No description available.

Usable privacy and security has become a significant area of interest for many people in both industry and academia. A better understanding of the knowledge and motivation is an important factor in the design of privacy and security tools. However, users of these tools are a heterogeneous group, and many past studies of user characteristics in the security and privacy domain have looked only at a small subset of factors to define differences between groups of users. The goal of this research is to critically look at the difference between people, their opinions and habits when it comes to issues of privacy and security. To address this goal, 32 in-depth qualitative interviews were conducted and analyzed to look at the heterogenous nature of this community. The participant’s attitudes and actions around the dimensions of knowledge about tools and of motivation for self-protection were used to cluster participants. The traits of these participant clusters are used to create a set of privacy and security personas, or prototypical privacy and security tool users. These personas are a tool for incorporating a broader understanding of the diversity of users into the design of privacy and security tools.

This thesis investigates ways to reduce or eliminate the necessity of user input to learning-based personal email spam filters. Personal spam filters have been shown in previous studies to yield superior effectiveness, at the cost of requiring extensive user training which may be burdensome or impossible. This work describes new approaches to solve the problem of building a personal spam filter that requires minimal user feedback. An initial study investigates how well a personal filter can learn from different sources of data, as opposed to user’s messages. Our initial studies show that inter-user training yields substantially inferior results to intra-user training using the best known methods. Moreover, contrary to previous literature, it is found that transfer learning degrades the performance of spam filters when the source of training and test sets belong to two different users or different times. We also adapt and modify a graph-based semi-supervising learning algorithm to build a filter that can classify an entire inbox trained on twenty or fewer user judgments. Our experiments show that this approach compares well with previous techniques when trained on as few as two training examples. We also present the toolkit we developed to perform privacy-preserving user studies on spam filters. This toolkit allows researchers to evaluate any spam filter that conforms to a standard interface defined by TREC, on real users’ email boxes. Researchers have access only to the TREC-style result file, and not to any content of a user’s email stream. To eliminate the necessity of feedback from the user, we build a personal autonomous filter that learns exclusively on the result of a global spam filter. Our laboratory experiments show that learning filters with no user input can substantially improve the results of open-source and industry-leading commercial filters that employ no user-specific training. We use our toolkit to validate the performance of the autonomous filter in a user study.

Researchers analyze images in areas such as geology, bat cardiovascular systems and art studies to verify their observations. Some images are hard to study as their details are not vivid; hence there is a need to enhance these images to facilitate their study while preserving their contents. This study is aimed at assisting the researchers in the Cardiovascular Systems Dynamic Laboratory at Texas A&M University by evaluating the importance of Image Enhancement (IE) for improving visibility of features. For this study the images were collected and manipulated using various IE techniques and were shown to the novice researchers who were asked to perform three different tasks. These tasks were representative of the research work conducted in the lab. The techniques that were selected aimed at reducing the problems that are usually associated with data obtained from microscopic feeds. A customized application was developed to expedite and automate the study. The results indicated that the researchers did not immmensely benefit from the improved visualization for easy tasks. However, their performance improved for tasks that required more practice and skill. Our approach contributes towards designing an effective training program for novice researchers in the lab. Moreover, it is promising for similar research in different fields of study.

Usable privacy and security has become a significant area of interest for many people in both industry and academia. A better understanding of the knowledge and motivation is an important factor in the design of privacy and security tools. However, users of these tools are a heterogeneous group, and many past studies of user characteristics in the security and privacy domain have looked only at a small subset of factors to define differences between groups of users. The goal of this research is to critically look at the difference between people, their opinions and habits when it comes to issues of privacy and security. To address this goal, 32 in-depth qualitative interviews were conducted and analyzed to look at the heterogenous nature of this community. The participant’s attitudes and actions around the dimensions of knowledge about tools and of motivation for self-protection were used to cluster participants. The traits of these participant clusters are used to create a set of privacy and security personas, or prototypical privacy and security tool users. These personas are a tool for incorporating a broader understanding of the diversity of users into the design of privacy and security tools.

Mobile devices are frequently used to view rich content while on the go. However, they have a tradeoff between increased screen size and portability; mobile devices, by definition, are constrained to a fraction of a desktop computer’s display area. This constraint means a user has to frequently navigate to content that lies outside the display. We present FlexView, a prototype system and set of interaction techniques, which allows users to navigate through depth-arranged large information spaces using display curvature as an additional input channel. FlexView augments the planar (X-Y) navigation currently performed by touch input with two forms of bend input to navigate through depth (Z). With leafing, the user holds one side of display and bends the opposite side. Squeezing involves gripping the display in one hand and applying pressure on both sides to create concave or convex curvatures, and supports concurrent interaction with touch input. We performed two evaluations to investigate the performance of FlexView’s interaction techniques. In Experiment 1, we measured the efficiency of participants when searching through pages of a document, and compared touch input to squeezing and leafing used in isolation. Experiment 2 introduced X-Y navigation in a pan-and-zoom pointing task where multi-touch pinch gestures were compared against squeezing and leafing for zoom operations. Panning, across all conditions, was performed with touch input using the index finger. Our experiments demonstrated that touch and bend interactions are comparable for navigation through depth-arranged content, and squeezing to zoom recorded the fastest times in the pan-and-zoom pointing task. Overall, FlexView allows users to easily browse depth-arranged information spaces without sacrificing traditional touch interactions. / Thesis (Master, Computing) -- Queen's University, 2012-09-10 13:28:18.984

