With the increasing number of smartphone devices and social media platforms, many users now have large personal image collections. As these collections expand, the task of organizing and retrieving specific images becomes increasingly challenging. To quantify the scope of this emerging problem, we conducted a quantitative survey study to gain insights into users' practices concerning personal image retrieval and organization. Initially, we conducted a survey on the Florida State University campus, primarily targeting undergraduate and graduate students. The survey questionnaire delved into various aspects of how users organize and retrieve their images. Then we implemented several machine learning models (decision trees, random forest, logistic regression, and XGboost) on the collected survey data to determine the existence of problems faced by the users to retrieve their personal images. XGboost performed better than the other models with an accuracy of 73%. The model also revealed that several factors, such as the frequency of encountering difficulties in finding photos both before and after sharing, the number of photos taken by users in the previous year, the number of photos posted on social media, and the total number of photos stored on users' laptops or desktops, were among the most critical features associated with personal image retrieval challenges.

Engineering the abundant phase interface is an effective strategy in enhancing the electrocatalytic activity of hybrid materials. Herein, CoFe-layered double hydroxide and Co-metal-organic framework (CoFe-LDH/Co-MOF) heteronanotube arrays are synthesized in situ via a straightforward three-step reaction process. The uniform hollow structure, substantial lattice defects and synergistic effects of CoFe-LDH and Co-MOF components, significantly promote the electronic and chemical structure of the heteronanotube arrays. Optimized heteronanotube arrays display exceptional performance in oxygen evolution reaction (OER) with ultralow overpotentials of 215 and 455 mV to deliver current densities of 10 mA cm−2 and 100 mA cm−2 with a small Tafel slope of 43 mV dec−1, and maintaining the electrocatalytic activity for a duration as long as 220 h, ranking it one of the top-performing MOF and non-noble-metal-based electrocatalysts for OER. Density functional theory calculations validate the reduction in free energy for the rate-determining step by the synergistic effect of Co-MOF and CoFe-LDH. This work establishes experimental and theoretical basis for promoting efficient water splitting through the design of heterostructures and the coupling of defects in catalysts.

The number of photographs taken worldwide is growing rapidly and steadily. While a small subset of these images is annotated and shared by users through social media platforms, due to the sheer number of images in personal photo repositories (shared or not shared), finding specific images remains challenging. This survey explores existing image retrieval techniques as well as photo-organizer applications to highlight their relative strengths in addressing this challenge.

The number of photographs taken worldwide is growing rapidly and steadily. While a small subset of these images is annotated and shared by users through social media platforms, due to the sheer number of images in personal photo repositories (shared or not shared), finding specific images remains challenging. This survey explores existing image retrieval techniques as well as photo-organizer applications to highlight their relative strengths in addressing this challenge.

This paper documents the findings of our analysis of the implementation of our six-year NSF S-STEM scholarship program. One major finding was that, for underrepresented students to major in computer science, knowing the major existed and understanding the nature of the program were the most important factors. Also, the academic support system and hands-on nature of the major had a significant impact on scholarship recipients' persistence in the major. Evidence demonstrated that scholarship recipients had a 10%+ higher year-to-year persistence rate from their freshmen to sophomore year than that of all computer science students of the same entering classes. For all computer science students, college computer science major GPAs were not strongly correlated with their high school GPAs, financial need, or ACT math scores. This paper also presents lessons learned and resulting recommendations for future new scholarship administrators, as our lessons can likely be applied to other grants that recruit and deal with underrepresented groups.

Traditional file system optimizations typically use a one-to-one mapping of logical files to their physical metadata representations. This mapping results in missed opportunities for a class of optimizations in which such coupling is removed.We have designed, implemented, and evaluated a composite-file file system, which allows many-to-one mappings of files to metadata, and we have explored the design space of different mapping strategies. Under webserver and software development workloads, our empirical evaluation shows up to a 27% performance improvement. This result demonstrates the promise of composite files.

