The evaluation of humor is essential for modern conversational agents. However, humor perception can be very individual and subjective.
Pavel Braslavski from Ural Federal University in Yekaterinburg and HSE University in Moscow talks about an evaluation of 30 dialog jokes from different sources by almost 300 participants — volunteers and crowd performers. He collected joke ratings along with participants' age, gender, language proficiency, and Big5 personality traits. Results show that crowdsourcing is a viable option for humor evaluation; demographics and joke topics can partly explain the variation in humor judgments