This thesis investigates ways to reduce or eliminate the necessity of user input to learning-based personal email spam filters. Personal spam filters have been shown in previous studies to yield superior effectiveness, at the cost of requiring extensive user training which may be burdensome or impossible. This work describes new approaches to solve the problem of building a personal spam filter that requires minimal user feedback. An initial study investigates how well a personal filter can learn from different sources of data, as opposed to user’s messages. Our initial studies show that inter-user training yields substantially inferior results to intra-user training using the best known methods. Moreover, contrary to previous literature, it is found that transfer learning degrades the performance of spam filters when the source of training and test sets belong to two different users or different times. We also adapt and modify a graph-based semi-supervising learning algorithm to build a filter that can classify an entire inbox trained on twenty or fewer user judgments. Our experiments show that this approach compares well with previous techniques when trained on as few as two training examples. We also present the toolkit we developed to perform privacy-preserving user studies on spam filters. This toolkit allows researchers to evaluate any spam filter that conforms to a standard interface defined by TREC, on real users’ email boxes. Researchers have access only to the TREC-style result file, and not to any content of a user’s email stream. To eliminate the necessity of feedback from the user, we build a personal autonomous filter that learns exclusively on the result of a global spam filter. Our laboratory experiments show that learning filters with no user input can substantially improve the results of open-source and industry-leading commercial filters that employ no user-specific training. We use our toolkit to validate the performance of the autonomous filter in a user study.

User generated reviews, like those found on Yelp and Amazon, have become important reference material in casual decision making, like dining, shopping and entertainment. However, very large amounts of reviews make the review reading process time consuming. A text visualization can speed up the review reading process. In this thesis, we present the clustered layout word cloud -- a text visualization that quickens decision making based on user generated reviews. We used a natural language processing approach, called grammatical dependency parsing, to analyze user generated review content and create a semantic graph. A force-directed graph layout was applied to the graph to create the clustered layout word cloud. We conducted a two-task user study to compare the clustered layout word cloud to two alternative review reading techniques: random layout word cloud and normal block-text reviews. The results showed that the clustered layout word cloud offers faster task completion time and better user satisfaction than the other two alternative review reading techniques. [Permission email from J. Huang removed at his request. GMc March 11, 2014] / Master of Science

Large, high resolution displays combine high pixel density with ample physical dimensions. Combination of these two factors creates a multi-scale workspace where object targeting requires both high speed and high accuracy for nearby and far apart targeting. Modern operating systems support dynamic control-display gain adjustment (i.e. cursor acceleration) that helps to maintain both speed and accuracy. However, very large high resolution displays require broad range of control-display gain ratios. Current interaction techniques attempt to solve the problem by utilizing multiple modes of interaction, where different modes provide different levels of pointer precision. We are investigating the question of the value of allowing users to dynamically choose granularity levels for continuous pointing within single mode of interaction via multiple acceleration curves. Our solution offers different cursor acceleration curves depending on the targeting conditions, thus broadening the range of control-display ratios. Our approach utilizes a consumer multitouch touchpad that allows fast and accurate detection of multiple fingers. A user can choose three different acceleration curves based on how many fingers are used for cursor positioning. Our goal is to investigate the effects of such multi-scale interaction and to compare it against standard single curve interaction. / Master of Science

Dark Patterns are a variety of different software designs that are used to manipulate and mislead the users of an application or service. These patterns range from making it harder to end a subscription service, adding additional charges to a purchase, or having the user give out data or personal information. With how widespread and varied dark patterns are, it led to us creating a way to detect and warn users of different dark patterns. In this study, we created Dark Pattern Detector, a Chrome extension that would help users detect and understand three different dark patterns: Hidden Costs, Disguised Ads, and Sneak into Basket. This extension was made to detect each of these patterns on any web page while not requiring any information from the user or their data. Study participants installed the extension and completed a series of tasks given to them that would occur on different websites containing the previous dark patterns. After completing the tasks, the users were surveyed to give feedback on what they thought of the extension and what suggestions for change they had. In the study, we had 40 participants and we found that 50% of the users were completely unfamiliar with dark patterns and that 77.5% have used extensions before. For the five tasks, each one had a majority of the participants successfully complete them. Finally, when asked about what they thought, the majority of the participants gave positive feedback claiming that they found the extension useful, interesting, and a good idea. Many participants also gave useful feedback about what changes or additions they would like to see. With our results, we can help users have a better understanding of dark patterns and have created a baseline for any future research done on dark pattern knowledge and detection. / Master of Science / Dark patterns are designs on the internet that websites use to trick its users. They may be used to hide advertisements, make the user spend more time or money on their website or more. Our goal was to create a way to help protect anyone on the internet and their information. For this study, we created a program called Dark Pattern Detector that would help the users see different dark patterns that appeared on websites. A study was conducted that had the participants use our program and give us feedback on what they thought of it as well as data on how well it worked. Out of the 40 participants, we found that half the users were unfamiliar with what dark patterns were. Once they completed the study, we saw that the majority of users were able to complete tasks while using our program and gave positive feedback. Seeing the positive feedback and results from our study, we believe that we can help users not get tricked by these patterns and help forward future research on Dark Patterns.