Privacy Settings Research Articles

Optimization of privacy-utility trade-offs under informational self-determination

The pervasiveness of Internet of Things results in vast volumes of personal data generated by smart devices of users (data producers) such as smart phones, wearables and other embedded sensors. It is a common requirement, especially for Big Data analytics systems, to transfer these large in scale and distributed data to centralized computational systems for analysis. Nevertheless, third parties that run and manage these systems (data consumers) do not always guarantee users’ privacy. Their primary interest is to improve utility that is usually a metric related to the performance, costs and the quality of service. There are several techniques that mask user-generated data to ensure privacy, e.g. differential privacy. Setting up a process for masking data, referred to in this paper as a ‘privacy setting’, decreases on the one hand the utility of data analytics, while, on the other hand, increases privacy. This paper studies parameterizations of privacy settings that regulate the trade-off between maximum utility, minimum privacy and minimum utility, maximum privacy, where utility refers to the accuracy in the estimations of aggregation functions. Privacy settings can be universally applied as system-wide parameterizations and policies (homogeneous data sharing). Nonetheless they can also be applied autonomously by each user or decided under the influence of (monetary) incentives (heterogeneous data sharing). This latter diversity in data sharing by informational self-determination plays a key role on the privacy-utility trajectories as shown in this paper both theoretically and empirically. A generic and novel computational framework is introduced for measuring privacy-utility trade-offs and their Pareto optimization. The framework computes a broad spectrum of such trade-offs that form privacy-utility trajectories under homogeneous and heterogeneous data sharing. The practical use of the framework is experimentally evaluated using real-world data from a Smart Grid pilot project in which energy consumers protect their privacy by regulating the quality of the shared power demand data, while utility companies make accurate estimations of the aggregate load in the network to manage the power grid. Over 20,000 differential privacy settings are applied to shape the computational trajectories that in turn provide a vast potential for data consumers and producers to participate in viable participatory data sharing systems.

Mobile App Tools for Identifying and Managing Mental Health Disorders in Primary Care.

Mental health apps are intriguing yet challenging tools for addressing barriers to treatment in primary care. In the current review, we seek to assist primary care professionals with evaluating and integrating mental health apps into practice. We briefly summarize two leading frameworks for evaluating mental health apps and conduct a systematic review of mental health apps across a variety of areas commonly encountered in primary care. Existing frameworks can guide professionals and patients through the process of identifying apps and evaluating dimensions such as privacy and security, credibility, and user experience. For specific apps, several problem areas appear to have relatively more scientific evaluation in the current app landscape, including PTSD, smoking, and alcohol use. Other areas such as eating disorders not only lack evaluation, but contain a significant subset of apps providing potentially harmful advice. Overall, individuals seeking mental health apps will likely encounter strengths such as symptom tracking and psychoeducational components, while encountering common weaknesses such as insufficient privacy settings and little integration of empirically-supported techniques. While mental health apps may have more promise than ever, significant barriers to finding functional, usable, effective apps remain for health professionals and patients alike.

Towards privacy preserving social recommendation under personalized privacy settings

Privacy leakage is an important issue for social relationships-based recommender systems (i.e., social recommendation). Existing privacy preserving social recommendation approaches usually allow the recommender to fully control users’ information. This may be problematic since the recommender itself may be untrusted, leading to serious privacy leakage. Besides, building social relationships requires sharing interests as well as other private information, which may lead to more privacy leakage. Although sometimes users are allowed to hide their sensitive private data using personalized privacy settings, the data being shared can still be abused by the adversaries to infer sensitive private information. Supporting social recommendation with least privacy leakage to untrusted recommender and other users (i.e., friends) is an important yet challenging problem. In this paper, we aim to achieve privacy-preserving social recommendation under personalized privacy settings. We propose PrivSR, a novel privacy-preserving social recommendation framework, in which user can model user feedbacks and social relationships privately. Meanwhile, by allocating different noise magnitudes to personalized sensitive and non-sensitive feedbacks, we can protect users’ privacy against untrusted recommender and friends. Theoretical analysis and experimental evaluation on real-world datasets demonstrate that our framework can protect users’ privacy while being able to retain effectiveness of the underlying recommender system.

