Tie-breaker: Using language models to quantify gender bias in sports journalism. (arXiv:1607.03895v1 [cs.CL])
Gender bias is an increasingly important issue in sports journalism. In this work, we propose a language-model-based approach to quantify differences in questions posed to female vs. male athletes, and apply it to tennis post-match interviews. We find that journalists ask male players questions that are generally more focused on the game when compared with the questions they ask their female counterparts.