Enter Note Done
Go to previous page in this tab
Session
  • Presentation | H31R: Frontier AI Models Transforming Water Science II Poster
  • Poster
  • Bookmark Icon
  • H31R-1326: From Text to Insight: Enhancing LLM Accuracy in Hydrology with Hybrid Retrieval and Knowledge Graphs
  • Schedule
    Notes
  • Board 1326‚ Hall EFG (Poster Hall)
    NOLA CC
    Set Timezone

Generic 'disconnected' Message
Author(s):
Aditya Sharma, University of Southern Mississippi (First Author, Presenting Author)
Md Ubayeid Ullah, University of Alabama
Jiaqi Gong, University of Alabama
Steven Burian, Cooperative Institute for Research to Operations in Hydrology (CIROH), Alabama Water Institute, The University of Alabama, Tuscaloosa, AL
Travis Loof, The University of Alabama


Scientists in fields like hydrology need accurate, detailed answers—not the kind of vague or incorrect responses LLMs often give. We created an AI system that reads research documents, builds smart maps of the knowledge inside, and then answers questions and we demonstrated the system using a body of documents from CIROH, showing that we can get better answers than GPT-4. It can even point out where research is missing or incomplete.



Scientific Discipline
Suggested Itineraries
Neighborhood
Type
Main Session
Discussion