PrivacyShield

Current in-situ privacy solution approaches are inadequate in protecting sensitive information. They either require extra configuration effort or lack the ability to configure user desired privacy settings. Based on in-depth discussions during a design workshop, we propose PrivacyShield, a mobile system for providing subtle just-in-time privacy provisioning. PrivacyShield leverages the screen I/O device (screen digitizer) of smartphones to recognize gesture commands, even when the phone's screen is turned off. Based on gesture command inputs, various privacy-protection policies can be configured on-the-fly. We develop a novel stroke-based approach to address the challenges in segmenting and recognizing gesture command inputs, which helps the system in achieving good usability and performance. PrivacyShield also provides developers with APIs to enable just-in-time privacy provisioning in their applications. We have implemented an energy efficient PrivacyShield prototype on the Android platform, including smartphones with and without a low-power co-processor. Evaluation results show that our gesture segmentation algorithm is fast enough for real-time performance (introducing less than 200ms processing latency) and accurate (achieving an accuracy of 95% for single-character gestures and 89% for even three-character gestures). We also build a non-touch-screen-based just-in-time privacy provisioning prototype called the wrist gesture method. We compare the performance of the two prototypes by doing a 6-week field study with 12 participants and show why a simplistic solution falls short in providing privacy configurations. We also report the participants' perceptions and reactions after the field study.

Probabilistic Properties of the Spatial Bloom Filters and Their Relevance to Cryptographic Protocols

The classical Bloom filter data structure is a crucial component of hundreds of cryptographic protocols. It has been used in privacy preservation and secure computation settings, often in conjunction with the (somewhat) homomorphic properties of ciphers such as Paillier's. In 2014, a new data structure extending and surpassing the capabilities of the classical Bloom filter has been proposed. The new primitive, called spatial Bloom filter (SBF) retains the hash-based membership-query design of the Bloom filter, but applies it to elements from multiple sets. Since its introduction, the SBF has been used in the design of cryptographic protocols for a number of domains, including location privacy and network security. However, due to the complex nature of this probabilistic data structure, its properties had not been fully understood. In this paper, we address this gap in knowledge and we fully explore the probabilistic properties of the SBF. In doing so, we define a number of metrics (such as emersion and safeness) useful in determining the parameters needed to achieve certain characteristics in a filter, including the false positive probability and inter-set error rate. This will in turn enable the design of more efficient cryptographic protocols based on the SBF, opening the way to their practical application in a number of security and privacy settings.

Personalized Privacy Assistants for the Internet of Things: Providing Users with Notice and Choice

As we interact with an increasingly diverse set of sensing technologies, it becomes difficult to keep up with the many different ways in which data about ourselves is collected and used. Study after study has shown that while people generally care about their privacy, they feel they have little awareness of-let alone control over-the collection and use of their data. This article summarizes ongoing research to develop and field privacy assistants designed to empower people to regain control over their privacy in the Internet of Things (IoT). Specifically, we focus on the infrastructure we have developed and deployed to support privacy assistants for the IoT. This infrastructure enables the assistants to discover IoT resources (sensors, apps, services, devices, and so on) in the vicinity of their users, and selectively inform users about the data practices associated with these resources. It also supports the discovery of user-configurable settings for IoT resources (opt in, opt out, data erasure, and so on) if there are any, enabling privacy assistants to help users configure their IoT experience in accordance with their privacy expectations. We also discuss how, using machine learning to build and refine models of users privacy expectations and preferences, we plan to developed personalized privacy assistants capable of selectively informing their users about the data practices they actually care about and of helping them configure associated privacy settings.

‘Little Islands of Empathy’: networked stories of gender diversity and multiple selves’

