric AI and natural language processing to achieve the highest quality results. We also leverage our pre-trained language models and transfer learning to rapidly build classifiers with minimal manual tagging.”
For this project, the Finch team built a custom classifier for DAU with more than 100 topics. From there, Finch for Text® will be used to instantly tag and classify the existing DAU catalog of instructional materials, as well as new content as it comes online. This will make the content immediately indexable and searchable, enabling the defense acquisition workforce to easily access content designed for them, and at the same time drive the utility of every piece of content placed on the site and furthering the DAU mission overall.