Skip-gram word embeddings in hyperbolic space
This work addresses the challenge of embedding words in hyperbolic space for natural language processing, but it is incremental as it builds on existing hyperbolic graph embedding methods without achieving clear superiority.
The paper tackled the problem of learning word embeddings in hyperbolic space by adapting the skip-gram negative-sampling architecture, resulting in hyperbolic embeddings that show potential in low dimensions but do not clearly outperform Euclidean ones on word similarity and analogy benchmarks.
Recent work has demonstrated that embeddings of tree-like graphs in hyperbolic space surpass their Euclidean counterparts in performance by a large margin. Inspired by these results and scale-free structure in the word co-occurrence graph, we present an algorithm for learning word embeddings in hyperbolic space from free text. An objective function based on the hyperbolic distance is derived and included in the skip-gram negative-sampling architecture of word2vec. The hyperbolic word embeddings are then evaluated on word similarity and analogy benchmarks. The results demonstrate the potential of hyperbolic word embeddings, particularly in low dimensions, though without clear superiority over their Euclidean counterparts. We further discuss subtleties in the formulation of the analogy task in curved spaces.