This article draws on a methodologically interesting case study called ‘Stories Beyond Gender’, in which a small group of trans* people collaborates in social media storytelling. Building on the possibilities manifest in other more explicitly personal-as-political genres like digital storytelling, I explore the potential of this facilitated workshop practice to establish meaningful connections across difference, forging affinities that may continue to flourish online. Furthermore, I offer some specific examples of the ways in which my own networked story-sharing online, in a zine and in an exhibition affirmed emergent complexity. I address the theme of this Special Issue by examining the ways in which social media, despite paradoxical fragmentation, can be used creatively to mobilise interest in public aspects of gender expression. However, sharing stories, especially those linked to stigmatised identities, whether online or off, is not without its complications. In the face of highly valued privacy, a lack of familiarity with ever-changing privacy settings or the affordances of specific platforms can pose an obstacle to online self-representation that stands in the way of visible civic engagement. While acknowledging that the trans-phobic consequences of online misadventures continue to be dire, I address the self-protective skills and sophisticated ways in which gender-diverse people curate emergent and past selves across intersecting social networks both on and offline. I argue that, at the intersections of post-digital and post-gender ways of being, we can observe emergent acceptance of multiple selves that are capable of being inconsistent without being incoherent. These representations exist in stark opposition to pop psychology’s premise of a singular authentic ‘inner truth’.

Net Benefits: Digital Inequities in Social Capital, Privacy Preservation, and Digital Parenting Practices of U.S. Social Media Users

Social media adoption has been shown to exhibit digital inequality: sociodemographic background has been highly correlated with usage rates. As social media use has also been shown to correlate with important benefits, chiefly social capital, low adoption rates can mean a portion of the population is not receiving these benefits. In this work, we examine the equity of social media benefits among users, introducing and validating the existence of two new social media benefits: learning of privacy-preserving behaviors and parental engagement in children's social media use; and explore the equity of their distribution among social media users. To draw generalizable conclusions, we use a probabilistic telephone survey (n=3,000), weighted to represent the responses of the U.S. population within 2.7%. We encouragingly find no difference in adoption of social media based on education and find that lower-income users are more likely to use social media. Yet, we find an inequality in benefits: older and less educated social media users report lower degrees of nearly all examined benefits. Further, we find preliminary suggestion of an inherited digital inequality: parents who use social media, especially those who are more educated and higher paid, are more likely both to help their children set up privacy settings and teach them safe posting behaviors.

Identifying Medication Management Smartphone App Features Suitable for Young Adults With Developmental Disabilities: Delphi Consensus Study.

BackgroundSmartphone apps can be a tool to facilitate independent medication management among persons with developmental disabilities. At present, multiple medication management apps exist in the market, but only 1 has been specifically designed for persons with developmental disabilities. Before initiating further app development targeting this population, input from stakeholders including persons with developmental disabilities, caregivers, and professionals regarding the most preferred features should be obtained.ObjectiveThe aim of this study was to identify medication management app features that are suitable to promote independence in the medication management process by young adults with developmental disabilities using a Delphi consensus method.MethodsA compilation of medication management app features was performed by searching the iTunes App Store, United States, in February 2016, using the following terms: adherence, medication, medication management, medication list, and medication reminder. After identifying features within the retrieved apps, a final list of 42 features grouped into 4 modules (medication list, medication reminder, medication administration record, and additional features) was included in a questionnaire for expert consensus rating. A total of 52 experts in developmental disabilities, including persons with developmental disabilities, caregivers, and professionals, were invited to participate in a 3-round Delphi technique. The purpose was to obtain consensus on features that are preferred and suitable to promote independence in the medication management process among persons with developmental disabilities. Consensus for the first, second, and third rounds was defined as ≥90%, ≥80%, and ≥75% agreement, respectively.ResultsA total of 75 responses were received over the 3 Delphi rounds—30 in the first round, 24 in the second round, and 21 in the third round. At the end of the third round, cumulative consensus was achieved for 60% (12/20) items in the medication list module, 100% (3/3) in the medication reminder module, 67% (2/3) in the medication administration record module, and 63% (10/16) in the additional features module. In addition to the medication list, medication reminder, and medication administration record features, experts selected the following top 3 most important additional features: automatic refills through pharmacies; ability to share medication information from the app with providers; and ability to share medication information from the app with family, friends, and caregivers. The top 3 least important features included a link to an official drug information source, privacy settings and password protection, and prescription refill reminders.ConclusionsAlthough several mobile apps for medication management exist, few are specifically designed to support persons with developmental disabilities in the complex medication management process. Of the 42 different features assessed, 64% (27/42) achieved consensus for inclusion in a future medication management app. This study provides information on the features of a medication management app that are most important to persons with developmental disabilities, caregivers, and professionals.

