Papers for

online platform designers

Papers whose findings have a practical use for this group, as judged from the abstract. Open a paper to read what it means in practice.

Expert users get more team opportunities and success on Kaggle forums

Who Shows Up, Who Benefits, and Who Wins? Expert-Novice Disparities in Team-Seeking Forums and Competition Performance on Kaggle

Abstract: Team formation is the first social gate to collaboration in online communities, yet it remains unclear whether public team-formation mechanisms broaden access to collaborators or reproduce existing hierarchies. We examine this question on Kaggle, a data science competition platform whose "Looking for a Team" (L4T) megathreads let participants publicly seek teammates for on-platform competitions. Grounded in Communities of Practice theory, we conduct two complementary studies of 18 competitions held in 2023. Study 1 analyzes the reply network among 671 megathread participants and finds that 89.3% are low-tier users, yet higher-tier users receive significantly more replies (Kruskal-Wallis H=16.46, p=.00246) and achieve higher rates of forum-visible teaming success (52.9% vs. 32.6%; chi-squared=10.53, p=.001), a proxy for eventual team membership. Linking megathread posters to verified team rosters from 16 competitions partially supports this proxy. Twenty-six final teams contained two or more megathread posters, indicating that megathread interaction sometimes, but not always, maps onto realized teaming. Study 2 links participation traces from 219 users to private leaderboard outcomes and finds that productive participation (submission count rho=.435, cross-competition participation rho=.247) is positively associated with performance, whereas megathread-specific discussion is not (rho=-.021). The submission-performance association appears stronger in higher tiers, although the novice subgroup estimate is imprecise. Together, the findings suggest that open access to team-seeking forums does not guarantee equal participation. Visible expertise shapes who attracts attention, and engagement in the community's core practice is more strongly linked to success than discussion alone. We discuss implications for designing more equitable team formation in online collaboration environments.

Fri 13 NovComputers and SocietyHuman-Computer Interaction
The gist
Finding teammates online can be hard, and this study looks at how people find team members on Kaggle, a data science contest site. Most people seeking teams are newcomers, but experienced users get more replies and more chances to join teams. Just talking in team forums doesn’t predict winning, but being active in contest work does. This suggests that open team-seeking spaces don’t always help everyone equally.
Open → 2611.18983v1

Wikipedia replies boost certain newcomers editing chances

Who Benefits from Wikipedia's Growth Mentorship? The Role of Newcomer Information Needs

Abstract: Evaluations of Wikipedia's programs for new editors usually ask whether a program works on average; we ask which newcomers benefit when their question to a mentor gets a reply. English Wikipedia's Growth mentorship system assigns every new account a volunteer mentor, and about a quarter of the questions newcomers send never get a reply. Using the first questions of 34,073 newcomers (2021 to 2026), we compare newcomers who received a reply with similar newcomers who did not, matched on 164 characteristics recorded before the question (propensity score stratification). (1) Does a reply help newcomers keep editing? Yes, in the short term: in the 14 days after the reply, newcomers who received one were 1.8 percentage points (pp) more likely to edit an article (22% vs. an estimated 20% without a reply) and 1.8pp more likely to edit on two or more days. Later in the first year, the only difference that remains significant after correction for multiple testing is in the share who edit an article at least once between six months and one year after the reply (+1.5pp). (2) Does the effect depend on what the newcomer asked? Yes. We sort questions by Morrison's (1993) information-seeking types, which earlier studies measured mostly through self-reports, using a language model checked against human labels. A reply has significant effects for questions about rules and norms (+3.4pp), questions that show the newcomer's earlier editing (+3.4pp) and specific how-to questions (+2.4pp). For newcomers who do not know what to work on, a reply has no detectable effect, and the effect is 4.2pp smaller than for newcomers with other questions. (3) Which replies are followed by more editing? Among newcomers who received a reply, those whose reply came sooner or stayed closer to the topic of their question edited more.

Sun 8 NovSocial and Information NetworksHuman-Computer Interaction
The gist
Not all newcomers to Wikipedia benefit equally from getting help from mentors. The authors found that when new editors ask questions and get answers, they tend to edit more in the short term, especially if their questions are about rules, editing tasks, or specific how-to issues. However, newcomers who just don’t know what to do don’t seem to benefit from replies. Also, faster and more focused replies lead to more editing. This helps us understand how to support new Wikipedia editors better.
Open → 2611.11083v1

Algorithms improve fairness measures in group decision making

Computational Aspects of Distortion

Abstract: The distortion framework in social choice theory allows quantifying the efficiency of (randomized) selection of an alternative based on the preferences of a set of agents. We make two fundamental contributions to this framework. First, we develop a linear-programming-based algorithm for computing the optimal randomized decision on a given instance, which is simpler and faster than the state-of-the-art solutions. For practitioners who may prefer to deploy a classical decision-making rule over the aforementioned optimal rule, we develop an algorithm based on non-convex quadratic programming for computing the exact distortion of any (and the best) randomized positional scoring rule. For a small number of alternatives, we find that the exact distortion bounds are significantly better than the asymptotic upper bounds established in prior literature and lead to different recommendations on which rules to use. These results rely on a novel characterization of the instances yielding the worst distortion, which may be of independent interest.

Tue 3 NovComputer Science and Game Theory
The gist
Choosing the best option for a group based on individual preferences can be tricky when preferences differ. The paper presents new algorithms that better calculate how efficient or fair a random choice method is, compared to past tools. One method uses linear programming for quick and simple results, while another finds exact fairness scores for popular voting rules. The authors also identify specific tricky cases that can cause the biggest fairness issues, helping understand when choices might be less fair.
Open → 2611.05257v1

AI voter models treat groups differently and can be unfair

AI Alignment for Digital Democracy: Fairer Voters Representation against LLM Biases

Abstract: AI representation is emerging in digital democracy, assisting in voter emulation and decision-making. Large language models (LLMs) behave differently depending on the model, the prompt context, and the voter information they receive, representing some groups more accurately than others and potentially making the emulation unfair to certain groups. We investigate demographic biases in LLM-emulated electorates: how accurately LLMs represent individual votes, whether some voter groups are represented more accurately than others, and whether AI alignment can reduce these disparities without sacrificing overall accuracy. Using three U.S. presidential elections, we compare a closed-source and an open-source LLM across voters' demographic features and test AI alignment strategies toward fairer voter representation. Removing gender from the voter profile reduces the largest group error of the models by 22-42%, and removing education reduces it by 64%, whereas removing self-reported race increases the mean group error by over 40%, all without lowering accuracy. Common alignment strategies, such as repeated queries or fairness instructions, do not reduce these errors and can double them.

Fri 23 OctComputers and SocietyArtificial Intelligence
The gist
Large language AI models are used to mimic how voters might choose in elections, but they don't represent all voter groups equally well, which can be unfair. The authors studied how these AI models perform on different demographic groups using past U.S. elections and found that removing some personal details from voter profiles can reduce errors for some groups without hurting accuracy. However, common methods to fix fairness problems often don't work and can even make things worse. This points to the challenge of making AI voter models both accurate and fair to everyone.
Open → 2610.32780v1

Iterated round robin methods may not always settle but some strategies guarantee fairness

On the Convergence of Iterated Round Robin

Abstract: Round Robin (RR) is a simple and widely used picking sequence mechanism for allocating indivisible goods, known to guarantee envy-freeness up to one item (EF1) when preferences are additive. Despite being manipulable (meaning that agents may have an incentive to strategically misreport their preferences), it is known to always admit at least one pure Nash equilibrium. In this paper, we study an iterated variant of RR, in which agents observe the outcome of the mechanism at each iteration and may revise their reported preferences through a best-response dynamics. We first consider a full-information setting, where agents observe both the outcome and the reported preferences of all the agents. In this setting, we show that cycles can occur even under strong restrictions on the manipulations carried out by the agents. We nonetheless identify a natural restriction (item-preserving swaps), under which convergence to a pure Nash equilibrium is guaranteed to exist for any number of agents and items. We also show that restricting to adjacent swaps guarantees convergence for two agents and a limited number of items, leaving the question open for unbounded number of items. We then turn to a partial-information setting, in which agents only observe the resulting allocation and must then infer the preference profiles of the others assuming consistency with the RR sequence. We study three behaviours guiding the agents when deciding how to manipulate under uncertainty on the other agents' preferences. We show that two of the studied behaviours, despite being quite natural, may cycle while a more cautious rule guarantees iterated RR to be strategyproof.

Fri 16 OctComputer Science and Game Theory
The gist
When sharing items among people, Round Robin is a common fair method where everyone takes turns picking items. The paper studies what happens when people change their stated preferences repeatedly to get better outcomes. The authors find that this process can sometimes get stuck in cycles and never settle on a stable sharing plan. But if people’s preference changes follow certain restrictions, the sharing method will eventually settle into a stable and fair outcome. They also examine situations where people don’t see all preferences and find some cautious strategies ensure fairness.
Open → 2610.23040v1

New fairness rules extend voter influence to complex preferences

Robust Proportionality Axioms for Additive Utilities

Abstract: We study the model of participatory budgeting, in which a subset of candidates is selected subject to a budget constraint. Voters express their preferences by assigning a numerical utility to each candidate. Our focus is on formal analysis of proportionality, a central notion of fairness in multi-winner elections. Intuitively, proportionality means that each group of voters should influence the election outcome roughly in proportion to its size. Recent "plus" axioms of proportionality are particularly appealing notions: they are verifiable in polynomial time, and they provide strong representation guarantees without requiring groups of voters to have cohesive preferences. However, their existing formulations are restricted to approval-based preferences. We generalise extended justified representation plus (EJR+), and Core plus (Core+) to general additive utilities. Our approach builds on a payment-based interpretation of proportionality: candidates' costs are shared among their supporters, with the aim of balancing voters' total payments.

Sun 11 OctComputer Science and Game Theory
The gist
When groups of people vote to pick projects within a budget, it's important that each group’s influence matches their size. The authors look at fairness rules that ensure this proportional influence even when voters rate projects with numbers instead of just yes or no. They generalize existing rules that worked only with simple approvals to handle these detailed ratings. Their method shares project costs among supporters fairly, balancing how much each voter 'pays' in influence. This gives stronger guarantees that different groups get fair representation.
Open → 2610.15516v1

Opposing comments boost engagement and shift attitudes on social media

The Influence of the Vocal Few: Evidence from Social Media Comments

Abstract: Online comment sections let a small number of vocal individuals reach far beyond their own networks. We conduct a large-scale field experiment on Facebook that randomizes the presence and stance of comments beneath posts for a racial justice organization, reaching around one million U.S. users. Opposing comments increase reactions, comments, and link clicks by 15-43 percent relative to no comments, whereas supportive comments have little effect. A complementary survey experiment shows that similar opposing comments make attitudes less progressive and reduce donations to the organization. Through a common feature of online platforms, the vocal few can exert outsized influence.

Sat 26 SeptSocial and Information Networks
The gist
Sometimes, a few people who speak loudly in online comment sections can change how many others react and behave. The researchers did an experiment on Facebook posts by a racial justice group and found that when people post opposing comments, more users click and respond compared to when no comments appear. However, comments that support the group don’t have much effect. Another survey showed that seeing opposing comments can make people less supportive and less willing to donate to the cause. This shows how a small number of vocal commenters can influence a lot of others online.
Open → 2609.32880v1

Proportional representation challenges with changing ranked votes over time

Proportional Representation in Temporal Voting with Ranked Preferences

Abstract: We study proportional representation in temporal voting, where one candidate is selected in each round. While prior work has focused on approval ballots, we consider ranked preferences, which may change over time. A natural approach treats each voter's top candidates as approved, but the right cutoff may differ across voters and rounds. We therefore require proportionality to hold for every admissible choice of cutoffs, whether fixed and common, common but varying across rounds, or set individually for each voter in each round. Combining these interpretations with temporal versions of justified representation (JR), proportional JR (PJR), extended JR (EJR), and proportionality for solid coalitions (PSC) gives us a hierarchy of axioms. We ask which of these axioms can be guaranteed, and with how much knowledge of the future. Unlike with approval ballots, no version of EJR can be guaranteed, and for the other axioms, flexibility in the cutoffs comes at a price. With a fixed common cutoff, JR, PJR, and PSC can be guaranteed, but only by rules that see all preferences in advance. Once the cutoff may vary across rounds, even such rules cannot guarantee JR or PSC for groups that agree in only some rounds. For groups that agree in every round, however, knowing only the number of rounds suffices for PJR in polynomial time, and PSC needs no knowledge of the future at all. Under individual cutoffs, no version of JR or PJR can be guaranteed, yet a rule as simple as serial dictatorship achieves PJR up to an additive loss that no rule can improve on, however much it knows. Natural preference restrictions restore exact guarantees. Finally, we show that checking our axioms is often coNP-complete; but perhaps surprisingly, a stronger axiom can be easier to check.

Thu 24 SeptComputer Science and Game TheoryArtificial Intelligence
The gist
This paper looks at how to fairly represent voters when decisions happen one by one over time and voters can rank their preferences differently each time. The authors find that ensuring fair representation is more complicated when voters rank rather than approve candidates, especially if the criteria for which candidates count as approved changes from round to round or voter to voter. Some fairness guarantees require knowing future voter preferences in advance, while others can be met with limited or no future knowledge depending on the situation. They also show that checking whether these fairness conditions are met can be computationally hard.
Open → 2609.30555v1

AI can help people talk and decide better in large groups

AI Should Facilitate Democratic Deliberation at Scale

Abstract: AI systems can strengthen democracy by supporting deliberation at scale by addressing cognitive, social, platform-design, and market-driven frictions, while preserving human agency. Unlike proposals such as liquid democracy that restructure representation through vote delegation, in this position paper, we argue that AI-assisted deliberation offers a more promising path by lowering barriers to meaningful engagement without substituting machine judgment for human choice. Drawing on evidence from online deliberation platforms and experimental research, we identify four guiding principles: preserving agency and autonomy, encouraging mutual respect, promoting equality and inclusiveness, and augmenting rather than substituting active citizenship. We also address critical challenges, including alignment, sycophancy, training bias, and over-reliance on AI systems. We call on the machine learning community to develop deliberation-focused AI systems evaluated not on engagement metrics but on their capacity to facilitate informed, representative, and friction-robust discourse.

Thu 17 SeptHuman-Computer InteractionArtificial IntelligenceComputation and Language
The gist
Making decisions together in big groups can be hard because people face many obstacles like misunderstandings or unfair attention. The authors argue that AI tools can help by making it easier for everyone to join in and share their views respectfully without replacing human choices. They suggest that these AI systems should respect people's freedom, encourage fairness, and support active participation. They also warn about problems like AI bias and overtrust to watch out for. Finally, they encourage AI builders to focus on tools that really improve thoughtful and honest discussions, not just increase clicks or comments.
Open → 2609.20059v1

Anchoring influences how groups reach decisions over time

Anchored Sequential Deliberation

Abstract: Sequential deliberation is a mechanism for collective decision making: at each round, a uniformly randomly selected pair is asked to revise a collective outcome, which then becomes the reference point for the next round. Existing theory by Fain et al.~\cite{fain2017sequential} treats the current outcome solely as the disagreement alternative in bargaining. Yet an existing draft, policy, or proposal might carry social influence and anchor participants' expressed positions toward the status quo. We introduce anchored sequential deliberation on a one-dimensional decision space. In each round, two participants with bliss points $U$ and $V$ shift their positions toward the previous outcome $O_{t-1}$ with anchoring strength $λ$, then Nash-bargain using $O_{t-1}$ as the disagreement alternative. The update simplifies to $O_t=(1-λ)\mathsf{Median}\{U,V,O_{t-1}\}+λO_{t-1}$. We establish a convergence--stability trade-off. For every population distribution and $λ<1$, the process has a unique stationary distribution. A monotone coupling yields a $1$-Wasserstein contraction factor of at most $\frac{1+λ}{2}$ and at least $λ$; thus, stronger anchoring slows mixing. On the other hand, stationary social cost weakly decreases with $λ$, although the worst-case distortion remains $\frac{1+\sqrt{2}}{2}$. We also identify a unique \emph{deliberative fixed point}, where the expected unanchored movement is zero, and prove that the stationary distribution concentrates around it as $λ\to 1$. For the uniform population, stationary distortion lies between $1+\frac{1-λ}{9+7λ}$ and $1+\frac{1-λ}{6(1+λ)}$, with both bounds approaching $1$ as $λ\to1$. Simulations for uniform and Beta populations show that stronger anchoring slows mixing, concentrates the stationary distribution, and lowers stationary distortion in these instances.

Tue 15 SeptComputer Science and Game TheoryMultiagent Systems
The gist
This paper looks at how groups make decisions when two people talk and update the group’s choice step by step. The authors study what happens if people are influenced or anchored by the previous decision outcome when negotiating. They find that stronger anchoring slows how fast opinions change but tends to keep decisions closer to the best overall choice. Their math shows how this balance works and predicts stable decisions around certain points. Simulations confirm that anchoring shapes both the speed and quality of group decisions.
Open → 2609.16673v1

Personality mix shapes social network polarization and collective smarts

Diverse Minds, Divided Networks? Personality Composition, Polarization, and Collective Intelligence in LLM-Based Social Simulations

Abstract: Simulated societies of large language model agents are used to study online polarization, and separately to study collective intelligence, but the two are rarely measured in the same system. It is therefore difficult to say whether a society's personality composition shapes both, or whether reducing polarization costs collective competence. We present TraitMix, an experimental design in which the Big Five composition of a simulated social network, both trait levels and trait heterogeneity, is a controlled experimental variable, and in which polarization and collective performance are measured in the same runs. Across 991 simulations of hundred-agent societies, spanning six contested topics and six language models, trait heterogeneity has the largest measured effects, acting in opposite directions on two faces of polarization: varied societies hold more dispersed opinions while being less segregated into camps, so homogeneous societies are not moderate but consensual echo chambers. Trait effects are not additive, as Agreeableness determines the sign of Openness, an interaction that replicates across models although the primary model's estimate is influence-driven. Contrary to the trade-off the study was designed to measure, no polarization measure predicts poorer collective performance, and cross-cutting interaction is the only one of four whose association with collective accuracy survives partialling on the aggregation identity. We report ablations removing two potential measurement circularities, an induction gate applied to every model, and the measures that failed them.

Fri 11 SeptComputation and Language
The gist
This paper looks at how the personalities of people in online groups affect how much they disagree and how well they work together. The authors used computer simulations with language model agents that mimic human traits like openness and agreeableness. They found that groups with more diverse personalities have opinions spread out but less likely to split into hostile sides, while similar groups tend to form echo chambers. Importantly, having less polarization did not mean the group made worse decisions, challenging the idea that reducing conflict harms group intelligence.
Open → 2609.12444v1

Efficient algorithm finds fair and nearly optimal voting committees

Finding Representative and Approximately Efficient Committees

Abstract: In approval-based committee voting, proportional approval voting (PAV) is a well-studied rule that combines proportional representation with Pareto efficiency. However, computing a PAV committee is NP-hard, raising a natural question: Can the proportionality and efficiency properties of PAV be achieved via computationally efficient procedures? We make two contributions toward answering this question. First, building on the known proportionality guarantees of the local-search-based variant of PAV (or local PAV), we systematically study its efficiency properties. We show that local PAV committees are weakly Pareto optimal, meaning that no other committee is strictly preferred by every voter. We also identify limitations: Local PAV guarantees only a $2$-approximation to fractional Pareto optimality ($2$-fPO) and a $2/3$-approximation to the optimal PAV score, and both bounds are tight. In contrast, global PAV is Pareto optimal and satisfies the stronger $α^\star$-fPO guarantee, where $α^\star \approx 1.346$ is the unique solution of $\int_0^{α^\star} \frac{1-e^{-y}}{y} \, dy = 1$, and this approximation is tight. Second, we design a polynomial-time algorithm that combines the best of these guarantees. The committee returned by our algorithm satisfies EJR$+$ (a proportionality guarantee), $α^\star$-fPO, and weak Pareto optimality. It also achieves a $0.79$-approximation to the optimal PAV score, matching the best possible polynomial-time approximation assuming $P \neq NP$. Our algorithm works by pipage rounding a concave relaxation of the PAV objective and using that committee to initialize local PAV, thereby combining global approximation guarantees with local search stability.

Mon 7 SeptComputer Science and Game Theory
The gist
Choosing a committee that fairly represents voters while being efficient is a tough computational problem. The authors study a popular voting rule called proportional approval voting (PAV) and analyze a local search version (local PAV) which is easier to compute. They show local PAV makes committees that are reasonably efficient but not perfectly so. Then, they present a new fast algorithm that combines good fairness with a nearly optimal efficiency guarantee, improving on previous methods and running in polynomial time.
Open → 2609.07554v1