NLP word2vec paper

论文1

[word2vec]Efficient Estimation of Word Representation in Vector Space

part1 Introduction

part2 Model Architectures

part3 New Log-linear Models

part4 Results

part5 Examples of the Learned Relationships

part6 Conclusion

part7 Follow-Up work

论文2

[word2vec]Distributed Representations of Words and Phrase and their Compositionality

part1 Introduction

part2 The Skip-gram Model

1. Hierarchical Softmax
2. Negative Sampling
3. Subsampling of Frequent Words

part3 Empirical Results

part4 Learning Phrases

1. Phrase Skip-Gram Results

part5 Additive Compositionality

part6 Comparison to Published Word Representations

part7 Conclusion