An In-Depth Evaluation of Federated Learning on Biomedical Natural Language Processing
This addresses privacy and data access challenges for medical NLP practitioners, but it is incremental as it evaluates existing FL methods on biomedical data.
The study tackled the problem of training language models in biomedical NLP under privacy constraints by evaluating federated learning on 2 tasks across 8 corpora, finding that FL models outperformed individual client models and sometimes matched pooled data performance, with pre-trained transformers showing resilience and FL models beating zero-/one-shot large language models in speed and performance.
Language models (LMs) such as BERT and GPT have revolutionized natural language processing (NLP). However, the medical field faces challenges in training LMs due to limited data access and privacy constraints imposed by regulations like the Health Insurance Portability and Accountability Act (HIPPA) and the General Data Protection Regulation (GDPR). Federated learning (FL) offers a decentralized solution that enables collaborative learning while ensuring data privacy. In this study, we evaluated FL on 2 biomedical NLP tasks encompassing 8 corpora using 6 LMs. Our results show that: 1) FL models consistently outperformed models trained on individual clients' data and sometimes performed comparably with models trained with polled data; 2) with the fixed number of total data, FL models training with more clients produced inferior performance but pre-trained transformer-based models exhibited great resilience. 3) FL models significantly outperformed large language models using zero-/one-shot learning and offered lightning inference speed.