Enhancing Active Learning for Annotation Sampling over Large-Scale Corpus via Vector-Based Indexing

Poster Number: P75
Presentation Time: 05:00 PM - 06:30 PM

Abstract Keywords: Natural Language Processing, Data Mining, Information Retrieval
Working Group: Natural Language Processing Working Group
Primary Track: Applications
Programmatic Theme: Clinical Informatics

Active learning offers a path to efficient NLP model development by strategically selecting informative data for annotation. However, traditional methods struggle with large clinical datasets due to the computational cost of processing the entire corpus in each sampling round. This work proposes a novel vector-based active learning approach for clinical note sampling for annotation. Our preliminary experiments demonstrated that this approach offers a promising solution.

Speaker(s):
Jianlin Shi, MS, MD
The Division of Epidemiology, School of Medicine, University of Utah; VA Salt Lake City Healthcare System