Factors influencing patients' satisfaction at different levels of health facilities in Bangladesh: Results from patient exit interviews.

There is a paucity in current literature about the level of patients’ satisfaction and factors influencing it in Bangladesh health system. We aimed to measure the level of patients’ satisfaction across different types and levels of healthcare facilities and to determine which factors influence this satisfaction level. A patient exit interview was carried out among 2207 patients attending selected health facilities in two administrative divisions of Bangladesh, namely Rajshahi and Sylhet. Information on healthcare experience and satisfaction with received care was collected through an electronic structured questionnaire. Information about ‘overall satisfaction with healthcare’ was collected on a 10-point scale and then dichotomized based on the median-split. Binomial logistic regressions, both simple and multivariable, were conducted to identify which factors contribute significantly to patients’ satisfaction. We found that 63.2% of the participants were satisfied with the healthcare service they received. Patients attending the private facilities had the highest level of satisfaction (i.e. 73%) and patients attending the primary care facilities had the lowest level of satisfaction (i.e. 52%). Factors like convenient opening hours, asking related questions to the providers, facility cleanliness and privacy settings were significantly associated with patients’ satisfaction. Being satisfied with facility cleanliness (multivariable OR 4.30; 95% CI: 3.29–5.62) and privacy settings (multivariable OR 1.68; 95% CI: 1.28–2.21) were the strongest predictors of patients’ satisfaction. In conclusion, a significant portion of the patients in Bangladesh are not satisfied with their received care. Patients’ satisfaction can be increased by focusing on improving facility cleanliness, privacy settings and providers’ interpersonal skills.

Safe sonet: a framework for building trustworthy relationships

The advent of easy to use services and the ability to bridge boundaries in space and time has increasingly changed our social lives and com-munication habits. Being unaware of their audience people on social networks inconsiderately share many personal items. Generally Social Networking Sites (SNS) users depend on the SNS platform for managing their social identities. Although SNS providers have introduced privacy settings to protect users against threats, these are insufficient and the access control models are difficult to be understood by novice users [1]. Unexpectedly, even those who are aware of the security and privacy threats and the preventive tools that combat those threats, lack the motivation to utilize security features to protect themselves. The paper discusses a frame-work (Safe SoNet) that aims to provide a plat-form for secure sharing of posts. The objective of this framework is twofold. Firstly, to analyze user behavior by retrieving user information from various social networking sites, addressing transliteration issues. Secondly, apply the user behavior to quantify the trust score of a connect, which would in turn form a decision making parameter to decide on the peer with whom the information can be securely shared.

Leveraging Content Sensitiveness and User Trustworthiness to Recommend Fine-Grained Privacy Settings for Social Image Sharing

To configure successful privacy settings for social image sharing, two issues are inseparable: 1) content sensitiveness of the images being shared; and 2) trustworthiness of the users being granted to see the images. This paper aims to consider these two inseparable issues simultaneously to recommend fine-grained privacy settings for social image sharing. For achieving more compact representation of image content sensitiveness (privacy), two approaches are developed: 1) a deep network is adapted to extract 1024-D discriminative deep features; and 2) a deep multiple instance learning algorithm is adopted to identify 280 privacy-sensitive object classes and events. Second, users on the social network are clustered into a set of representative social groups to generate a discriminative dictionary for user trustworthiness characterization. Finally, both the image content sensitiveness and the user trustworthiness are integrated to train a tree classifier to recommend fine-grained privacy settings for social image sharing. Our experimental studies have demonstrated both the efficiency and the effectiveness of our proposed algorithms.

Personalized Privacy-Preserving Social Recommendation

Privacy leakage is an important issue for social recommendation. Existing privacy preserving social recommendation approaches usually allow the recommender to fully control users' information. This may be problematic since the recommender itself may be untrusted, leading to serious privacy leakage. Besides, building social relationships requires sharing interests as well as other private information, which may lead to more privacy leakage. Although sometimes users are allowed to hide their sensitive private data using privacy settings, the data being shared can still be abused by the adversaries to infer sensitive private information. Supporting social recommendation with least privacy leakage to untrusted recommender and other users (i.e., friends) is an important yet challenging problem. In this paper, we aim to address the problem of achieving privacy-preserving social recommendation under personalized privacy settings. We propose PrivSR, a novel framework for privacy-preserving social recommendation, in which users can model ratings and social relationships privately. Meanwhile, by allocating different noise magnitudes to personalized sensitive and non-sensitive ratings, we can protect users' privacy against the untrusted recommender and friends. Theoretical analysis and experimental evaluation on real-world datasets demonstrate that our framework can protect users' privacy while being able to retain effectiveness of the underlying recommender system.

