DIPA2: An Image Dataset with Cross-cultural Privacy Perception Annotations

The world today is increasingly visual. Many of the most popular online social networking services are largely powered by images, making image privacy protection a critical research topic in the fields of ubiquitous computing, usable security, and human-computer interaction (HCI). One topical issue is understanding privacy-threatening content in images that are shared online. This dataset article introduces DIPA2, an open-sourced image dataset that offers object-level annotations with high-level reasoning properties to show perceptions of privacy among different cultures. DIPA2 provides 5,897 annotations describing perceived privacy risks of 3,347 objects in 1,304 images. The annotations contain the type of the object and four additional privacy metrics: 1) information type indicating what kind of information may leak if the image containing the object is shared, 2) a 7-point Likert item estimating the perceived severity of privacy leakages, and 3) intended recipient scopes when annotators assume they are either image owners or allowing others to repost the image. Our dataset contains unique data from two cultures: We recruited annotators from both Japan and the U.K. to demonstrate the impact of culture on object-level privacy perceptions. In this paper, we first illustrate how we designed and performed the construction of DIPA2, along with data analysis of the collected annotations. Second, we provide two machine-learning baselines to demonstrate how DIPA2 challenges the current image privacy recognition task. DIPA2 facilitates various types of research on image privacy, including machine learning methods inferring privacy threats in complex scenarios, quantitative analysis of cultural influences on privacy preferences, understanding of image sharing behaviors, and promotion of cyber hygiene for general user populations.

With the development of the mobile internet, service providers obtain data and resources through a large number of terminal user devices. They use private data for business empowerment, which improves the user experience while causing users’ privacy disclosure. Current research ignores the impact of disclosing user non-sensitive attributes under a single scenario of data sharing and lacks consideration of users’ privacy preferences. This paper constructs a data-sharing privacy metrics model based on information entropy and group privacy preferences. Use information theory to model the correlation of the privacy metrics problem, the improved entropy weight algorithm to measure the overall privacy of the data, and the analytic hierarchy process to correct user privacy preferences. Experiments show that this privacy metrics model can better quantify data privacy than conventional methods, provide a reliable evaluation mechanism for privacy security in data sharing and publishing scenarios, and help to enhance data privacy protection.

Content-based multi-source encrypted image retrieval in clouds with privacy preservation

Many cloud platforms emerge to meet urgent requirements for large-volume personal image store, sharing and search. Though most would agree that images contain rich sensitive information (e.g., People, location and event) and people's privacy concerns hinder their participation into untrusted services, today's cloud platforms provide little support for image privacy protection. Facing large-scale images from multiple users, it is extremely challenging for the cloud to maintain the index structure and schedule parallel computation without learning anything about the image content and indices. In this work, we introduce a novel system PIC: a Privacy-preserving Image search system on Cloud, which is a step towards feasible cloud services which provide secure content-based large-scale image search with fine-grained access control. Users can search on others' images if they are authorized by the image owners. Majority of the computationally intensive jobs are handled by the cloud, and a querier can now simply send the query and receive the result. Specially, to deal with massive images, we design our system suitable for distributed and parallel computation and introduce several optimizations to further expedite the search process. Our security analysis and prototype system evaluation results show that PIC successfully protects the image privacy at a low cost of computation and communication.

Many cloud platforms emerge to meet urgent requirements for large-volume personal image store, sharing and search. Though most would agree that images contain rich sensitive information (e.g., people, location and event) and people’s privacy concerns hinder their participation into untrusted services, today’s cloud platforms provide little support for image privacy protection. Facing large-scale images from multiple users, it is extremely challenging for the cloud to maintain the index structure and schedule parallel computation without learning anything about the image content and indices. In this work, we introduce a novel system PIC: A Privacy-preserving Image search system on Cloud, which is a step towards feasible cloud services which provide secure content-based large-scale image search with fine-grained access control. Users can search on others’ images if they are authorized by the image owners. Majority of the computationally intensive jobs are handled by the cloud, and a querier can now simply send the query and receive the result. Specially, to deal with massive images, we design our system suitable for distributed and parallel computation and introduce several optimizations to further expedite the search process. Our security analysis and prototype system evaluation results show that PIC successfully protects the image privacy at a low cost of computation and communication.

