SmartSearch is a deep learning tool developed by NETL/MATRIC to assist with data discovery for research. Designed to transform how scientists and engineers connect to relevant data resources, SmartSearch was designed to learn from a corpus of seed resources, such as documents, zip folders, html resources, etc. to parse digital data stores and drive data discovery. SmartSearch is a deep learning tool that automates data discovery by analyzing training content to find new, related content on enterprise data stores, the world-wide web and local file shares. The tool was recently re-engineered using a multi-cloud approach to offer an infinitely-scalable solution for data discovery. Using a combination of artificial intelligence, natural language processing, and parallel processing, the tool can ingest, analyze, discover, catalog, and provide relevance analyses on returned content. 

SmartSearch has been used to rapidly assemble open-source data to characterize global oil and gas infrastructure to help mitigate fugitive methane emissions, to assemble the world’s largest open-source dataset of carbon storage related resources, as well as for materials data and information discovery.

Presenter Bios

Dr. Kelly Rose is technical director for the Science-based AI/ML Institute (SAMI), based at the U.S. Department of Energy’s National Energy Technology Laboratory (NETL), and a geo-data science researcher with over twenty years experience developing data-driven methods and models to address energy and environmental challenges at NETL.

Vic Baker is a senior systems engineer at the Mid-Atlantic Technology, Research and Innovation Center (MATRIC); lead architect for SmartSearch, which has been used by DOE HQ for eval of GCP platform for ATO; and technical lead for architecture and implementation of high availability EDX deployment (first of its kind for NETL).