Design and Evaluation of a Retrieval Augmented Artificial Intelligence Approach for Biomedical Literature Extraction
Abstract
The volume of biomedical literature indexed in PubMed exceeds the practical capacity of manual review, while standalone generative models produce fluent responses without verifiable provenance. Although biomedical literature evolves and increases continuously, clinicians and researchers primarily rely on static literature reviews and keyword based search systems. These traditional approaches lack semantic understanding, structured retrieval, automated knowledge freshness monitoring, and integrated validation workflows. This thesis presents the design, implementation, and evaluation of a domain adapted Retrieval Augmented Generation platform tailored for extraction of records related to thrombosis risk factors from PubMed literature. The proposed system integrates semantic embedding based retrieval, vector database storage, configurable large language model generation, and a continuous DOI level ingestion scheduler. A human in the loop curation workflow is incorporated through annotation tooling, enabling validation and iterative improvement of extracted knowledge. The platform follows modular architectural principles emphasizing reproducibility, observability, and configuration driven experimentation. A multi phase evaluation framework assesses retrieval quality, answer grounding, extraction accuracy, and operational reproducibility using qrels based ranking metrics, RAGAS metrics, extraction metrics, and recorded run artifacts. The June 2026 generation benchmark evaluates…
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.