Corpus Frequencies in Morphological Inflection: Do They Matter?

TL;DR

This paper investigates the impact of incorporating corpus frequency information into morphological inflection tasks, proposing methods that improve model performance by reflecting real-world word usage distributions.

Contribution

It introduces frequency-aware training, combines lemma-disjoint splits with frequency weighting, and compares token versus type accuracy for more realistic evaluation.

Findings

Frequency-aware training outperforms uniform sampling in 26 of 43 languages.

Token accuracy better reflects real-world performance on frequent words.

Incorporating frequency information improves generalization and evaluation metrics.

Abstract

The traditional approach to morphological inflection (the task of modifying a base word (lemma) to express grammatical categories) has been, for decades, to consider lexical entries of lemma-tag-form triples uniformly, lacking any information about their frequency distribution. However, in production deployment, one might expect the user inputs to reflect a real-world distribution of frequencies in natural texts. With future deployment in mind, we explore the incorporation of corpus frequency information into the task of morphological inflection along three key dimensions during system development: (i) for train-dev-test split, we combine a lemma-disjoint approach, which evaluates the model's generalization capabilities, with a frequency-weighted strategy to better reflect the realistic distribution of items across different frequency bands in training and test sets; (ii) for evaluation, we complement the standard type accuracy (often referred to simply as accuracy), which treats all items equally regardless of frequency, with token accuracy, which assigns greater weight to frequent words and better approximates performance on running text; (iii) for training data sampling, we introduce a method novel in the context of inflection, frequency-aware training, which explicitly incorporates word frequency into the sampling process. We show that frequency-aware training outperforms uniform sampling in 26 out of 43 languages.