Vietnamese Word Vector
- The pretrained word vectors are available here (using HCMUT email to access).
- There are 2 types of word vector: unigram and segmented.
- Each of them is trained using 3 methods: CBOW, Skip-gram and GloVe
- To use the unigram word vector, use the
clean_text
function inpreprocess.py
to prepare the texts before mapping them to word vectors. (requireregex
python package). - To use the segmented version, use VnCoreNLP for word segmentation before mapping them to word vectors.