Decision aids based on artificial intelligence (AI) induce a wide range of outcomes when they are deployed in uncertain environments. In this paper, we investigate how users' trust in recommendations from an AI decision aid is impacted over time by designs that expose uncertainty in predicted outcomes. Unlike previous work, we focus on gig driving - a real-world, repeated decision-making context. We report on a longitudinal mixed-methods study (n=51) where we measured gig drivers' trust as they interacted with an AI-based schedule recommendation tool. Our results show that participants' trust in the tool was shaped by both their first impressions of its accuracy and their longitudinal interactions with it; and that task-aligned framings of uncertainty improved trust by allowing participants to incorporate uncertainty into their decision-making processes. Additionally, we observed that trust depended on their characteristics as drivers, underscoring the need for more in situ studies of AI decision aids.

Privacy policies are often lengthy and complex legal documents, and are difficult for many people to read and comprehend. Recent research efforts have explored automated assistants that process the language in policies and answer people's privacy questions. This study documents the importance of two different types of reasoning necessary to generate accurate answers to people's privacy questions. The first is the need to support taxonomic reasoning about related terms commonly found in privacy policies. The second is the need to reason about regulatory disclosure requirements, given the prevalence of silence in privacy policy texts. Specifically, we report on a study involving the collection of 749 sets of expert annotations to answer privacy questions in the context of 210 different policy/question pairs. The study highlights the importance of taxonomic reasoning and of reasoning about regulatory disclosure requirements when it comes to accurately answering everyday privacy questions. Next we explore to what extent current generative AI tools are able to reliably handle this type of reasoning. Our results suggest that in their current form and in the absence of additional help, current models cannot reliably support the type of reasoning about regulatory disclosure requirements necessary to accurately answer privacy questions. We proceed to introduce and evaluate different approaches to improving their performance. Through this work, we aim to provide a richer understanding of the capabilities automated systems need to have to provide accurate answers to everyday privacy questions and, in the process, outline paths for adapting AI models for this purpose.

The human brain’s remarkable efficiency in solving puzzles through pictorial information processing serves as a valuable inspiration for computational puzzle solving. In this study, we present a nucleation algorithm for automated puzzle solving, developed based on statistical analysis of an empirical database. This algorithm effectively solves puzzles by choosing pieces with infrequent and iridescent edges as nucleation centers, followed by the identification of neighboring pieces with high resemblances from the remaining puzzle pieces. For the 8 different pictures examined in this study, both empirical data and computer simulations consistently demonstrate a power-law relationship between solving time and the number of puzzle pieces, with an exponent less than 2. We explain this relationship through the nucleation model and explore how the exponent is influenced by the color pattern of the puzzle picture. Moreover, our investigation of puzzle-solving processes reveals distinct principal pathways, akin to protein folding behavior. Our study contributes to the development of a cognitive model for human puzzle solving and color pattern recognition.

We introduce Monte Carlo Forest Search (MCFS), a class of reinforcement learning (RL) algorithms for learning policies in {tree MDPs}, for which policy execution involves traversing an exponential-sized tree. Examples of such problems include proving unsatisfiability of a SAT formula; counting the number of solutions of a satisfiable SAT formula; and finding the optimal solution to a mixed-integer program. MCFS algorithms can be seen as extensions of Monte Carlo Tree Search (MCTS) to cases where, rather than finding a good path (solution) within a tree, the problem is to find a small tree within a forest of candidate trees. We instantiate and evaluate our ideas in an algorithm that we dub Knuth Synthesis, an MCFS algorithm that learns DPLL branching policies for solving the Boolean satisfiability (SAT) problem, with the objective of achieving good average-case performance on a given distribution of unsatisfiable problem instances. Knuth Synthesis is the first RL approach to avoid the prohibitive costs of policy evaluations in an exponentially-sized tree, leveraging two key ideas: first, we estimate tree size by randomly sampling paths and measuring their lengths, drawing on an unbiased approximation due to Knuth (1975); second, we query a strong solver at a user-defined depth rather than learning a policy across the whole tree, to focus our policy search on early decisions that offer the greatest potential for reducing tree size. We matched or exceeded the performance of a strong baseline on three well-known SAT distributions, facing problems that were two orders of magnitude more challenging than those addressed in previous RL studies.

