Disengagement from social media platforms encompasses various behaviours, such as reducing activity or abandoning them altogether. At the far end of this spectrum, permanently deleting user accounts holds substantial implications for individuals, the people they contact, and the platform itself. However, the motivations underlying this choice, which involves erasing all content and social connections rather than simply abandoning the platform, remain unclear. To address this research gap, we analysed 4.5 million active users on Tumblr, a microblogging service. We assessed the likelihood of users deleting their accounts using binary logistic regression in general and in specific clusters of users. Our findings reveal that users are more inclined to delete their accounts if they have previously removed content from their profiles, highlighting the role of account deletion in cleaning and controlling personal information. Additionally, we observed robust peer correlations, indicating that users are more likely to delete their accounts if those they follow have already taken this step. We discuss these findings in the context of non-use theories and propose the development of checking-out processes that can enhance the user experience associated with account deletion.

Differential Privacy (DP) has emerged as the standard for privacy-preserving analysis of individual-level data. Despite growing attention in the research community to the operationalization of DP through the selection of the privacy budget ɛ , little is known about how DP obfuscation affects users’ disclosure decisions in data market scenarios. These decisions may be context-specific and vary with privacy preferences, eliciting disparate data valuations across individuals. Through a choice-based conjoint analysis ( N 1 = 588 ), simulating realistic data markets, we analyzed how varying DP protection levels influence individual decision-making of participation in data collection under DP. Our findings show that personal reward and the guaranteed DP protection had the strongest influence on participants’ selection of a data collection scenario. Surprisingly, the type of disclosed data had the least influence on participants’ decisions to disclose personal data, a trend consistent across participants from different countries. Furthermore, increasing the DP protection level by a single unit reduced the preferred compensation price by over 60% for the same level of user utility, with marginal effects diminishing exponentially at higher DP levels. Our results were then confirmed in an online study ( N 2 = 146 ) involving real data disclosure with actual payments, using our original scenario framing. Our findings can support context-specific DP configuration and help data practitioners improve decision-making associated with privacy protection in differentially private systems, balancing the trade-off between DP and compensation costs.

An increasing number of technologies depend on the large-scale collection of individual-level data, whether for gathering statistical insights from billions of users or training AI models. However, reliance on personal data raises privacy concerns that, in turn, limit the collection and analysis essential to these technologies. Differential Privacy (DP) has gained traction in both academia and industry, ensuring privacy by adding carefully crafted noise to data or its outputs based on a pre-defined DP parameter$\varepsilon$. As real-world implementations emerge, we can examine how DP is practically used beyond academic settings, supporting industry adoption and expanding knowledge on DP applications. Using a systematic process, we comprehensively surveyed the deployed parameters of DP configurations in both commercial and governmental implementations ($n=140$) and compared them to those employed in academic research. We also propose a high-level taxonomy for DP configuration, capturing practical implementations of differentially private Machine Learning (ML) and Federated Learning (FL) applications, highlighting key factors, including the privacy unit and$\varepsilon$. Our results show that, on average,$\varepsilon$values utilized in the industry span a wider range than those in academic research, with distinct configuration policies for governmental and commercial organizations. Moreover, we identified contrasting reasoning behind$\varepsilon$selection across deployment environments, alongside insufficient transparency in industry disclosures of DP parameters and limited support for user-oriented configuration. Finally, we discuss how the collected knowledge can be used to create methodological guidelines for the configuration of DP in real-world environments, supporting the vision of an Epsilon Registry.

Product managers are central figures in digital product development, coordinating teams and prioritizing features. Despite their influence, little research explores how their decisions affect user experience, especially in integrating social values into product architecture. Employing a mixed-methods framework, we conducted semi-structured interviews with 20 product managers and an online survey with an additional 81, all based in Israel. Our study identifies four unique strategies product managers utilize to balance business goals, user satisfaction, and ethical considerations. The survey data further substantiates the prevalence of these strategies across diverse sectors, confirming they reflect industry-wide approaches in the Israeli tech sector rather than isolated practices. To conclude, we emphasize how “soft resistance” tactics, such as adjusting data interpretations based on personal values, impact digital product designs. Moreover, our findings highlight that maintaining an ethical reputation in the job market can be pivotal in shaping product design.

The size of Android applications is getting larger to fulfill the requirements of various users. However, not all the features of the applications are needed and desired by a specific user. The unnecessary and non-desired features can increase the attack surface and consume system resources such as storage and memory. To address this issue, we propose a framework, MiniMon, to debloat unnecessary features from an Android app based on the logs of specific users' interactions with the app.

