English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

A common explanation for negative user impacts of content recommender systemsis misalignment between the platform's objective and user welfare. In thiswork, we show that misalignment in the platform's objective is not the onlypotential cause of unintended impacts on users: even when the platform'sobjective is fully aligned with user welfare, the platform's learning algorithmcan induce negative downstream impacts on users. The source of these userimpacts is that different pieces of content may generate observable userreactions (feedback information) at different rates; these feedback rates maycorrelate with content properties, such as controversiality or demographicsimilarity of the creator, that affect the user experience. Since differencesin feedback rates can impact how often the learning algorithm engages withdifferent content, the learning algorithm may inadvertently promote contentwith certain such properties. Using the multi-armed bandit framework withprobabilistic feedback, we examine the relationship between feedback rates anda learning algorithm's engagement with individual arms for different no-regretalgorithms. We prove that no-regret algorithms can exhibit a wide range ofdependencies: if the feedback rate of an arm increases, some no-regretalgorithms engage with the arm more, some no-regret algorithms engage with thearm less, and other no-regret algorithms engage with the arm approximately thesame number of times. From a platform design perspective, our results highlightthe importance of looking beyond regret when measuring an algorithm'sperformance, and assessing the nature of a learning algorithm's engagement withdifferent types of content as well as their resulting downstream impacts.

Can Probabilistic Feedback Drive User Impacts in Online Platforms?