A Survey of Embedding Space Alignment Methods for Language and Knowledge Graphs
arXiv:2010.13688
Abstract
Neural embedding approaches have become a staple in the fields of computer vision, natural language processing, and more recently, graph analytics. Given the pervasive nature of these algorithms, the natural question becomes how to exploit the embedding spaces to map, or align, embeddings of different data sources. To this end, we survey the current research landscape on word, sentence and knowledge graph embedding algorithms. We provide a classification of the relevant alignment techniques and discuss benchmark datasets used in this field of research. By gathering these diverse approaches into a singular survey, we hope to further motivate research into alignment of embedding spaces of varied data types and sources.
arXiv comment27 pages, 2 figures
References in corpus (16)
- Efficient Estimation of Word Representations in Vector Space
- Sequence to Sequence Learning with Neural Networks
- Exploiting Similarities among Languages for Machine Translation
- Snorkel: Rapid Training Data Creation with Weak Supervision
- An efficient framework for learning sentence representations
- Representation Learning on Graphs: Methods and Applications
- Traversing Knowledge Graphs in Vector Space
- Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond
- Cross-lingual Entity Alignment via Joint Attribute-Preserving Embedding
- Loss in Translation: Learning Bilingual Word Mapping with a Retrieval Criterion
- One-Shot Relational Learning for Knowledge Graphs
- Extreme Extraction: Only One Hour per Relation
- Few-Shot Knowledge Graph Completion
- Bilingual Lexicon Induction with Semi-supervision in Non-Isometric Embedding Spaces
- A Structural Representation Learning for Multi-relational Networks
- Lost in Evaluation: Misleading Benchmarks for Bilingual Dictionary Induction