Interactive Digital TV (iDTV) is an emerging technology that faces problems that are inherent to it; for example the lack of users’ experience interacting with television content. The knowledge constructed from the Human Computer Interaction (HCI) field could be an ally for dealing with interaction design for the iDTV context. This work sought to map out the main issues that have been addressed in the iDTV and HCI fields in recent years, aiming at finding ways of bringing HCI to typical iDTV interaction issues. A data collection and analysis of tag clouds created from titles found in the full programs of two major conferences in the field of HCI (ACM CHI and IFIP Interact), and the major conference in the field of iDTV (EuroITV), complemented with other ACM-DL iDTV publications revealed the individual characteristics of HCI and iDTV publications, as well as their similarities and differences. Thus, this study offers a view of iDTV relative to the HCI field as revealed by the publications words.

The Computer-Human Interaction and Software Engineering (CHISE) series of edited volumes originated from a number of workshops and discussions over the latest research and developments in the field of Human Computer Interaction (HCI) and Software Engineering (SE) integration, convergence and cross-pollination. A first volume in this series (CHISE Volume I – Human-Centered Software Engineering: Integrating Usability in the Development Lifecycle) aims at bridging the gap between the field of SE and HCI, and addresses specifically the concerns of integrating usability and user-centered systems design methods and tools into the software development lifecycle and practices. This has been done by defining techniques, tools and practices that can fit into the entire software engineering lifecycle as well as by defining ways of addressing the knowledge and skills needed, and the attitudes and basic values that a user-centered development methodology requires. The first volume has been edited as Vol. 8 in the Springer HCI Series (Seffah, Gulliksen and Desmarais, 2005).

To enhance the performance of transportation systems a wide range of computer applications and computer software have been developed, which help users into the decision-making process. This paper intends to go into the field of human-computer interaction and transportation, by carrying out a literature review in order to recognize the usefulness of human-computer interaction and the way that it is used in the topic of transportation. human-computer interaction perspectives are applicable to describe, to analyze, to improve issues as user experience, user security, urban development, business model and so for. Understanding the way, a system is utilized is crucial to achieve the purpose of its design.

Billions images are shared daily on social networks. When shared with an inappropriate audience, user-generated images can, however, compromise users' privacy and may have severe consequences, such as dismissals. To address this issue, different solutions were proposed, ranging from graphical user interfaces to Deep Learning (DL) models to alert users based on image sensitivity prediction. Although these models show promising results, they are evaluated on datasets relying on small participants' samples. To address this limitation, we first introduce SensitivAlert, a dataset that re-annotates the previously annotated images from two existing datasets, but using a German-speaking cohort of 907 participants. We then leverage it to classify images according to two sensitivity classes---private or public---using recent transformer-based DL models. In our evaluation, we first consider consensus-based generic models using our dataset as benchmark based on image content itself and its associated user tags. Moreover, we show that our fine-tuned models trained on our dataset better reflect users' image privacy conceptions. We finally focus on individual user's privacy estimation by investigating three approaches: (1) a generic approach based on participants' consensus for fine-tuning, (2) a user-wise approach based on user's privacy preferences only, and (3) a hybrid approach that combines individual preferences with consensus-based preferences. Our results finally show that the generic and hybrid approaches outperform the user-wise one for most users, thus ensuring the feasibility of image privacy prediction preferences at the individuals' level.

E-learning has the potential to support the development of expertise in clinical reasoning by being able to provide students with interactive learning experiences, exposure to multiple cases, and opportunities for deliberate practice with tailored feedback. This review focuses on two important but underappreciated factors necessary for successful e-learning, i.e. the management of the learner’s cognitive load and the usability of the technology interface. Cognitive load theory views learning as involving active processing of information by working memory via separate visual and auditory channels. This system is of very limited capacity and any cognitive load that does not directly contribute to learning is considered extraneous and likely to impede learning. Researchers in cognitive load theory have provided evidence-based instructional design principles to reduce extraneous cognitive load and better manage the cognitive processing necessary for learning. Usability is a concept from the field of human-computer interaction which describes how easy technology interfaces are to use, and is routinely evaluated and optimised in the software development industry. This is seldom the case when e-learning resources are developed, especially in the area of medical education. Poor usability limits the potential benefit of educational resources, as learners experience difficulties with the technology interface while simultaneously dealing with the challenges of the content presented. Practitioners in the field of human-computer interaction have provided guidelines and methods for evaluating and optimising the usability of e-learning materials. The fields of cognitive load theory and human-computer interaction share a common goal in striving to reduce extraneous cognitive load. The load induced by poor usability of e-learning materials can be viewed as a specific component of extraneous cognitive load, adding to any load resulting from poor instructional design. The guidelines from these two fields are complementary and, if correctly implemented, may substantially improve the impact of our e-learning resources on the development of the clinical reasoning skills of students.

While recent research on intelligent transportation systems including vehicular communication systems has focused on technical aspects, little research work has been conducted on drivers’ privacy perceptions and preferences. Understanding the driver’s privacy perceptions and preferences will allow researchers to design usable privacy and identity management systems offering user privacy choices and controls for intelligent transportation systems. We conducted in-depth semi-structured interviews with 17 Swedish drivers to analyse their privacy perceptions and preferences for intelligent transportation systems, particularly for user control and for privacy trade-offs with cost, safety and usability. We also compare our results from the interviews with Swedish drivers with results from interviews that we conducted previously with South African drivers. Our cross-cultural comparison shows that perceived privacy implications, the drivers’ willingness to share location information under certain conditions with other parties, as well as their appreciation of Privacy Enhancing Technologies differ significantly across drivers with different cultural backgrounds. We further discuss the cultural impact on privacy preferences, including those for privacy trade-offs, and the implications of our results for usable privacy-enhancing Identity Management for future vehicular communication systems. In particular, we provide recommendations for suitable pre-defined privacy options to be offered to users with different cultural backgrounds enabling them to easily make privacy-related control choices.

More and more users openly share their information on online websites, with the resulting privacy issues being under scrutiny. Content such as a user’s personal data and location information is often asked for before posting to enforce the user’s privacy preferences; however, little attention has been paid to the lack of content (e.g., images) posted by the user. Even if privacy preferences are requested before images are published, publishers often remain unaware of the extent of privacy leakage associated with their data. To this end, we provide an image privacy metric scheme that incorporates users’ privacy preferences, with the core idea of assisting users in making data publishing decisions. First, we propose privacy-specific spatial attention mechanisms that can effectively improve the prediction accuracy. Next, we integrate set pair analysis (SPA) theory and use the network output as the privacy value. Finally, we combine a user study to understand the privacy preferences of different users with respect to these attributes and combine it with principal component analysis to correct and enforce user privacy preferences. Our model is trained with the ability to predict privacy risk end-to-end, thus being able to guide the user in sharing data in open platforms. We use the image privacy dataset, VISPR, to predict privacy information better than other methods.

Building management systems tout numerous benefits, such as energy efficiency and occupant comfort but rely on vast amounts of data from various sensors. Advancements in machine learning algorithms make it possible to extract personal information about occupants and their activities beyond the intended design of a non-intrusive sensor. However, occupants are not informed of data collection and possess different privacy preferences and thresholds for privacy loss. While privacy perceptions and preferences are most understood in smart homes, limited studies have evaluated these factors in smart office buildings, where there are more users and different privacy risks. To better understand occupants’ perceptions and privacy preferences, we conducted twenty-four semi-structured interviews between April 2022 and May 2022 on occupants of a smart office building. We found that data modality features and personal features contribute to people’s privacy preferences. The features of the collected modality define data modality features – spatial, security, and temporal context. In contrast, personal features consist of one’s awareness of data modality features and data inferences, definitions of privacy and security, and the available rewards and utility. Our proposed model of people’s privacy preferences in smart office buildings helps design more effective measures to improve people’s privacy.

The fields of Human Computer Interaction (HCI) and Information Visualization (InfoVis) could complement one another, rather than being treated as two different disciplines, in order to provide visual and interactive aspects for improved information visualization. A model is proposed in this light that allows both researchers and users to analyze a pronunciation learning-oriented task, based on analysis of three reference models for information visualization Card, McKinlay-Shneiderman; Chi; and Wunsche. These models are linear and although they offer data representation-focused approaches, they recognize the role of the user in their different stages.The three visualization models highlight such important aspects as representation, interaction and perception - elements of study that are essential in the field of HCI - in order to create a suitable representation of the information that enables a detailed understanding to be acquired and developed, and experiences to be communicated, wherein the user can interact more easily with the information and get the most from the visualization tool.This article aims to address both research fields. Information visualization is focused on visual representation of acoustic signals of the voice, based upon multidimensional data, while HCI takes into account user-oriented visual and interactive aspects. It is sought to integrate the two research fields in the proposed model, so that acoustic voice signal data are represented on a two dimensional plane and the user can visually comprehend as many aspects of the voice as possible, being able to recognize the quality of the pronunciation through information succinctly represented by graphic attributes.The model is comprised of stages and interaction mechanisms that combine to represent four views. Each view has a different makeup, in order to structure the data and display them using graphical attributes, such as color, position, shape, size, text, orientation and texture. These help in analysis and evaluation and in how to represent the connections between the various visual components in order to establish relationships between data.The main objective of the model is to represent a large number of aspects of the voice, so that in pronunciation training tasks a person can understand pronunciation quality visually. The model represents the results of comparing an input signal recorded by a microphone with a correctly pronounced signal set.Views constitute the final stage of the model. Data are thus presented so that the user can see and interpret representations of phoneme pronunciations. The views enable the user to learn about different aspects of the voice and recognize pronunciation quality visually. Each view represents different information: the first reveals information rapidly using facial gestures to represent moods - a happy face for good pronunciation and a sad face when the pronunciation is poor, making use of facial attributes such as eyes, eyebrows, nose and mouth. These represent values of voice characteristics through their size, shape, position or orientation. The second view permits a quick inspection of data, making comparison between correct pronunciation prototypes and the test signal and employing graphical attributes such as color and shape to interpret degrees of similarity of characteristics between phonemes. A third view illustrates similarities between data associated with a color, position and shape, using a visualization technique called Self-organizing Maps. Finally, a fourth view shows a diagonal line to indicate correlations between a set of phonemes relative to the phoneme pronounced. Each view has a level of interpretation and understanding depending on the degree of experience of the user - from a first view oriented for a child to a fourth view for a person skilled in signal modeling.The four views offer different visual explorations and levels of interpretation, so that the user can understand aspects of the voice and gain knowledge about the quality of the pronunciation. The model involves interaction with the user, which takes account of key aspects that might contribute to an easy decoding of information, in order to provide a better understanding of the task. They also allow the user to learn about the different aspects of the acoustic signal of the voice by means of various combinations of visualization and help an understanding of the data, taking full advantage of visual perception skills to discover patterns.The proposed model considers aspects of design in the display interface, aspects that could reduce the cognitive effort required to understand the graphic representation, so that users can devote their cognitive ability to understanding that which is being represented.Information visualization and HCI are two areas of research that can support each another with the aim of creating a visual representation oriented to the user, in such a way that as much information as possible can be gained and interpreted with the minimum of effort.

The systematic study of human-computer interaction has arguably been the most significant factor driving the exponential increase in technology acceptance, diffusion, and utilization, over the past two decades, as well as the technology-driven productivity gains that have benefited a full spectrum of organizations. Human Computer Interaction: Concepts, Methodologies, Tools, and Applications penetrates the human computer interaction (HCI) field with more breadth and depth of comprehensive research than any other publication. The four-volume set contains more than 200 authoritative works from over 250 leading experts in the field of human computer interaction. This groundbreaking collection contains significant chapters in topics such as Web logs, technology influences, and human factors of information systems and technologies. Human Computer Interaction: Concepts, Methodologies, Tools, and Applications is the single vital source for practitioners and academicians aspiring to expand their knowledge of HCI. Key Features:A? Authoritative contributions from over 250 international experts A? Comprehensive coverage of each specific topic, highlighting recent trends and describing the latest advances A? Cross referencing of key terms, figures and information pertinent to human computer interaction A? In-depth analysis of issues, concepts, new trends, and advanced technologies A? More than 5,000 references to existing literature and research in human computer interaction A? Organized by topic and indexed, making it a convenient method of reference for all IT/IS scholars and professionals A? Over 200 chapters covering the most incisive and critical research issues in human computer interaction