Human attention has become a critical resource for the effective design of smart services in which control may move back and forth between humans and computers. To avoid errors in critical conditions when the mental load is high, computer systems need to manage ongoing interruptions. In particular, the effect of interruptions can be mitigated with previews of computer-generated notifications. While previews have been used to increase engagement, research on their potential to mitigate the effect of interruptions is scarce. Using an experiment based on a game environment with varying task loads, we investigated the effect of previews on mitigating interruptions at several levels of mental load. We found interruptions that displayed previews added less to participants’ mental load but did not improve their overall performance. These results were consistent in all levels of task load. We summarize the article by discussing how previews can be designed to minimize the negative effects of interruptions.

Differential Privacy (DP) is often presented as a strong privacy-enhancing technology with broad applicability and advocated as a de-facto standard for releasing aggregate statistics on sensitive data. However, in many embodiments, DP introduces a new attack surface: a malicious entity entrusted with releasing statistics could manipulate the results and use the randomness of DP as a convenient smokescreen to mask its nefariousness. Since revealing the random noise would obviate the purpose of introducing it, the miscreant may have a perfect alibi. To close this loophole, we introduce the idea of \textit{Verifiable Differential Privacy}, which requires the publishing entity to output a zero-knowledge proof that convinces an efficient verifier that the output is both DP and reliable. Such a definition might seem unachievable, as a verifier must validate that DP randomness was generated faithfully without learning anything about the randomness itself. We resolve this paradox by carefully mixing private and public randomness to compute verifiable DP counting queries with theoretical guarantees and show that it is also practical for real-world deployment. We also demonstrate that computational assumptions are necessary by showing a separation between information-theoretic DP and computational DP under our definition of verifiability.

People regularly rely on social support from family, friends, and the public when mitigating security and privacy risks, even if mainstream technologies hardly support these interactions. In this paper, we evaluated Meerkat, a mobile application that allows users to receive support through screenshot capturing, marking, and messaging. In a field experiment (n = 65), we tested how Meerkat helps users face phishing attempts and examined it by receiving help from close social connections and community volunteers. Our findings show that while users could learn from both types of helpers, they were significantly more willing to rely on advice from close connections. We evaluate several criteria for successful support interactions, showing that learning is significantly correlated with specific properties of the support interaction, such as the length of the messages. We conclude the paper by discussing how our findings can be used to design community-based applications.

Android applications are getting bigger with an increasing number of features. However, not all the features are needed by a specific user. The unnecessary features can increase the attack surface and cost additional resources (e.g., storage and memory). Therefore, it is important to remove unnecessary features from Android applications. However, it is difficult for the end users to fully explore the apps to identify the unnecessary features, and there is no off-the-shelf tool available to assist users to debloat the apps by themselves. In this work, we propose AutoDebloater to debloat Android applications automatically for end users. AutoDebloater is a web application that can be accessed by end-users through a web browser. In particular, AutoDebloater can automatically explore an app and identify the transitions between activities. Then, AutoDebloater will present the Activity Transition Graph to users and ask them to select the activities they do not want to keep. Finally, AutoDebloater will remove the activities that are selected by users from the app. We conducted a user study on five Android apps downloaded from three categories (i.e., Finance, Tools, and Navigation) in Google Play and F-Droid. The results show that users are satisfied with AutoDebloater in terms of the stability of the debloated apps and the ability of AutoDebloater to identify features that are never noticed before. The tool is available at http://autodebloater.club. The code is available at https://github.com/jiakun-liu/autodebloater/ and the demonstration video can be found at https://youtu.be/Gmz0-p2n9D4.

During the COVID-19 pandemic, many countries have developed and deployed contact tracing technologies to curb the spread of the disease by locating and isolating people who have been in contact with coronavirus carriers. Subsequently, understanding why people install and use contact tracing apps is becoming central to their effectiveness and impact. This paper analyzes situations where centralized mass surveillance technologies are deployed simultaneously with a voluntary contact tracing mobile app. We use this parallel deployment as a natural experiment that tests how attitudes toward mass deployments affect people's installation of the contact tracing app. Based on a representative survey of Israelis (n=519), our findings show that positive attitudes toward mass surveillance were related to a reduced likelihood of installing contact tracing apps and an increased likelihood of uninstalling them. These results also hold when controlling for privacy concerns about the contact tracing app, attitudes toward the app, trust in authorities, and demographic properties. Similar reasoning may also be relevant for crowding out voluntary participation in data collection systems.

