Linguistically inspired morphological inflection with a sequence to sequence model

TL;DR

This paper explores a linguistically inspired sequence-to-sequence neural model that learns morphological inflections by modeling stems and affixes as character blocks, showing improved performance especially on unseen words across multiple languages.

Contribution

It introduces a character-morpheme-based seq2seq model that unifies linguistic theory with neural methods for morphological inflection, particularly excelling in low-resource and unseen word scenarios.

Findings

Small improvement in inflecting known lemmas (+0.68%)

Steady better performance on unknown words (+3.7%)

Enhanced accuracy in low-resource settings (+1.09%)

Abstract

Inflection is an essential part of every human language's morphology, yet little effort has been made to unify linguistic theory and computational methods in recent years. Methods of string manipulation are used to infer inflectional changes; our research question is whether a neural network would be capable of learning inflectional morphemes for inflection production in a similar way to a human in early stages of language acquisition. We are using an inflectional corpus (Metheniti and Neumann, 2020) and a single layer seq2seq model to test this hypothesis, in which the inflectional affixes are learned and predicted as a block and the word stem is modelled as a character sequence to account for infixation. Our character-morpheme-based model creates inflection by predicting the stem character-to-character and the inflectional affixes as character blocks. We conducted three experiments on creating an inflected form of a word given the lemma and a set of input and target features, comparing our architecture to a mainstream character-based model with the same hyperparameters, training and test sets. Overall for 17 languages, we noticed small improvements on inflecting known lemmas (+0.68%) but steadily better performance of our model in predicting inflected forms of unknown words (+3.7%) and small improvements on predicting in a low-resource scenario (+1.09%)