Accountability in algorithmic decision making

Diakopoulos, Nicholas

doi:10.1145/2844110

Cited by 404 publications

(197 citation statements)

References 12 publications

Supporting

Mentioning

170

Contrasting

Unclassified

Order By: Relevance

“…Several design perspectives assume that embracing ethical competence is vital for good design [32,35]. Some argue that game designers, as creators of art, media and cultural artifacts, have a moral and social obligation to understand the impact their design choices have on societythat there are consequences to what is designed and therefore consumed by others [19,40,72,73,74]. Others argue that self-reflection in design is essential for personal and professional growth.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Ethical Stewardship: Designing Serious Games Seriously

Sandovar

Braad

Streicher

et al. 2016

Entertainment Computing and Serious Games

View full text Add to dashboard Cite

Abstract. In this chapter, we propose an ethical framework for serious game design, which we term the Ecosystem for Designing Games Ethically (EDGE). EDGE expands on Zagal's categorization of ethical areas in game design by incorporating the different contexts of design and their use. In addition, we leverage these contexts to suggest four guidelines that support Ethical Stewardship in serious game design. We conclude by discussing a number of specific areas in which ethics plays a role in serious game design. These include games in (a) a military context, (b) the consideration of privacy issues, and (c) the evaluation of game design choices.

show abstract

Section: Discussionmentioning

confidence: 99%

“…The average player is likely unaware of that and hence not in a position to consider the implications of this hidden purpose. Laying open agendas of an algorithm and making choices on an algorithm transparent is the goal of an area recently termed IT Algorithm Accountability [39,40].…”

Section: Serious Games and Educationmentioning

confidence: 99%

Ethical Stewardship: Designing Serious Games Seriously

Sandovar

Braad

Streicher

et al. 2016

Entertainment Computing and Serious Games

View full text Add to dashboard Cite

show abstract

“…These summaries are then used by news media for their stories. Similarly, Facebook uses algorithms to select content for users' newsfeeds (Diakopoulos, 2016). In both cases, users risk having their view of the world influenced by biases or errors in the algorithms, whether intended (e.g., for commercial gain) or accidental.…”

Section: Background: Algorithmic Biasmentioning

confidence: 99%

“…Issues like those above have led to calls for algorithmic accountability and transparency in the sense of the right of society to check that key algorithms behave ethically (Diakopoulos, 2016). Transparency can be impractical due to the complexity of an algorithm -if few understand it -or unethical because it uses private information, such as search engine log files, to learn its behaviour (Ananny & Crawford, in press).…”

Section: Background: Algorithmic Biasmentioning

confidence: 99%

Gender bias in machine learning for sentiment analysis

Thelwall

2018

OIR

View full text Add to dashboard Cite

Purpose: This paper investigates whether machine learning induces gender biases in the sense of results that are more accurate for male authors than for female authors. It also investigates whether training separate male and female variants could improve the accuracy of machine learning for sentiment analysis. Design/methodology/approach: This article uses ratings-balanced sets of reviews of restaurants and hotels (3 sets) to train algorithms with and without gender selection. Findings: Accuracy is higher on female-authored reviews than on male-authored reviews for all data sets, so applications of sentiment analysis using mixed gender datasets will over represent the opinions of women. Training on same gender data improves performance less than having additional data from both genders. Practical implications:End users of sentiment analysis should be aware that its small gender biases can affect the conclusions drawn from it and apply correction factors when necessary. Users of systems that incorporate sentiment analysis should be aware that performance will vary by author gender. Developers do not need to create gender-specific algorithms unless they have more training data than their system can cope with. Originality/value: This is the first demonstration of gender bias in machine learning sentiment analysis.

show abstract

“…First is planning oversight. When major new or revised algorithmic systems are being developed, an independent oversight review could require implementers to submit an algorithms impact statement (11,12). This document would be similar to the environmental impact statements that are now required for major construction programs.…”

Section: A Proactive Approachmentioning

confidence: 99%