The legacy storage data path is largely structured in black-box layers and has four major limitations: (1) functional redundancies across layers, (2) poor crosslayer coordination and data tracking, (3) presupposition of high-latency storage devices, and (4) poor support for new storage data models. We introduce Tags, a unifying primitive that can be used throughout the storage data path. This white-box approach enables all storage layers to coordinate and track data using shared data structures that are constructed through the Tags API. Our case studies show that by eliminating redundant services, our Tagsbased key-value store can outperform LevelDB by 20170% when inserting and deleting 100-byte key-value pairs. We also build a Tags-based file system (TagFS) to demonstrate the usability and robustness of Tags. In addition, we build per-file secure deletion via TagFS to show data-path-wide coordination and data tracking.

Journaling has been a popular reliability tool for file systems. However, journaling may involve writing updates twice, once to the journal, and once to potentially random update locations. The journal granularity tends to be coarse, leading to more bytes written than necessary. We introduce Fine-grained Journal Store (FJS). Finegrained Journal Store uses the journal as the final storage location for updates, thus eliminating double and random writes. Fine-grained journaling units are utilized to reduce the number of bytes written compared to the number of bytes updated. To limit the memory overhead of tracking small updatable units, FJS uses a range-based lookup table. In addition, FJS reuses existing journaling constructs to avoid the need for additional byte-addressable NVM storage devices. We extended ext4 and jbd2 to prototype FJS. Our results show that FJS can outperform ext4 by up to 15x and reduce the write amplification of common metadata types by up to 5.8x. CCS Concepts •Information systems~Record storage systems

We have designed, prototyped, and evaluated the Legend file system, which exploits the ability to regenerate lost files to improve reliability. Unlike RAID-5 and RAID-6, Legend degrades gracefully as the number of failed storage devices increases. Legend also mitigates the high storage capacity overhead imposed by n-way replications. Combined with existing methods, Legend can form another line of defense against data loss.

Privacy is an important concern for mobile computing. Users might not understand the privacy implications of their actions and therefore not alter their behavior depending on where they move, when they do so, and who is in their surroundings. Since empirical data about the privacy behavior of users in mobile environments is limited, we conducted a survey study of ~600 users recruited from Florida State University and Craigslist. Major findings include: (1) People often exercise little caution preserving privacy in mobile computing environments; they perform similar computing tasks in public and private. (2) Privacy is orthogonal to trust; people tend to change their computing behavior more around people they know than strangers. (3) People underestimate the privacy threats of mobile apps, and comply with permission requests from apps more often than operating systems. (4) Users' understanding of privacy is different from that of the security community, suggesting opportunities for additional privacy studies.

Power is becoming a major concern when designing storage systems for platforms ranging from mobile devices to data centers. While many solutions exist, different solutions make very different tradeoffs between energy savings and storage performance, capacity, reliability, cost of ownership, etc. This survey walks through layers of the legacy storage stack, exploring tradeoffs made by a representative set of energy-efficient storage approaches. The survey also points out architectural implications of implementing energy-efficient solutions at various storage layers.

The amount of sensitive data stored on electronic media increases as the use of computers and mobile devices becomes more prevalent. For example, home computers and devices may store financial information (e.g., Quicken files or tax documents), usernames and passwords, private correspondence (e.g., emails or chat logs), and personal media files (e.g., pictures or videos). Business computers and devices may store sensitive client data and trade secrets. Government computers and devices may store personally identifiable data on citizens and various classified materials. As the amount of digital sensitive information accrues, the need for the ability to securely remove this information increases. Short of physically destroying the entire storage medium, existing secure-deletion solutions tend to be piecemeal at best – they may only work for one type of storage or file system, may force the user to delete all files instead of selective files, may require the added complexities of encryption and key storage, may require extensive changes and additions to the computer's operating system or storage firmware, and may not handle system crashes gracefully. This dissertation introduces TrueErase, a holistic secure-deletion framework that irrevocably deletes data and metadata. At heart, TrueErase is an information-propagation framework that works alongside of legacy operating system components for easier integration. Through its design, implementation, verification, and evaluation on both a hard drive and emerging solid-state storage, TrueErase shows that it is possible to construct a holistic, per-file, encryption-free, secure-deletion framework that accommodates different storage media and legacy file systems, requires limited changes to legacy systems, and handles common crash scenarios. The experience of building TrueErase further contributes insight into the mechanisms and complexities of the legacy operating system storage data path.

