AutoExtend: Extending Word Embeddings to Embeddings for Synsets and Lexemes

2015/01/01 by Sascha Rothe, Hinrich Schütze, Sch\"utze, Hinrich · 1 citation
Computer Science · Mathematics · #Artificial intelligence #Computer science #Image (mathematics) #Lexeme #Linguistics #Mathematics #Multimodal Machine Learning Applications #Natural Language Processing Techniques #Natural language processing #Pure mathematics #Similarity (geometry) #Space (punctuation) #Tensor (intrinsic definition) #Topic Modeling #Vector space #Word (group theory) #Word-sense disambiguation #WordNet

paper · pdf · doi:10.3115/v1/p15-1173

openalex publication_date 2015/07/04 · openalex created_date 2019/07/30 · openalex updated_date 2026/08/05

Abstract

We present AutoExtend, a system to learn embeddings for synsets and lexemes. It is flexible in that it can take any word embeddings as input and does not need an additional training corpus. The synset/lexeme embeddings obtained live in the same vector space as the word embeddings. A sparse tensor formalization guarantees efficiency and parallelizability. We use WordNet as a lexical resource, but AutoExtend can be easily applied to other resources like Freebase. AutoExtend achieves state-of-the-art performance on word similarity and word sense disambiguation tasks.

Citations

• SimLex-999: Evaluating Semantic Models With (Genuine) Similarity Estimation
• Exploiting Similarities among Languages for Machine Translation
• Glove: Global Vectors for Word Representation
• A Convolutional Neural Network for Modelling Sentences

Cited by

• Word-sense disambiguation [wikipedia]

Related

• AutoExtend: Combining Word Embeddings with Semantic Resources
• An Ensemble Method to Produce High-Quality Word Embeddings (2016)
• Dense Embeddings Preserving the Semantic Relationships in WordNet
• Extending and Improving Wordnet via Unsupervised Word Embeddings
• Lex2vec: making Explainable Word Embeddings via Lexical Resources