Identifying Private Content for Online Image Sharing

I present the outline of my dissertation work, Identifying Private Content for Online Image Sharing. Particularly, in my dissertation, I explore learning models to predict appropriate binary privacy settings (i.e., private, public) for images, before they are shared online. Specifically, I investigate textual features (user-annotated tags and automatically derived tags), and visual semantic features that are transferred from various layers of deep Convolutional Neural Network (CNN). Experimental results show that the learning models based on the proposed features outperform strong baseline models for this task on the Flickr dataset of thousands of images.

Is a Fitbit a Diary? Self-Tracking and Autobiography

Is a Fitbit a Diary? Self-Tracking and Autobiography

Why do college students prefer Facebook, Twitter, or Instagram? Site affordances, tensions between privacy and self-expression, and implications for social capital

Whereas the bulk of research on social media has taken a granular approach, targeting specific behaviors on one site, or to a lesser extent, multiple sites, the current study aimed to holistically examine the social media landscape, exploring questions about who is drawn to popular social media sites, why they prefer each site, and the social consequences of site preference. Survey data was collected from 663 college students regarding their use and preference for Facebook, Instagram, or Twitter. Results highlight the popularity of Instagram for college students, and women in particular. Personal characteristics such as gender, age, affordances on specific sites, and privacy concerns predicted social media preference. Expanding upon the privacy paradox, we found that participants who preferred Twitter were more likely to have a public (vs. private) profile, reported higher levels of self-disclosure, and indicated more bridging social capital. Participants who preferred Facebook reported lower levels of self-disclosure, but higher levels of bonding social capital compared to those who preferred Instagram. These findings suggest that associations between privacy settings, disclosure, and social capital vary as a function of both user motivations and the affordances of specific social media sites.

Twitter Users’ Privacy Concerns: What do Their Accounts’ First Names Tell Us?

Abstract Purpose In this paper, we describe how gender recognition on Twitter can be used as an intelligent business tool to determine the privacy concerns among users, and ultimately offer a more personalized service for customers who are more likely to respond positively to targeted advertisements. Design/methodology/approach We worked with two different data sets to examine whether Twitter users’ gender, inferred from the first name of the account and the profile description, correlates with the privacy setting of the account. We also used a set of features including the inferred gender of Twitter users to develop classifiers that predict user privacy settings. Findings We found that the inferred gender of Twitter users correlates with the account’s privacy setting. Specifically, females tend to be more privacy concerned than males. Users whose gender cannot be inferred from their provided first names tend to be more privacy concerned. In addition, our classification performance suggests that inferred gender can be used as an indicator of the user’s privacy preference. Research limitations It is known that not all twitter accounts are real user accounts, and social bots tweet as well. A major limitation of our study is the lack of consideration of social bots in the data. In our study, this implies that at least some percentage of the undefined accounts, that is, accounts that had names non-existent in the name dictionary, are social bots. It will be interesting to explore the privacy setting of social bots in the Twitter space. Practical implications Companies are investing large amounts of money in business intelligence tools that allow them to know the preferences of their consumers. Due to the large number of consumers around the world, it is very difficult for companies to have direct communication with each customer to anticipate market changes. For this reason, the social network Twitter has gained relevance as one ideal tool for information extraction. On the other hand, users’ privacy preference needs to be considered when companies consider leveraging their publicly available data. This paper suggests that gender recognition of Twitter users, based on Twitter users’ provided first names and their profile descriptions, can be used to infer the users’ privacy preference. Originality/value This study explored a new way of inferring Twitter user’s gender, that is, to recognize the user’s gender based on the provided first name and the user’s profile description. The potential of this information for predicting the user’s privacy preference is explored.

