Don't Forget the Long Tail! A Comprehensive Analysis of Morphological Generalization in Bilingual Lexicon Induction
This addresses a key limitation in machine translation for handling rare word forms, though it is incremental as it builds on existing models with a constraint-based enhancement.
The study investigated whether state-of-the-art bilingual lexicon induction models can generalize to rare morphological inflections, finding that performance drops significantly on such forms, but adding a simple morphological constraint during training improves results.
Human translators routinely have to translate rare inflections of words - due to the Zipfian distribution of words in a language. When translating from Spanish, a good translator would have no problem identifying the proper translation of a statistically rare inflection such as habláramos. Note the lexeme itself, hablar, is relatively common. In this work, we investigate whether state-of-the-art bilingual lexicon inducers are capable of learning this kind of generalization. We introduce 40 morphologically complete dictionaries in 10 languages and evaluate three of the state-of-the-art models on the task of translation of less frequent morphological forms. We demonstrate that the performance of state-of-the-art models drops considerably when evaluated on infrequent morphological inflections and then show that adding a simple morphological constraint at training time improves the performance, proving that the bilingual lexicon inducers can benefit from better encoding of morphology.