People’s location data are continuously tracked from various devices and sensors, enabling an ongoing analysis of sensitive information that can violate people’s privacy and reveal confidential information. Synthetic data have been used to generate representative location sequences yet to maintain the users’ privacy. Nonetheless, the privacy-accuracy tradeoff between these two measures has not been addressed systematically. In this article, we analyze the use of different synthetic data generation models for long location sequences, including extended short-term memory networks (LSTMs), Markov Chains (MC), and variable-order Markov models (VMMs). We employ different performance measures, such as data similarity and privacy, and discuss the inherent tradeoff. Furthermore, we introduce other measurements to quantify each of these measures. Based on the anonymous data of 300 thousand cellular-phone users, our work offers a road map for developing policies for synthetic data generation processes. We propose a framework for building data generation models and evaluating their effectiveness regarding those accuracy and privacy measures.

Older adults face increased safety challenges, such as targeted online fraud and phishing, contributing to the growing technological divide between them and younger adults. Social support from family and friends is often the primary way older adults receive help, but it may also lead to reliance on others. We have conducted an exploratory study to investigate older adults' attitudes and experiences related to mobile social support technologies for mobile safety. We interviewed 18 older adults about their existing support experiences and used the think-aloud method to gather data about a prototype for providing social support during mobile safety challenges. Our findings point to the potential of mobile technology to increase older adults' ability to mitigate mobile safety challenges through active learning from close social connections. We discuss how to support technology can address helpers' intolerance and overcome the challenges of physical distance.

Installing security applications is a common way to protect against malicious apps, phishing emails, and other threats in mobile operating systems. While these applications can provide essential security protections, they also tend to access large amounts of people's sensitive information. Therefore, individuals need to evaluate the trade-off between the security features and the privacy invasion when deciding on which protection mechanisms to use. In this paper, we examine factors affecting the willingness to install mobile security applications by taking into account the invasion levels and security features of cyber-security applications. To this end, we propose a visual language that depicts the coverage of different security features as well as privacy intrusiveness levels. Our user study (n=300) shows that users assessing security applications find their trade-off balance in highly secure apps with a medium level of privacy invasion. The results indicate that a low privacy invasion might signal that the security application provides less security. We discuss these findings in the context of understanding the trade-off between privacy and security.

Contemporary mobile devices continuously interrupt people with notifications in various and changing physical environments. As different places can have different social setting, understanding how disturbing an interruption might be to people around the user is not a straightforward task. To understand how users perceive disturbance in their social environment, we analyze the results of a 3-week user study with 50 participants using the experience sampling method and log analysis. We show that perceptions of disturbance are strongly related to the social norms surrounding the place, such as whether the place is considered private or public, even when controlling for the number of people around the user. Furthermore, users' perceptions of disturbance are also related to the activity carried out on the phone, and the subjective perceptions of isolation from other people in the space. We conclude the paper by discussing how our findings can be used to design new mobile devices that are aware of the social norms and their users' environmental context.

Privacy by design (PbD) is a policy measure that calls for embedding privacy into the design of technologies at early stages of the development process and throughout its lifecycle. By introducing privacy considerations into the technological design, PbD delegates responsibility over privacy to those in charge of the design, namely software developers who design information technologies (hereafter called developers). Thus, for PbD to be a viable option, it is important to understand developers' perceptions, interpretation and practices as to informational privacy.

Privacy in Online Social Networks (OSNs) is a dynamic concept, contingent on changes in technology and usage norms. Social influence is a major avenue for adopting online behaviors in general and privacy practices in particular. In this study, we examine how the source of influence affects the perceived behavioral intention to adopt privacy behavior. Our findings are based on a randomized experiment (167 U.S.-based Amazon Mechanical Turk workers) using a custom Facebook application that collects feedback from participants regarding their intention to adopt privacy practices from different types of sources, including authoritative organizations and friends with varying tie strength correlative. Our results show that the source of social influence affects the susceptibility to adopt certain privacy behaviors and that there are different patterns of influence for security and privacy norms. More interestingly, susceptibility is modulated by the privacy perceptions of the user: users with high perceived behavioral control are more susceptible to peer influence. Additionally, we show that the intention to adopt privacy practices is correlated with the intention to further influence other people.

