Table 4.

Description of traditional classifiers and parameter settings used in the experiments (the same parameter settings were used for the following 3 targets: relevance, commercial, and sentiment).

Classifiers	Scikit-learn functions (version)	Parameter values
Logistic regression	sklearn.linear_model.LogisticRegression (0.20.3)	All default values except C=0.001
Random forest	sklearn.ensemble.RandomForestClassifier (0.20.3)	All default values except max_features=“sqrt”
Support vector machine	sklearn.linear_model.SGDClassifier (0.20.3)	All default values except α=.01
Naive Bayes	sklearn.naive_bayes.MultinomialNB (0.20.3)	All default values