On the Transferability of Neural Models of Morphological Analogies

Safa Alsaidi (1) , Amandine Decker (2) , Puthineath Lay (3) , Esteban Marquer (1) , Pierre-Alexandre Murena (4) , Miguel Couceiro (1)

1 ORPAILLEUR - Knowledge representation, reasonning
2 SEMAGRAMME - Semantic Analysis of Natural Language
3 LORIA - NLPKD - Department of Natural Language Processing & Knowledge Discovery
4 HIIT - Helsinki Institute for Information Technology

Analogical proportions are statements expressed in the form "A is to B as C is to D" and are used for several reasoning and classification tasks in artificial intelligence and natural language processing (NLP). In this paper, we focus on morphological tasks and we propose a deep learning approach to detect morphological analogies. We present an empirical study to see how our framework transfers across languages, and that highlights interesting similarities and differences between these languages. In view of these results, we also discuss the possibility of building a multilingual morphological model.

Mots clés: morphological analogy; deep learning; transferability; analogy classification

Domaines: Intelligence artificielle [cs.AI]; Apprentissage [cs.LG]

Safa Alsaidi, Amandine Decker, Puthineath Lay, Esteban Marquer, Pierre-Alexandre Murena, et al.. On the Transferability of Neural Models of Morphological Analogies. AIMLAI 2021 - workshop on Advances in Interpretable Machine Learning and Artificial Intelligence, Sep 2021, Bilbao/Virtual, Spain. pp.76-89. ⟨hal-03313591⟩