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

Table 10

Obtained results (accuracy, %) for age and gender classification on the PAN author profiling 2016 Spanish training corpus under 10-fold cross-validation.

Feature set	Age
	LR-NP	LR-WP	SVM-NP	SVM-WP
D2V (1-gram)	44.40	46.00	44.80	43.60
D2V (1 + 2-grams)	47.20	52.40	46.40	51.20
D2V (1 + 2 + 3-grams)	51.60	56.00	48.80	57.20
Character 3-grams	50.80	52.00	48.00	47.60
Bag-of-Words	48.00	47.60	44.00	48.40
Feature set	Gender
	LR-NP	LR-WP	SVM-NP	SVM-WP
D2V (1-gram)	71.20	68.00	67.60	64.80
D2V (1 + 2-grams)	69.60	71.60	68.00	70.40
D2V (1 + 2 + 3-grams)	70.40	75.60	69.20	73.60
Character 3-grams	68.00	69.60	61.60	63.20
Bag-of-Words	66.40	63.60	58.80	72.00