Evaluating group fairness in online tutoring rankings
Mostra el registre complet Registre parcial de l'ítem
- dc.contributor.author Rando Ramírez, Javier
- dc.date.accessioned 2021-11-03T11:18:20Z
- dc.date.available 2021-11-03T11:18:20Z
- dc.date.issued 2021-06
- dc.description Tutors: Carlos Castillo i Emilia Gómezca
- dc.description.abstract This project addresss the evaluation of group fairness in commercial search engines that may lead to discrimination against certain groups of individuals. More precisely, we measure fairness with respect to gender and nationality in a set of platforms where users can search for second language teachers completely online. These websites must rank available teachers for visitors, and disparate exposure may lead to uneven economic benefit. We evaluate if teachers belonging to protected groups are fairly represented in the first positions of the rankings, and we conduct a statistical analysis of price depending on nationality and gender across languages and platforms. Our results put forward that although most lists are fair for all groups, there are some worrisome exceptions. Finally, we found out that women and people from high-income countries charge significantly higher fees for their classes.ca
- dc.format.mimetype application/pdf*
- dc.identifier.uri http://hdl.handle.net/10230/48884
- dc.language.iso engca
- dc.rights © Tots els drets reservatsca
- dc.rights.accessRights info:eu-repo/semantics/openAccessca
- dc.subject.keyword Algorithmic fairnessen
- dc.subject.keyword Ranking biasen
- dc.subject.keyword Data miningen
- dc.subject.keyword Information retrievalen
- dc.subject.keyword Machine learningen
- dc.subject.keyword Language tutoringen
- dc.title Evaluating group fairness in online tutoring rankingsca
- dc.type info:eu-repo/semantics/bachelorThesisca