The Pro version
includes all example code bundled in, +30 pages of additional content
, and addresses two new important topics.
1. Backpropagation training of word2vec.
The original eBook explains all of the modifications made to a neural network to implement word2vec, but what does the weight update process actually look like? This is covered in detail in Chapter 5 - Backpropagation, and is even implemented from scratch in the included example code.
2. Subword information & fastText
Did you know that Mikolov, the main author of word2vec, has published further work on word2vec in the form of the fastText library from Facebook? Chapter 6 - Subwords and fastText explains the enhancements made and the trade-offs it makes. The example code will have you training a word model with fastText, and comparing the results to the original word2vec to see these trade-offs first hand.