SwellShark: A Generative Model for Biomedical Named Entity Recognition without Labeled Data

TL;DR

SwellShark is a novel framework that leverages weak supervision and generative modeling to develop high-accuracy biomedical NER systems without requiring manually labeled data, significantly reducing annotation effort.

Contribution

The paper introduces SwellShark, a method that uses biomedical resources as function primitives for weak supervision and a generative model to produce large-scale labeled datasets without manual annotation.

Findings

Achieves competitive NER performance without hand-labeled data

Reduces annotation time from weeks to 24 hours for drug name extraction

Matches state-of-the-art supervised benchmarks in biomedical NER

Abstract

We present SwellShark, a framework for building biomedical named entity recognition (NER) systems quickly and without hand-labeled data. Our approach views biomedical resources like lexicons as function primitives for autogenerating weak supervision. We then use a generative model to unify and denoise this supervision and construct large-scale, probabilistically labeled datasets for training high-accuracy NER taggers. In three biomedical NER tasks, SwellShark achieves competitive scores with state-of-the-art supervised benchmarks using no hand-labeled training data. In a drug name extraction task using patient medical records, one domain expert using SwellShark achieved within 5.1% of a crowdsourced annotation approach -- which originally utilized 20 teams over the course of several weeks -- in 24 hours.