On Using Distribution-Based Compositionality Assessment to Evaluate Compositional Generalisation in Machine Translation
This work addresses the problem of evaluating compositional generalization in real-world NLP tasks for researchers and practitioners, though it is incremental as it applies an existing framework to a new domain.
The authors tackled the lack of real-world benchmarks for compositional generalization in machine translation by using a distribution-based framework to automatically split the Europarl corpus into training and test sets with divergent dependency relations, resulting in a fully-automated procedure for creating natural language benchmarks.
Compositional generalisation (CG), in NLP and in machine learning more generally, has been assessed mostly using artificial datasets. It is important to develop benchmarks to assess CG also in real-world natural language tasks in order to understand the abilities and limitations of systems deployed in the wild. To this end, our GenBench Collaborative Benchmarking Task submission utilises the distribution-based compositionality assessment (DBCA) framework to split the Europarl translation corpus into a training and a test set in such a way that the test set requires compositional generalisation capacity. Specifically, the training and test sets have divergent distributions of dependency relations, testing NMT systems' capability of translating dependencies that they have not been trained on. This is a fully-automated procedure to create natural language compositionality benchmarks, making it simple and inexpensive to apply it further to other datasets and languages. The code and data for the experiments is available at https://github.com/aalto-speech/dbca.