A Graph-Based Approach for Mitigating Multi-Sided Exposure Bias in Recommender Systems

Fairness is a critical system-level objective in recommender systems that has been the subject of extensive recent research. A specific form of fairness is supplier exposure fairness, where the objective is to ensure equitable coverage of items across all suppliers in recommendations provided to users. This is especially important in multistakeholder recommendation scenarios where it may be important to optimize utilities not just for the end user but also for other stakeholders such as item sellers or producers who desire a fair representation of their items. This type of supplier fairness is sometimes accomplished by attempting to increase aggregate diversity to mitigate popularity bias and to improve the coverage of long-tail items in recommendations. In this article, we introduce FairMatch, a general graph-based algorithm that works as a post-processing approach after recommendation generation to improve exposure fairness for items and suppliers. The algorithm iteratively adds high-quality items that have low visibility or items from suppliers with low exposure to the users’ final recommendation lists. A comprehensive set of experiments on two datasets and comparison with state-of-the-art baselines show that FairMatch, although it significantly improves exposure fairness and aggregate diversity, maintains an acceptable level of relevance of the recommendations.

The molecular forms of BMP15 in a range of mammalian species

<p>BMP15 is an oocyte-secreted growth factor that is critical for ovarian follicular development and fertility in mammals. To improve our understanding of the species-specific effects of BMP15 on fertility, the aim of this study was to elucidate the molecular forms of BMP15 in mammals with high (rat and pigs) and low (sheep and deer) litter sizes. Western blotting experiments were undertaken using a monoclonal antibody directed against a semi-conserved area on the mature region of E.coli-expressed ovine BMP15. Within the oocyte lysate of all species tested, the predominant form was promature BMP15, although mature BMP15 was also present. Within oocyte-conditioned media of sheep, pigs and deer, the promature:mature BMP15 ratio dropped significantly from what was seen inside the oocyte, with promature BMP15 only slightly predominant in deer, and relatively equal amounts of each protein were present in pigs and sheep. No BMP15 was detected in the oocyte-conditioned media of rats. Cross-linking studies did not provide any evidence of BMP15/BMP15 homodimers or BMP15/GDF9 heterodimers in any species tested. Intra-oocyte levels of both promature and mature BMP15, from highest to lowest, were detected in deer, followed by sheep and pigs (similar), and then rats (significantly lower). In the oocyte-conditioned media, with the exception of the rat where BMP15 was absent, there was a similar pattern of promature BMP15 levels observed between species, whereas mature BMP15 levels showed no inter-species variation. In summary, the molecular forms and relative amounts of BMP15 protein differ across species. Generally, high BMP15 levels were associated with low litter size; however the pig is an exception. Furthermore, the molecular forms observed in this study differed from those reported in studies using recombinant BMP15 suggesting that recombinantly-expressed protein may not provide a fair representation of native BMP15.</p>

The molecular forms of BMP15 in a range of mammalian species

<p>BMP15 is an oocyte-secreted growth factor that is critical for ovarian follicular development and fertility in mammals. To improve our understanding of the species-specific effects of BMP15 on fertility, the aim of this study was to elucidate the molecular forms of BMP15 in mammals with high (rat and pigs) and low (sheep and deer) litter sizes. Western blotting experiments were undertaken using a monoclonal antibody directed against a semi-conserved area on the mature region of E.coli-expressed ovine BMP15. Within the oocyte lysate of all species tested, the predominant form was promature BMP15, although mature BMP15 was also present. Within oocyte-conditioned media of sheep, pigs and deer, the promature:mature BMP15 ratio dropped significantly from what was seen inside the oocyte, with promature BMP15 only slightly predominant in deer, and relatively equal amounts of each protein were present in pigs and sheep. No BMP15 was detected in the oocyte-conditioned media of rats. Cross-linking studies did not provide any evidence of BMP15/BMP15 homodimers or BMP15/GDF9 heterodimers in any species tested. Intra-oocyte levels of both promature and mature BMP15, from highest to lowest, were detected in deer, followed by sheep and pigs (similar), and then rats (significantly lower). In the oocyte-conditioned media, with the exception of the rat where BMP15 was absent, there was a similar pattern of promature BMP15 levels observed between species, whereas mature BMP15 levels showed no inter-species variation. In summary, the molecular forms and relative amounts of BMP15 protein differ across species. Generally, high BMP15 levels were associated with low litter size; however the pig is an exception. Furthermore, the molecular forms observed in this study differed from those reported in studies using recombinant BMP15 suggesting that recombinantly-expressed protein may not provide a fair representation of native BMP15.</p>

