পদ্ধতির তুলনা করুন

	বহুভাষিক ভ্যারিয়েশনাল অটোএনকোডার	বহুভাষিক বাক্য এমবেডিং
ক্ষেত্র	গভীর শিখন	গভীর শিখন
পরিবার	Machine learning	Machine learning
উদ্ভবের বছর	2017-2018	2019–2022
প্রবর্তক	Multiple research groups (Lample, Conneau et al.; Zhao et al.)	Reimers, N. & Gurevych, I.; Feng, F. et al. (Google)
ধরন	Generative latent-variable model	Cross-lingual representation learning
মৌলিক উৎস	Zhao, T., Zhang, Y., & Eskenazi, M. (2018). Zero-shot dialog generation with cross-domain latent actions. In Proceedings of the 19th Annual SIGdial Meeting on Discourse and Dialogue (pp. 1-10). ACL.	Reimers, N. & Gurevych, I. (2020). Making Monolingual Sentence Embeddings Multilingual using Knowledge Distillation. Proceedings of EMNLP 2020, 4512–4525.
অপর নাম	ML-VAE, cross-lingual VAE, multilingual latent variable model, multilingual generative autoencoder	multilingual sentence representations, cross-lingual sentence embeddings, mSE, multilingual semantic embeddings
সম্পর্কিত	5	5
সারসংক্ষেপ	A Multilingual Variational Autoencoder (ML-VAE) extends the standard VAE framework to handle multiple languages within a shared probabilistic latent space. Language-specific encoders map text from each language into a common continuous representation, while language-specific decoders reconstruct or translate that text. This enables cross-lingual generation, style transfer, and representation learning with or without parallel corpora.	Multilingual sentence embeddings map sentences from many languages into a single shared vector space so that semantically equivalent sentences — regardless of language — land close together. Models such as LaBSE, multilingual Sentence-BERT, and mUSE have made it practical to compare, retrieve, and classify text across 50 to 100+ languages without translating anything first.