Advanced search


Carlos D. Martínez-Hinarejos, Ruxandra Dumitru. Detection of sexist language by using text classification techniques. Proceedings of IberSpeech 2014, 2014. pp. 159-168.

Sexist language is the use of language that reinforces differences between women and men. Sexist language has a strong social influence, since propagates beliefs that get reflected in social behaviour, and therefore must be avoided in order to guarantee equal opportunities for women and men. This is specially important in public institutions documents. Nowadays, detecting and correcting sexist language in official documents is a hand-made task. However, this task could benefit of the use of Text Classification techniques in order to detect sexist expressions and allow their correction. We present the use of several text classifiers in the detection of sexist language in Spanish official documents. Results show that the automatic detection guarantees a low rate of sexist expressions in the final document (about 10%) with an admissible extra review work.