CLLGApr 16, 2021

To Share or not to Share: Predicting Sets of Sources for Model Transfer Learning

arXiv:2104.08078v20.00663 citations
AI Analysis55

This addresses the challenge of selecting effective transfer sources for sequence labeling tasks across domains, which is incremental as it builds on prior similarity-based ranking methods.

The paper tackles the problem of predicting useful transfer sources in low-resource settings to avoid negative transfer, proposing methods based on model similarity and support vector machines that achieve performance increases of up to 24 F1 points.

In low-resource settings, model transfer can help to overcome a lack of labeled data for many tasks and domains. However, predicting useful transfer sources is a challenging problem, as even the most similar sources might lead to unexpected negative transfer results. Thus, ranking methods based on task and text similarity -- as suggested in prior work -- may not be sufficient to identify promising sources. To tackle this problem, we propose a new approach to automatically determine which and how many sources should be exploited. For this, we study the effects of model transfer on sequence labeling across various domains and tasks and show that our methods based on model similarity and support vector machines are able to predict promising sources, resulting in performance increases of up to 24 F1 points.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes