Budget-Xfer: Budget-Constrained Source Language Selection for Cross-Lingual Transfer to African Languages

TL;DR

Budget-Xfer introduces a framework for optimizing source language selection and data allocation under fixed budgets in cross-lingual transfer learning, improving NLP for African low-resource languages.

Contribution

It formulates source language selection as a budget-constrained resource allocation problem and evaluates strategies across multiple NLP tasks and languages.

Findings

Multi-source transfer outperforms single-source transfer.

Differences among multi-source strategies are modest and often non-significant.

Embedding similarity as a selection proxy is task-dependent, sometimes less effective than random selection.

Abstract

Cross-lingual transfer learning enables NLP for low-resource languages by leveraging labeled data from higher-resource sources, yet existing comparisons of source language selection strategies do not control for total training data, confounding language selection effects with data quantity effects. We introduce Budget-Xfer, a framework that formulates multi-source cross-lingual transfer as a budget-constrained resource allocation problem. Given a fixed annotation budget B, our framework jointly optimizes which source languages to include and how much data to allocate from each. We evaluate four allocation strategies across named entity recognition and sentiment analysis for three African target languages (Hausa, Yoruba, Swahili) using two multilingual models, conducting 288 experiments. Our results show that (1) multi-source transfer significantly outperforms single-source transfer (Cohen's d = 0.80 to 1.98), driven by a structural budget underutilization bottleneck; (2) among multi-source strategies, differences are modest and non-significant; and (3) the value of embedding similarity as a selection proxy is task-dependent, with random selection outperforming similarity-based selection for NER but not sentiment analysis.