NucScholar Search Engine

Created by Walid Younes and Bethany Goldblum at the Lawrence Berkeley National Laboratory and Juan Manfredi at the Air Force Institute of Technology. Please report any problems to Walid Younes.

The NucScholar project uses Natural Language Processing (NLP) to automatically retrieve, categorize, and recommend nuclear science papers. The goal of this effort is to provide the groundwork for a shift to a fully automated workflow for nuclear science literature searches, enabling increased efficiency in the nuclear data pipeline and accelerating data throughput for a wide range of applications.

Additional information about the project can be found here, along with a listing of all project team members.

The NucScholar project is supported by the U.S. Department of Energy/Office of Science.

News No scheduled maintenance


User dashboard • Hello, you are currently signed in as guest

    To log in, follow this link
    To change password, follow this link
    To log out, follow this link

• This website provide interactive tools designed to generate training data for natural language processing models tailored to the nuclear science lexicon. The database consists of 71708 individual sentences automatically extracted from 108 arXiv articles. The following tools are available:

    Search arXiv papers by meaning

    Ask questions to find answers within the arXiv papers