Research Article

Improving Feature Representation Based on a Neural Network for Author Profiling in Social Media Texts

Table 8

Obtained results (accuracy, %) for gender classification on the PAN author profiling 2015 Italian training corpus under 10-fold cross-validation.

Feature set Gender
LR-NPLR-WPSVM-NPSVM-WP

D2V (1-gram) 71.0571.0571.0568.42
D2V (1 + 2-grams) 71.0571.0568.4271.05
D2V (1 + 2 + 3-grams) 78.9581.5878.9581.58
Character 3-grams 84.2184.2184.2184.21
Bag-of-Words 76.3278.9578.9578.95