Online social networks (OSNs) make information accessible for unlimited periods and provide easy access to past information by arranging information in time lines or by providing sophisticated search mechanisms. Despite increased concerns over the privacy threat that is posed by digital memory, there is little knowledge about retrospective privacy: the extent to which the age of the exposed information affects sharing preferences. In this article, we investigate how information aging impacts users’ sharing preferences on Facebook. Our findings are based on a between-subjects experiment (n = 272), in which we measured the impact of time since first publishing an OSN post on its sharing preferences. Our results quantify how willingness to share is lower for older Facebook posts and show that older posts have lower relevancy to the user’s social network and are less representative of the user’s identity. We show that changes in the user’s social circles, the occurrence of significant life changes and a user’s young age are correlated with a further decrease in the willingness to keep sharing past information. We discuss our findings by juxtaposing digital memory theories and privacy theories and suggest a vision for mechanisms that can help users manage longitudinal privacy.

The way users manage access to their information and computers has a tremendous effect on the overall security and privacy of individuals and organizations. Usually, access management is conducted using a choice architecture , a behavioral economics concept that describes the way decisions are framed to users. Studies have consistently shown that the design of choice architectures, mainly the selection of default options, has a strong effect on the final decisions users make by nudging them toward certain behaviors. In this article, we propose a method for optimizing access control choice architectures in online social networks. We empirically evaluate the methodology on Facebook, the world's largest online social network, by measuring how well the default options cover the existing user choices and preferences and toward which outcome the choice architecture nudges users. The evaluation includes two parts: (a) collecting access control decisions made by 266 users of Facebook for a period of 3 months; and (b) surveying 533 participants who were asked to express their preferences regarding default options. We demonstrate how optimal defaults can be algorithmically identified from users’ decisions and preferences, and we measure how existing defaults address users’ preferences compared with the optimal ones. We analyze how access control defaults can better serve existing users, and we discuss how our method can be used to establish a common measuring tool when examining the effects of default options.

Cloud computing offers a compelling vision of computation, enabling an unprecedented level of data distribution and sharing. Beyond improving the computing infrastructure, cloud computing enables a higher level of interoperability between information systems, simplifying tasks such as sharing documents between coworkers or enabling collaboration between an organization and its suppliers. While these abilities may result in significant benefits to users and organizations, they also present privacy challenges due to unwanted exposure of sensitive information. As information-sharing processes in cloud computing are complex and domain specific, configuring these processes can be an overwhelming and burdensome task for users. This article investigates the feasibility of configuring sharing processes through a small and representative set of canonical configuration options. For this purpose, we present a generic method, named SCON-UP (Simplified CON-figuration of User Preferences). SCON-UP simplifies configuration interfaces by using a clustering algorithm that analyzes a massive set of sharing preferences and condenses them into a small number of discrete disclosure levels. Thus, the user is provided with a usable configuration model while guaranteeing adequate privacy control. We describe the algorithm and empirically evaluate our model using data collected in two user studies (n = 121 and n = 352). Our results show that when provided with three canonical configuration options, on average, 82% of the population can be covered by at least one option. We exemplify the feasibility of discretizing sharing levels and discuss the tradeoff between coverage and simplicity in discrete configuration options.

People-Nearby applications are becoming a popular way for individuals to search for new social relations in their physical vicinity. This paper presents the results of a qualitative study, based on 25 interviews, examining how privacy and locality are managed in these applications. We describe how location is used as a grounding mechanism, providing a platform for honest and truthful signals in the challenging process of forming new social relations. We discuss our findings by suggesting theoretical frameworks that can be used to analyze the social space induced by the applications, as well as to inform the design of new technologies that foster the creation of new social ties.

Online social networks provide access to the user's information for long periods of time after the information's initial publication. In this paper, we investigate the relation between information aging and its sharing preferences on Facebook. Our findings are based on a survey of 193 Facebook users, in which we asked users to specify their sharing preferences and intentions towards posts that were published in different periods of time (from the time of the survey and up to 24 months prior to the time of the survey.) Our results show that willingness to share significantly drops with the time passed since publishing the post. The occurrence of life changes, such as graduating from college or moving to a new town, is correlated with a further decrease in the willingness to share. We discuss our findings by relating it to information aging theories and privacy theories. Finally, we use our results to reflect on privacy mechanisms for long-term usage of online social networks, such as expiry date for content and historical information reviewing processes.

'People-nearby' applications for meeting new people online are some of the most popular examples of systems that lead people from an online interaction to an offline interaction. This paper provides a critical review of available applications, and identifies three key properties that are essential for the applications: physical location, identity management, and trust. The paper suggests open research questions that can explain the success of these applications and guide the design of new technologies that encourage offline interactions.