Understanding and managing data privacy in the digital world can be challenging for sighted users, let alone blind and low-vision (BLV) users. There is limited research on how BLV users, who have special accessibility needs, navigate data privacy, and how potential privacy tools could assist them. We conducted an in-depth qualitative study with 21 US BLV participants to understand their data privacy risk perception and mitigation, as well as their information behaviors related to data privacy. We also explored BLV users' attitudes towards potential privacy question answering (Q A) assistants that enable them to better navigate data privacy information. We found that BLV users face heightened security and privacy risks, but their risk mitigation is often insufficient. They do not necessarily seek data privacy information but clearly recognize the benefits of a potential privacy Q A assistant. They also expect privacy Q A assistants to possess cross-platform compatibility, support multi-modality, and demonstrate robust functionality. Our study sheds light on BLV users' expectations when it comes to usability, accessibility, trust and equity issues regarding digital data privacy.

Monte Carlo Tree Search (MCTS) is an effective approach for finding low-cost paths through any large combinatorial space that can naturally be structured as a search tree. However, some combinatorial problems do not have a natural interpretation as searches for a good path. For example, solving a CSP can be represented as a path (assign variables sequentially and check the solution); however, proving that no solution exists (via existing methods) requires enumerating multiple paths to build out a “proof tree” demonstrating that every possible variable assignment leads to a conflict. Rather than finding a good path (solution) within a tree, the search problem becomes searching for a small proof tree within a forest of candidate trees. In this paper we develop Monte Carlo Forest Search (MCFS), an algorithm for finding small search trees. Our method leverages the benefits of the best MCTS approaches and further introduces two key ideas. First, we estimate tree size via the linear (i.e., path-based) and unbiased approximation from Knuth (1975). Second, we query a strong solver at a user-defined depth rather than learning a policy across the whole tree, in order to (1) reduce the variance of our tree-size estimates and (2) focus our policy search on early decisions, which offer the greatest potential for reducing tree size. We evaluated our approach on the Boolean satisfiability (SAT) problem, and found that it matched or improved performance over a strong baseline on two well-known distributions (\texttt{sgen}, \texttt{random}). Notably, we improved walltime by 9\% on \texttt{sgen} over the \texttt{kcnfs} solver and even further over the strongest UNSAT solver from the 2021 SAT competition.

Traffic simulators are used to generate data for learning in intelligent transportation systems (ITSs). A key question is to what extent their modelling assumptions affect the capabilities of ITSs to adapt to various scenarios when deployed in the real world. This work focuses on two simulators commonly used to train reinforcement learning (RL) agents for traffic applications, CityFlow and SUMO. A controlled virtual experiment varying driver behavior and simulation scale finds evidence against distributional equivalence in RL-relevant measures from these simulators, with the root mean squared error and KL divergence being significantly greater than 0 for all assessed measures. While granular real-world validation generally remains infeasible, these findings suggest that traffic simulators are not a deus ex machina for RL training: understanding the impacts of inter-simulator differences is necessary to train and deploy RL-based ITSs.

