A Hierarchical Neural-Network-Based Document Representation Approach for Text Classification
Table 2
Statistics of datasets (the vocabulary in datasets has gone through data cleaning, excluding single characters and punctuations as well as retaining only the lemmatized words).
