Pre-trained word embeddings using ukWaC corpus.
This repository includes links to download pre-trained word embeddings of different models using ukWaC corpus. These embeddings have been used in the following research papaers:
- Compositional Approaches for Representing Relations Between Words: A Comparative Study
- Why does PairDiff work? – A Mathematical Analysis of Bilinear Relational Compositional Operators for Analogy Detection
- Learning Relation Representations from Word Representations
- Continous Bag of Words (CBOW) 300 dimensions, download
- Skip-Gram (SG) 300 dimensions, download
- Global Vectors (GloVe) 300 dimensions, Download
- Latent Semantic Analysis with Singular Value Decomposition (LSA-SVD), download
- Latent Semantic Analysis with Non-negative Matrix Factorisation (LSA-NMF), download
To read word embeddings, use wordrep.py script in this repository.