UMUAI Special issue on: Fair, Accountable, and Transparent Recommender Systems
For extended abstracts: 19th of June (extended)
For full papers (for accepted abstracts): Aug. 2nd, 2019
Any updates to the special issue will be published here.
Full papers to be submitted via the UMUAI editorial manager: https://www.editorialmanager.com/umui/Default.aspx
BACKGROUND AND SCOPE
This special issue addresses research on responsible design, maintenance, evaluation, and study of recommender systems. It is a venue for work that has evolved out of recent workshops and conferences (e.g, FairUMAP, FATRec, FATML, FAT*) on fair, accountable, and transparent (FAT) recommender systems. In particular, it addresses what it means for a recommender system to be responsible, and how to assess the social and human impact of recommender systems. The questions addressed under each criterion are seen as follows:
- Fairness: what might ‘fairness’ mean in the context of recommendation? How could a recommender be unfair, and how could we measure such unfairness?
- Accountability: to whom, and under what standard, should a recommender system be accountable? How can or should it and its operators be held accountable? What harms should such accountability be designed to prevent?
- Transparency: what is the value of transparency in recommendation, and how might it be achieved? How might it trade off with other important concerns?
The special issue covers several aspects of FAT recommendation. Of particular interest are case studies of successful FAT practices in domains with large societal impact (e.g., healthcare, insurance, lending, news, educational systems), but also with large financial impact (e.g., ecommerce sites, travel booking sites, job search sites, dating sites, etc.). The scope of the special issue includes but is not limited to:
- Fairness of user and item models (e.g., low confidence recommendations, disbalanced data, measures of diversity, low confidence recommendations)
- Accountability of user and item models (e.g., accountability by or for different stakeholders, requirements on modeling to enable accountability)
- Transparency of user and item models (e.g., explanatory needs for different user groups, explaining individual and global consumptions patterns)
- Fairness of recommendations (e.g., trade-offs between criteria, bias for classes of items or users)
- Accountability of recommendations (e.g., mechanisms for reporting/accounting, balancing filtering and completeness)
- Transparency of recommendations (e.g., explanatory visualizations, user control, comparing explanatory aims)
- Methodologies to assess Fairness (e.g., metrics for balance, diversity, and other social welfare criteria; evaluation simulations; assessing stakeholder specific bias)
- Methodologies to assess Accountability (e.g., metrics and user studies of accountability mechanisms)
- Methodologies to assess Transparency (e.g., metrics and evaluation frameworks for assessing the impact of interface or interaction strategies)
- Impacts of Fairness practices (e.g., balancing needs of different groups of users or stakeholders in recommender systems)
- Impacts of Accountability practices (e.g., mechanisms for reporting data and models or decisions about them)
- Impacts of Transparency practices (e.g., counterfactuals and what-if recommendations)
PAPER SUBMISSION & REVIEW PROCESS
Submissions will be pre‐screened for topical fit based on extended abstracts.
(deadlines after abstract submission are only indicative)
abstracts (up to three pages in journal format)
Deadline for extended abstracts: June 19th, 2019 extended
Notification about extended abstracts: June 26th, 2019
Deadline for full manuscript submission: August 16th, 2019 extended
Notification 1st cycle: October 14, 2019 (estimation only)
Deadline for revised manuscripts: December 12, 2019
Notification 2nd and final cycle: January 17, 2020
Deadline for camera-ready manuscripts: February 28, 2020