Bilingual Lexicon Induction and Adaptive Techniques for Cross-Lingual Model Enhancement

Abstract

Cross-lingual models, which have contributed greatly to languages convergence in different applications, have been observed to possess viability. This research presents new ways to amplify the operations of cross-lingual models through the invention of the adaptive system of representations and the usage of the bilingual strategy applied. Our methodology examines several techniques, namely, the creation of more comprehensive and reliable language representations and tools for the automatic generation of bilingual lexicons. Our experimental results show the success of using these methods in the improvement of the model’s performance through various language profiles.