Speaker
Description
Literature search engines have become an integral and indispensable part of academic research. Search engines like google scholar rely on powerful AI/ML tools to return results that match the user's intended meaning as closely as possible, but they are better suited for general-purpose queries. As a result, they tend to overwhelm the user searching for domain-specific information with irrelevant hits. For this reason, the Nuclear Science References (NSR) database [1], hosted and maintained by the National Nuclear Data Center at Brookhaven National Laboratory (BNL), has become a standard search engine in the field. The goal of the NucScholar project is to build on the capabilities of NSR by automating onerous archival tasks needed to construct a database of nuclear science journal articles, and by augmenting the search/retrieval interface through the use of natural language processing (NLP) tools.
In this talk, I will present the current status of the project. I will focus in particular on the automated archiving functionality and on the NLP capability to perform both semantic and question-answering searches, and their implementation using deep learning models.
This work was performed under the auspices of the U.S. Department of Energy by Lawrence Berkeley National Laboratory under Contract DE-AC02-05CH11231. This work was supported in part by the by the U.S. Department of Energy, National Nuclear Security Administration, Office of Defense Nuclear Nonproliferation Research and Development (DNN R&D) through the Nuclear Science and Security Consortium under Award Numbers DE-NA0003180 and DE-NA0003996.
[1] B. Pritychenko et al., NIM A 640, 213 (2011).
