The Effect of Data Partitioning Strategy on Model Generalizability: A Case Study of Morphological Segmentation
This work addresses the challenge of realistic model evaluation for researchers in computational linguistics, though it is incremental as it builds on existing partitioning strategies.
The study tackled the problem of evaluating data partitioning strategies for model generalization in morphological segmentation, finding that models trained on random splits achieved higher scores and more consistent rankings when tested on new data.
Recent work to enhance data partitioning strategies for more realistic model evaluation face challenges in providing a clear optimal choice. This study addresses these challenges, focusing on morphological segmentation and synthesizing limitations related to language diversity, adoption of multiple datasets and splits, and detailed model comparisons. Our study leverages data from 19 languages, including ten indigenous or endangered languages across 10 language families with diverse morphological systems (polysynthetic, fusional, and agglutinative) and different degrees of data availability. We conduct large-scale experimentation with varying sized combinations of training and evaluation sets as well as new test data. Our results show that, when faced with new test data: (1) models trained from random splits are able to achieve higher numerical scores; (2) model rankings derived from random splits tend to generalize more consistently.