MUSE aligns monolingual embeddings with an unsupervised orthogonal map

Conneau et al. (MUSE) show two independently, monolingually trained fastText embedding spaces are related by a single orthogonal linear map, recoverable with no parallel data [conneau-etal-2018-muse-word-translation-without-parallel-data] An adversarially-trained map is iteratively refined via orthogonal Procrustes on its own mutual-nearest-neighbor pairs [conneau-etal-2018-muse-word-translation-without-parallel-data] The resulting map achieves bilingual dictionary induction accuracy rivaling or exceeding supervised baselines across several language pairs [conneau-etal-2018-muse-word-translation-without-parallel-data] It is a foundational precursor to later per-language affine cross-lingual-overlap and vec2vec-style universal-alignment findings [conneau-etal-2018-muse-word-translation-without-parallel-data]