This paper presents a study that aims to answer two important questions related to targeted location-sharing privacy attacks: (1) given a group of users and their social graph, is it possible to predict which among them is likely to reveal most about their whereabouts, and (2) given a user, is it possible to predict which among her friends knows most about her whereabouts. To answer these questions we analyse the privacy policies of users of a real-time location sharing application, in which users actively shared their location with their contacts. The results show that users who are central to their network are more likely to reveal most about their whereabouts. Furthermore, we show that the friend most likely to know the whereabouts of a specific individual is the one with most common contacts and/or greatest number of contacts.

We present the design, implementation, and evaluation of Caché, a system that offers location privacy for certain classes of location-based applications. The core idea in Caché is to periodically pre-fetch potentially useful location-enhanced content well in advance. Applications then retrieve content from a local cache on the mobile device when it is needed. This approach allows an end-user to make use of location-enhanced content while only revealing to third-party content providers a large geographic region rather than a precise location. In this paper, we present an analysis that examines tradeoffs in terms of storage, bandwidth, and freshness of data. We then discuss the design and implementation of an Android service embodying these ideas. Finally, we provide two evaluations of Caché. One measures the performance of our approach with respect to privacy and mobile content availability using real-world mobility traces. The other focuses on our experiences using Caché to enhance user privacy in three open source Android applications.

This paper examines the location traces of 489 users of a location sharing social network for relationships between the users' mobility patterns and structural properties of their underlying social network. We introduce a novel set of location-based features for analyzing the social context of a geographic region, including location entropy, which measures the diversity of unique visitors of a location. Using these features, we provide a model for predicting friendship between two users by analyzing their location trails. Our model achieves significant gains over simpler models based only on direct properties of the co-location histories, such as the number of co-locations. We also show a positive relationship between the entropy of the locations the user visits and the number of social ties that user has in the network. We discuss how the offline mobility of users can have implications for both researchers and designers of online social networks.

The rapid adoption of location tracking and mobile social networking technologies raises significant privacy challenges. Today our understanding of people's location sharing privacy preferences remains very limited, including how these preferences are impacted by the type of location tracking device or the nature of the locations visited. To address this gap, we deployed Locaccino, a mobile location sharing system, in a four week long field study, where we examined the behavior of study participants (n=28) who shared their location with their acquaintances (n=373.) Our results show that users appear more comfortable sharing their presence at locations visited by a large and diverse set of people. Our study also indicates that people who visit a wider number of places tend to also be the subject of a greater number of requests for their locations. Over time these same people tend to also evolve more sophisticated privacy preferences, reflected by an increase in time- and location-based restrictions. We conclude by discussing the implications our findings.

No abstract available.

In this work, we propose a two-step, context-based semantic approach to the problem of matching and ranking Web services for possible service composition. We present an analysis of different methods for classifying Web services for possible composition and supply a context-based semantic matching method for ranking these possibilities. Semantic understanding of Web services may provide added value by identifying new possibilities for compositions of services. The semantic matching ranking approach is unique since it provides the Web service designer with an explicit numeric estimation of the extent to which a possible composition ldquomakes sense.rdquo First, we analyze two common methods for text processing, TF/IDF and context analysis; and two types of service description, free text and WSDL. Second, we present a method for evaluating the proximity of services for possible compositions. Each Web service WSDL context descriptor is evaluated according to its proximity to other services' free text context descriptors. The methods were tested on a large repository of real-world Web services. The experimental results indicate that context analysis is more useful than TF/IDF. Furthermore, the method evaluating the proximity of the WSDL description to the textual description of other services provides high recall and precision results.

Web service discovery is one of the main applications of semantic Web services, which extend standard Web services with semantic annotations. Current discovery solutions were developed in the context of automatic service composition. Thus, the “client” of the discovery procedure is an automated computer program rather than a human, with little, if any, tolerance to inexact results. However, in the real world, services which might be semantically distanced from each other are glued together using manual coding. In this article, we propose a new retrieval model for semantic Web services, with the objective of simplifying service discovery for human users. The model relies on simple and extensible keyword-based query language and enables efficient retrieval of approximate results, including approximate service compositions. Since representing all possible compositions and all approximate concept references can result in an exponentially-sized index, we investigate clustering methods to provide a scalable mechanism for service indexing. Results of experiments, designed to evaluate our indexing and query methods, show that satisfactory approximate search is feasible with efficient processing time.