I plan to edit the bag of word page in the following ways:

Firstly, adding a session "N-gram model" to describe the relationship between bag-of-words representation and n-gram representation. The former can be viewed as a special case of N-gram model with n=1.

Secondly, I think it is better to provide some background information for the bag-of-word model. Like how it is used in reality by various softwares.

Thirdly, as one of the first step for feature generation, it would be useful to describe the workflow of using BoW representation in a text mining project, including parsing, stemming, removing stopwords, etc.