Sign in Subscribe
Artificial intelligence

Artificial Intelligence System to Organize Scientific Data Receives $400,000 NSF CAREER Award

Strategic Insights

— 1 min read

Virginia Tech assistant professor of computer science, Xuan Wang, has received a $400,000 National Science Foundation Faculty Early Career Development Program (CAREER) award to build an accurate and trustworthy artificial intelligence system that can quickly and accurately extract usable information from troves of published scientific research and other data. Wang's five-year project, titled "Automated, Quality, and Trustworthy Scientific Information Extraction from Massive Text Data," aims to tackle the growing problem of extracting patterns and meaning from the vast and expanding trove of scientific data available to researchers. By using artificial intelligence, Wang hopes to speed up meaningful discoveries in any scientific field and improve health outcomes.

Key Takeaways:

  • Xuan Wang, assistant professor of computer science, has received a $400,000 National Science Foundation CAREER award to develop an artificial intelligence system that can accurately extract information from scientific data.
  • The project, titled "Automated, Quality, and Trustworthy Scientific Information Extraction from Massive Text Data," aims to tackle the problem of extracting patterns and meaning from scientific data, speeding up discoveries in various fields.
  • Wang's team will collaborate with PubMed and Children's National Hospital to create an automatic extraction system for electronic health records, improving health outcomes.
  • The project will target systems that can work across multiple science and engineering fields, including biology, physics, math, and other disciplines.
  • Wang will compare the effectiveness of large and small language models to create systems that can mine databases for information and structure it to deliver accurate overviews of existing knowledge.
  • Large language models are powerful but expensive and can make mistakes, while small models are simpler and cheaper to use but require tuning.

Statistics:

  • $400,000: The amount of the National Science Foundation CAREER award received by Xuan Wang.
  • 5 years: The duration of the CAREER grant.
  • 38 million: The number of citations for biomedical literature on PubMed.
  • 2 forms of AI: Large and small language models, which will be compared for their effectiveness in creating systems that can extract usable information from scientific data.
  • 10%: The rate at which large language models can hallucinate or provide information not included in the database.

Sources:

  • Virginia Tech news article
  • National Science Foundation CAREER award
  • PubMed database

Read more