Semantics-based sensitive topic diffusion detection framework towards privacy aware online social networks

The advent of sharing sensitive information via Online Social Networks (OSN) has jeopardized the user to the extent that the privacy of millions of OSN users could well be compromised, with their data openly available in the public domain. Evidently, users lack in data privacy and the access control mechanisms available to avoid the risk of disclosure. Therefore a framework that automatically preserves the user privacy to detect sensitive topic and minimize the risk of sensitive information disclosure risk beyond the current privacy sceneries offered by OSN service providers is required. In this paper, we present a three-fold sanitization framework which precisely detects sensitive topics semantically using statistical topic model scheme which incorporates standard knowledge bases for tagging the sensitive topics discovered. The interaction documents from location-of-interest are subjected to SSAR–LDA using Gibbs Sampling to identify sensitive topic clusters with high location entropy. The experimental result shows, (i) the sensitive topic clusters are identified with very high accuracy, (ii) despite the redaction approach, which eliminate the sensitive term, our proposed scheme enhance the privacy preserving policy by replacing the sensitive terms with suitable hierarchical generalization fetched from knowledge bases (iii) the probability of Kullback–Leibler (KL) divergence between sensitive and generalized sanitization terms on Twitter, with negligible information disclosure risk is acceptable, and (iv) the sanitization carried out for 10 sensitive topics, from 4500 user posts of 790 Twitter users, demonstrated high precision and recall, which can be correlated with advanced privacy settings for OSN users in the near future.

Affordances-in-practice: An ethnographic critique of social media logic and context collapse.

Drawing on data gathered during ethnographic fieldwork in Mardin, a medium-sized town in southeast Turkey, this article shows that social media users actively appropriate online platforms and change privacy settings in order to keep different social spheres and social groups apart. Keeping different online social contexts distinct from each other is taken for granted as a way of using social media in Mardin. By contrast, social media scholars have extensively discussed the effects of social media in terms of context collapse. The article highlights how context collapse is the result of patterns of usage within Anglo-American contexts and not the consequence of a platform’s architecture or social media logic. It then suggests a theoretical refinement of affordances, and proposes the concept of affordances-in-practice.

Critical assessment of pediatric neurosurgery patient/parent educational information obtained via the Internet.

OBJECTIVE The Internet is used frequently by patients and family members to acquire information about pediatric neurosurgical conditions. The sources, nature, accuracy, and usefulness of this information have not been examined recently. The authors analyzed the results from searches of 10 common pediatric neurosurgical terms using a novel scoring test to assess the value of the educational information obtained. METHODS Google and Bing searches were performed for 10 common pediatric neurosurgical topics (concussion, craniosynostosis, hydrocephalus, pediatric brain tumor, pediatric Chiari malformation, pediatric epilepsy surgery, pediatric neurosurgery, plagiocephaly, spina bifida, and tethered spinal cord). The first 10 "hits" obtained with each search engine were analyzed using the Currency, Relevance, Authority, Accuracy, and Purpose (CRAAP) test, which assigns a numerical score in each of 5 domains. Agreement between results was assessed for 1) concurrent searches with Google and Bing; 2) Google searches over time (6 months apart); 3) Google searches using mobile and PC platforms concurrently; and 4) searches using privacy settings. Readability was assessed with an online analytical tool. RESULTS Google and Bing searches yielded information with similar CRAAP scores (mean 72% and 75%, respectively), but with frequently differing results (58% concordance/matching results). There was a high level of agreement (72% concordance) over time for Google searches and also between searches using general and privacy settings (92% concordance). Government sources scored the best in both CRAAP score and readability. Hospitals and universities were the most prevalent sources, but these sources had the lowest CRAAP scores, due in part to an abundance of self-marketing. The CRAAP scores for mobile and desktop platforms did not differ significantly (p = 0.49). CONCLUSIONS Google and Bing searches yielded useful educational information, using either mobile or PC platforms. Most information was relevant and accurate; however, the depth and breadth of information was variable. Search results over a 6-month period were moderately stable. Pediatric neurosurgery practices and neurosurgical professional organization websites were inferior (less current, less accurate, less authoritative, and less purposeful) to governmental and encyclopedia-type resources such as Wikipedia. This presents an opportunity for pediatric neurosurgeons to participate in the creation of better online patient/parent educational material.