Discriminatory Expressions to Improve Model Comprehensibility in Short Documents

Microblogging sites are being used as analysis avenues due to their peculiarities (promptness, short texts...). Lately, researchers have focused mainly in classification performance rather than interpretability. When the problem requires transparency, it is necessary build interpretable pipelines, and even though, resulting models too complex be considered comprehensible, making impossible for humans understand actual decisions. This paper presents a feature selection mechanism that able improve comprehensibility by using less but more meaningful features. Results show our proposal better most stable one terms of accuracy, generalisation microblogging context.