Over the past decade, researchers have started to explore the use of NLP to develop tools aimed at helping the public, vendors, and regulators analyze disclosures made in privacy policies. With the introduction of new privacy regulations, the language of privacy policies is also evolving, and disclosures made by the same organization are not always the same in different languages, especially when used to communicate with users who fall under different jurisdictions. This work explores the use of language technologies to capture and analyze these differences at scale. We introduce an annotation scheme designed to capture the nuances of two new landmark privacy regulations, namely the EU's GDPR and California's CCPA/CPRA. We then introduce the first bilingual corpus of mobile app privacy policies consisting of 64 privacy policies in English (292K words) and 91 privacy policies in German (478K words), respectively with manual annotations for 8K and 19K fine-grained data practices. The annotations are used to develop computational methods that can automatically extract "disclosures" from privacy policies. Analysis of a subset of 59 "semi-parallel" policies reveals differences that can be attributed to different regulatory regimes, suggesting that systematic analysis of policies using automated language technologies is indeed a worthwhile endeavor.

MSClustering is an efficient software package for visualizing and analyzing complex networks in Cytoscape. Based on the distance matrix of a network that it takes as input, MSClustering automatically displays the minimum span clustering (MSC) of the network at various characteristic levels. To produce a view of the overall network structure, the app then organizes the multi-level results into an MSC tree. Here, we demonstrate the package's phylogenetic applications in studying the evolutionary relationships of complex systems, including 63 beta coronaviruses and 197 GPCRs. The validity of MSClustering for large systems has been verified by its clustering of 3481 enzymes. Through an experimental comparison, we show that MSClustering outperforms five different state-of-the-art methods in the efficiency and reliability of their clustering.

Abstract Browsing privacy tools can help people protect their digital privacy. However, tools which provide the strongest protections—such as Tor Browser—have struggled to achieve widespread adoption. This may be due to usability challenges, misconceptions, behavioral biases, or mere lack of awareness. In this study, we test the effectiveness of nudging interventions that encourage the adoption of Tor Browser. First, we test an informational nudge based on protection motivation theory (PMT), designed to raise awareness of Tor Browser and help participants form accurate perceptions of it. Next, we add an action planning implementation intention, designed to help participants identify opportunities for using Tor Browser. Finally, we add a coping planning implementation intention, designed to help participants overcome challenges to using Tor Browser, such as extreme website slowness. We test these nudges in a longitudinal field experiment with 537 participants. We find that our PMT-based intervention increased use of Tor Browser in both the short- and long-term. Our coping planning nudge also increased use of Tor Browser, but only in the week following our intervention. We did not find statistically significant evidence of our action planning nudge increasing use of Tor Browser. Our study contributes to a greater understanding of factors influencing the adoption of Tor Browser, and how nudges might be used to encourage the adoption of Tor Browser and similar privacy enhancing technologies.

Traffic signal control (TSC) is a high-stakes domain that is growing in importance as traffic volume grows globally. An increasing number of works are applying reinforcement learning (RL) to TSC; RL can draw on an abundance of traffic data to improve signalling efficiency. However, RL-based signal controllers have never been deployed. In this work, we provide the first review of challenges that must be addressed before RL can be deployed for TSC. We focus on four challenges involving (1) uncertainty in detection, (2) reliability of communications, (3) compliance and interpretability, and (4) heterogeneous road users. We show that the literature on RL-based TSC has made some progress towards addressing each challenge. However, more work should take a systems thinking approach that considers the impacts of other pipeline components on RL.

Vagueness and ambiguity in privacy policies threaten the ability of consumers to make informed choices about how businesses collect, use, and share their personal information. The California Consumer Privacy Act (CCPA) of 2018 was intended to provide Californian consumers with more control by mandating that businesses (1) clearly disclose their data practices and (2) provide choices for consumers to opt out of specific data practices. In this work, we explore to what extent CCPA's disclosure requirements, as implemented in actual privacy policies, can help consumers to answer questions about the data practices of businesses. First, we analyzed 95 privacy policies from popular websites; our findings showed that there is considerable variance in how businesses interpret CCPA's definitions. Then, our user survey of 364 Californian consumers showed that this variance affects the ability of users to understand the data practices of businesses. Our results suggest that CCPA's mandates for privacy disclosures, as currently implemented, have not yet yielded the level of clarity they were designed to deliver, due to both vagueness and ambiguity in CCPA itself as well as potential non-compliance by businesses in their privacy policies.

Vagueness and ambiguity in privacy policies threaten the ability of consumers to make informed choices about how businesses collect, use, and share their personal information. The California Consumer Privacy Act (CCPA) of 2018 was intended to provide Californian consumers with more control by mandating that businesses (1) clearly disclose their data practices and (2) provide choices for consumers to opt out of specific data practices. In this work, we explore to what extent CCPA's disclosure requirements, as implemented in actual privacy policies, can help consumers to answer questions about the data practices of businesses. First, we analyzed 95 privacy policies from popular websites; our findings showed that there is considerable variance in how businesses interpret CCPA's definitions. Then, our user survey of 364 Californian consumers showed that this variance affects the ability of users to understand the data practices of businesses. Our results suggest that CCPA's mandates for privacy disclosures, as currently implemented, have not yet yielded the level of clarity they were designed to deliver, due to both vagueness and ambiguity in CCPA itself as well as potential non-compliance by businesses in their privacy policies.

Strangely enough, it is possible to use machine learning models to predict the satisfiability status of hard SAT problems with accuracy considerably higher than random guessing. Existing methods have relied on extensive, manual feature engineering and computationally complex features (e.g., based on linear programming relaxations). We show for the first time that even better performance can be achieved by end-to-end learning methods — i.e., models that map directly from raw problem inputs to predictions and take only linear time to evaluate. Our work leverages deep network models which capture a key invariance exhibited by SAT problems: satisfiability status is unaffected by reordering variables and clauses. We showed that end-to-end learning with deep networks can outperform previous work on random 3-SAT problems at the solubility phase transition, where: (1) exactly 50% of problems are satisfiable; and (2) empirical runtimes of known solution methods scale exponentially with problem size (e.g., we achieved 84% prediction accuracy on 600-variable problems, which take hours to solve with state-of-the-art methods). We also showed that deep networks can generalize across problem sizes (e.g., a network trained only on 100-variable problems, which typically take about 10 ms to solve, achieved 81% accuracy on 600-variable problems).

We present an integrated approach to cluster and visualize character networks in fiction with the aid of computational and statistical methods. An unsupervised clustering algorithm, minimum span clustering (MSC), was applied to cluster fictional characters at various characteristic resolutions based on their activities in the novel. As a demonstration, we study the character network in Dream of the Red Chamber, the greatest novel in Chinese literature. The character network of the novel is found to exhibit properties of scale-free and small-world networks. Based on unsupervised cluster analyses, we construct and visualize the community structure of the network, and find a three-tiered structure of core, secondary, and peripheral characters. By treating the network as a weighted graph, we further analyze the centralities of characters to determine their importance in the network, and find that betweenness centrality, as a measure of characters’ control over the flow of the narrative, is differentiated from other centrality measures for Dream of the Red Chamber. We believe that these analytic methods provide beneficial tools for applications such as autonomous novel writing.

Automated methods for the analysis, modeling, and visualization of large‐scale scientometric data provide measures that enable the depiction of the state of world scientific development. We aimed to integrate minimum span clustering (MSC) and minimum spanning tree methods to cluster and visualize the global pattern of scientific publications (PSP) by analyzing aggregated Science Citation Index (SCI) data from 1994 to 2011. We hypothesized that PSP clustering is mainly affected by countries' geographic location, ethnicity, and level of economic development, as indicated in previous studies. Our results showed that the 100 countries with the highest rates of publications were decomposed into 12 PSP groups and that countries within a group tended to be geographically proximal, ethnically similar, or comparable in terms of economic status. Hubs and bridging nodes in each knowledge production group were identified. The performance of each group was evaluated across 16 knowledge domains based on their specialization, volume of publications, and relative impact. Awareness of the strengths and weaknesses of each group in various knowledge domains may have useful applications for examining scientific policies, adjusting the allocation of resources, and promoting international collaboration for future developments.