Vicente, our graduate student and collaborators, have developed a machine learning framework that automatically identifies when scientific facilities like telescopes are used in astrophysics papers, addressing a long-standing challenge in tracking research facility usage.

The framework analyzes scientific text using natural language processing and Support Vector Machine classification to accurately detect when missions like Kepler and TESS are used in research, achieving 92.9% accuracy. This innovative approach saves valuable research time by eliminating the need for manual classification while being adaptable for use across various scientific facilities worldwide. To see more and read the full paper: https://arxiv.org/abs/2411.00987.

Collaborators:

• - Wolfgang Kerzendorf (Department of Computational Mathematics, Science, and Engineering; Department of Physics and Astronomy, Michigan State University)
• - Brian Cherinka (Space Telescope Science Institute)
• - Joshua V. Shields (Department of Physics and Astronomy, Michigan State University)
• - Annie Didier (Outrider, Golden)
• - Katharina von der Wense (Department of Computer Science, University of Colorado Boulder; Institute of Computer Science, Johannes Gutenberg University Mainz)

Gracie Tvrdik presented her research on the evolution of collaboration dynamics in Astrophysics at the annual ACRES REU (Research Experiences for Undergraduates) Symposium. Her outstanding presentation earned her the 1st Place Outstanding Talk Award.

Key points of Gracie's research:

1. Utilized the NASA/ADS (Astrophysics Data System) dataset to analyze research dynamics in astrophysics over time.
2. Developed innovative methods to extract country information from affiliations, enabling accurate tracking of changes in research collaborations.
3. Collaborated with graduate student Vicente Amado Olivo and undergraduate Ben Mellon from our research group.

The project outcomes showcase Gracie's exceptional research skills and hard work. Congratulations to Gracie for her outstanding achievement and to Vicente for his guidance throughout the project.

Our graduate student Vicente Amado Olivo attended the 3rd International Conference on the Science of Science and Innovation (ICSSI) on July 1 - 3, at the National Academy of Sciences in Washington, DC. This interdisciplinary conference brought together leading experts in the field of science of science and innovation, providing a multi-channel platform to bring together both producers (scientists from industry and academia) and consumers (policymakers, publishers, funders, administrators, etc.) of the field. Vicente presented his work in identifying unique researchers from the literature utilizing machine learning to solve the author ambiguity challenge.

Vicente has successfully defended his comprehensive examination titled 'Leveraging Machine Learning to Revolutionize Astrophysics Research Processes: From Peer Review to Research Management'. His work explores how computational techniques like machine learning can enhance various aspects of the research process in astrophysics.

This upcoming August, Benjamin Mellon and fellow undergraduate at MSU, Aardarsh Swaminathan, will intern in Oslo, Norway under Dr. John Aiken.

The interns will apply modern AI tools to identify acoustic events occuring in a combined hydrophone and geophone data set acquired in the Oman Drilling Project's Multiborehole observatory.

The students will use a neural network to classify physical occurences for later classification based on limited validation data.

The interns are enthusiastic to use cutting edge computing resources to collaboratively devolope AI software solutions for a geophyscal problem.

Bea, an undergraduate in the group, is excited to join Blue Cross Blue Shield of Michigan's Analytics Center of Excellence as a Summer 2024 Intern! Bea has previously worked with big data and different projects in the Deepthought Initiative Collaboration (DTI). That includes conducting text classification of abstracts with machine learning, website development with GitHub and web template engines, and creating map visualizations of researcher migration paths. She is eager to explore a new industry and help create impact with her skills. Even more so, Bea is excited to learn and grow as a data scientist through this upcoming experience.

The Deepthought Initiative (DTI) collaboration applies computational techniques and tools to analyze the vast literature in scientific research, with a specific focus on leveraging machine learning to understand astrophysics literature. Vicente Amado Olivo's research revolves around tracking astrophysics publications and identifying astronomers worldwide through publications on arXiv. Being part of a research group comprising astrophysicists, Vicente's work in computational meta-research ensures that the tools developed by DTI are directly applicable to the needs of astrophysicists. To further their research in astrophysics meta-research, Vicente is visiting ESAC, a world-leading space agency, in order to gain firsthand experience with their current research systems, understand their future needs, actively participate in a proposal cycle, and collaborate with the librarians at ESA.

Wolfgang Kerzendorf and collaborators organized a workshop on peer review at the European Southern Observatory (ESO) in Garching, Germany. The workshop aimed to address the persistent challenges faced by peer review in the scientific community, including the difficulty of identifying experts in specialized fields and the increasing number of researchers and publications. Experts from organizations such as ESO, ESA, ALMA, STScI, NASA, and NOIRLab came together to discuss the state of peer review and explored innovative approaches for a digital and interconnected science community. The workshop consisted of four main sessions and was held at the ESO Headquarters from February 6 to 10, 2023. Vicente Amado Olivo presented a contributed talk and you can find the slides see link in references.