Do Linguistic Features Help Deep Learning? The Case of Aggressiveness in Mexican Tweets

In the last years, the control of online user generated content is becoming a priority, because of the increase of online aggressiveness and hate speech legal cases. Considering the complexity and the importance of this issue, this paper presents an approach that combines the deep learning framework with linguistic features for the recognition of aggressiveness in Mexican tweets. This approach has been evaluated relying ona collection of tweets released by the organizers of the shared task about aggressiveness detection in the context of the Ibereval 2018 evaluation campaign. The use of a benchmark corpus allows to compare the results with those obtained by Ibereval 2018 participant systems. However, looking at the achieved results, linguistic features seem not to help the deep learning classification for this task.