Stakes of Knowing the Truth: A Motivational Perspective on the Popularity of a Controversial Scientific Theory

The aim of this article is to provide a different perspective on people's beliefs regarding controversial scientific information. We emphasize that, although people generally aim at getting a fair representation of reality, accuracy about scientific issues only matters to the extent that individuals perceive it as useful to achieve their own goals. This has important consequences in terms of how anti-science attitudes as well as epistemically questionable beliefs must be interpreted, which has consequences for addressing misinformation. We argue that most people who endorse scientific misinformation are not truly interested in its accuracy, and rather that plausibility at face value often suffices when it is meant to be used for social purposes only. We illustrate this view with the example of hydroxychloroquine, which was considered as potential treatment for Covid-19, and which has been the subject of much media hype and public concern, particularly in France.

Toward Learning Trustworthily from Data Combining Privacy, Fairness, and Explainability: An Application to Face Recognition

In many decision-making scenarios, ranging from recreational activities to healthcare and policing, the use of artificial intelligence coupled with the ability to learn from historical data is becoming ubiquitous. This widespread adoption of automated systems is accompanied by the increasing concerns regarding their ethical implications. Fundamental rights, such as the ones that require the preservation of privacy, do not discriminate based on sensible attributes (e.g., gender, ethnicity, political/sexual orientation), or require one to provide an explanation for a decision, are daily undermined by the use of increasingly complex and less understandable yet more accurate learning algorithms. For this purpose, in this work, we work toward the development of systems able to ensure trustworthiness by delivering privacy, fairness, and explainability by design. In particular, we show that it is possible to simultaneously learn from data while preserving the privacy of the individuals thanks to the use of Homomorphic Encryption, ensuring fairness by learning a fair representation from the data, and ensuring explainable decisions with local and global explanations without compromising the accuracy of the final models. We test our approach on a widespread but still controversial application, namely face recognition, using the recent FairFace dataset to prove the validity of our approach.

Connecting Stereotypes of South Asians in the American Media to South Asian Teenagers’ Internalized Racism

2020 put the entire world upside down in its call for racial equity and justice. During these unveiling times, minorities in America have come forward in protest of racial and ethnic stereotypes which Hollywood still plays into. This research paper capitalized on the removal and protest of these characters and dove more specifically into South Asian stereotypes which have been reinforced by television show and movie characters. The claim in these cries of disapproval for these stereotypes was that it caused internalized racism in people who were brought up watching these stereotypes forced upon them. As building a body of knowledge progressed, the unmade connection of stereotypes to internalized racism in formal research became apparent, and therefore the research paper intended to see if there was a correlation between South Asian stereotypes and internalized racism. In order to identify internalized racism among South Asian teenagers, an open-ended survey was used as a method, followed by a mixed analysis to identify self-stereotyping indicators which would denote internalized racism. The research in the end did find a staunch correlation between the South Asian stereotypes and internalized racism in today’s South Asian teenagers. While today’s teenagers cannot rid themselves of the stereotypes already imposed on them, future South Asian children might possibly have accurate and fair representation in subsequent years.

‘Difference Is the One Thing That We Have in Common’

Television production is championing diversity in representation with record numbers compared to previous years. Netflix’s Sense8 is definitely amongst the highest scoring shows as concerns intersectional representation. Such remarkable representation was worth the 2016 GLAAD Outstanding Drama Series award, a prize granted to the most diverse television shows. However, this applause is geared solely to numerical representations while current academic discussion focuses more on the concept of fair representations. Not only is being represented of importance, but how one is represented. The present paper employs photovoice and photo elicitation to investigate how Sense8 fans articulate what constitutes a fair representation of queer gender identities within the show. The present research addresses two gaps in the literature. First, a methodological one: the employment of creative visual methodologies to transcend the limitations of the most common methods used for audience research —interviews and focus groups. Secondly, this study follows the contemporary conversation around fair representation by addressing what is now a gap in the existing literature on queer television: what is fair representation from an audience perspective? The results of this study show how audiences’ perspectives on fair representation differ from those formulated in public and academic debates.