Grep-BiasIR: A Dataset for Investigating Gender Representation-Bias in Information Retrieval Results

Krieg, Klara; Parada-Cabaleiro, Emilia; Medicus, Gertraud; Lesota, Oleg; Schedl, Markus; Rekabsaz, Navid

doi:10.48550/arxiv.2201.07754

Cited by 4 publications

(7 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Future work might also try to extend the developed experimental setup to include other SE-related concepts found to contain biases, such as automated query suggestions [3]. Here, our Grep-BiasIR dataset [16] opens the possibility to conduct further related experiments. Till then, as one of the first studies to explore effects of perceived gender biases in retrieval results on relevance judgements, this study presents an initial empirical contribution.…”

Section: Discussionmentioning

confidence: 99%

“…Data. We used a subset of queries and documents provided by the recently released Grep-BiasIR dataset [16]. The Grep-BiasIR dataset provides 118 bias-sensitive queries.…”

Section: Experiments Setupmentioning

confidence: 99%

“…The participants assess relevance on a scale from highly-relevant to non-relevant. The experiments are conducted based on a set of queries and documents from the recently release Grep-BiasIR dataset [16]. We repeat the experiments on two settings.…”

Section: Introductionmentioning

confidence: 99%

“…Bias-sensitive refers to a gender-neutral query whose bias in its retrieval results is considered as socially problematic[16,23].…”

mentioning

confidence: 99%

See 3 more Smart Citations

Do Perceived Gender Biases in Retrieval Results Affect Relevance Judgements?

Krieg¹,

Parada-Cabaleiro²,

Schedl³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

This work investigates the effect of gender-stereotypical biases in the content of retrieved results on the relevance judgement of users/annotators. In particular, since relevance in information retrieval (IR) is a multi-dimensional concept, we study whether the value and quality of the retrieved documents for some bias-sensitive queries can be judged differently when the content of the documents represents different genders. To this aim, we conduct a set of experiments where the genders of the participants are known as well as experiments where the participants' genders are not specified. The set of experiments comprise of retrieval tasks, where participants perform a rated relevance judgement for different search query and search result document compilations. The shown documents contain different gender indications and are either relevant or non-relevant to the query. The results show the differences between the average judged relevance scores among documents with various gender contents. Our work initiates further research on the connection of the perception of gender stereotypes in users with their judgements and effects on IR systems, and aim to raise awareness about the possible biases in this domain.

show abstract

Section: Discussionmentioning

confidence: 99%

“…Data. We used a subset of queries and documents provided by the recently released Grep-BiasIR dataset [16]. The Grep-BiasIR dataset provides 118 bias-sensitive queries.…”

Section: Experiments Setupmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

“…Bias-sensitive refers to a gender-neutral query whose bias in its retrieval results is considered as socially problematic[16,23].…”

mentioning

confidence: 99%

See 2 more Smart Citations

Do Perceived Gender Biases in Retrieval Results Affect Relevance Judgements?

Krieg¹,

Parada-Cabaleiro²,

Schedl³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…In contrast to this work, we study bias mitigation from the consumer side. More recently, based on adversarial training, Wu et al [30] explore the mitigation of consumer bias in news recommendation, and several recent studies [16,17,23,33] approach fairness in the representation of gender-related documents in information retrieval. Our work extends these studies by introducing a novel bias-aware recommendation model based on variational autoencoders.…”

Section: Introductionmentioning

confidence: 99%

Unlearning Protected User Attributes in Recommendations with Adversarial Training

Ganhör

Penz

Rekabsaz

et al. 2022

Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval

Self Cite

View full text Add to dashboard Cite

Collaborative filtering algorithms capture underlying consumption patterns, including the ones specific to particular demographics or protected information of users, e. g., gender, race, and location. These encoded biases can influence the decision of a recommendation system (RS) towards further separation of the contents provided to various demographic subgroups, and raise privacy concerns regarding the disclosure of users' protected attributes. In this work, we investigate the possibility and challenges of removing specific protected information of users from the learned interaction representations of a RS algorithm, while maintaining its effectiveness. Specifically, we incorporate adversarial training into the state-of-the-art MultVAE architecture, resulting in a novel model, Adversarial Variational Auto-Encoder with Multinomial Likelihood (Adv-MultVAE), which aims at removing the implicit information of protected attributes while preserving recommendation performance. We conduct experiments on the MovieLens-1M and LFM-2b-DemoBias datasets, and evaluate the effectiveness of the bias mitigation method based on the inability of external attackers in revealing the users' gender information from the model. Comparing with baseline MultVAE, the results show that Adv-MultVAE, with marginal deterioration in performance (w. r. t. NDCG and recall), largely mitigates inherent biases in the model on both datasets. CCS CONCEPTS• Information systems → Collaborative filtering; • Computing methodologies → Neural networks.

show abstract