Smartphones are sensor-rich and Internet-enabled. With their on-board sensors, web services, social media, and external biosensors, smartphones can provide contextual information about the device, user, and environment, thereby enabling the creation of rich, biologically driven applications. We introduce ContextProvider, a framework that offers a unified, query-able interface to contextual data on the device. Unlike other context-based frameworks, ContextProvider offers interactive user feedback, self-adaptive sensor polling, and minimal reliance on third-party infrastructure. ContextProvider also allows for rapid development of new context and bio-aware applications. Evaluation of ContextProvider shows the incorporation of an additional monitoring sensor into the framework with fewer than 100 lines of Java code. With adaptive sensor monitoring, power consumption per sensor can be reduced down to 1% overhead. Finally, through the use of context, accuracy of data interpretation can be improved by up to 80%.

The ability to delete sensitive data securely from electronic storage is becoming an increasing concern. However, current per-file deletion solutions tend to be limited to one segment of the operating system’s storage data path, and may leave behind sensitive data when interacting with storage components such as journaling, file-system caches, and certain storage media such as NAND flash. This work introduces TrueErase, a secure-deletion framework. Through design, implementation, verification, and evaluation, we show that it is possible to build a legacy-compatible full-storage-data-path framework that performs per-file secure deletion, works with common file systems and emerging solid-state storage, while surviving common system failures.

The specification of the sporadic server real-time scheduling policy in the IEEE POSIX standard is defective, and needs to be corrected. Via experiments using a POSIX sporadic server implementation under Linux, as well as simulations, we have shown and confirmed previously unreported defects. We propose and demonstrate a corrected sporadic server formulation that eliminates these defects without changes to the syntax of the API or any significant increase in implementation complexity.

Reliable software systems typically have a version release mechanism that is well organized and documented. This can be drawn upon to predict release timelines, which is helpful in gauging the quality of the software development and maintenance activity. In this paper we present initial results from developing and applying Predictux – a decision-tree-based framework to predict release times of Linux kernel versions. We compare predictions from the framework with actual data and discuss our future plans for refining Predictux further.

Modern storage consists of many optimizations along the data path, from the front-end file system cache and disk layout management to the back-end disk controller cache and on-disk caching. Various components generally exploit the temporal and spatial localities of a workload to improve performance. However, how localities of a workload are transformed by each optimization is not well understood, resulting in designs that are more reflective of the understanding of the front-end workload than that of the locality characteristics immediately before the component. This research studies how file system caching transforms temporal and spatial localities. This understanding helps us quantify how before-cache workloads (e.g., references sent from an application to the OS) properly stress after-cache components (e.g., disks), how synthesized before-cache workloads have different after-cache effects than original workloads from which they are derived, and what opportunities remain for further after-cache optimizations. Quantifying localities is challenging for three reasons: (1) File system caching induces system traffic due to prefetching, buffered writes, page replacements, metadata accesses, and system events, which are sensitive to time and memory resources. Therefore, accesses before cache do not have one-to-one mapping to accesses after cache. (2) The semantics of locality depends on the granularity of analysis. Accesses before cache are in files, directories, and links. Accesses after cache are in blocks. Locality computed based on the distance between file references tends to be poorer than that based on blocks, since blocks tend to be referenced sequentially within files. (3) Locality metrics need to be comparable across workloads and environments. A workload with a “90% spatial locality” on a 50-GB drive should show meaningful behaviors when being applied to a 100-GB drive. Existing research evaluates caching, temporal locality, and spatial locality independently, with certain limitations. The cache-hit-rate metric [3] is used to compare caching policies without showing how workloads are transformed. Stack and block distances [1] can measure how localities are transformed by caches, but these metrics are highly sensitive to system settings, making results difficult to compare across workloads and environments. Some studies evaluate the effects of caching algorithms and cache sizes on the reference stream, but often exclude system traffic from analyses [2]. We propose two metrics—temporal and spatial affinities: ( ) dist reference affinity temporal _ 10 log 1 _ 10 + =

While wireless vehicular ad hoc networks (VANETs) are attracting greater commercial interest, current research has not adequately captured the real-world urban constraints in VANET deployment. In this work, we evaluated the feasibility and benefits of deploying a VANET in urban settings with a wireless mesh backbone infrastructure. We modeled urban street layouts, traffic rules, RF attenuation due to physical obstacles, and the use of multiple radio channels Our results indicate that the performance improves in dense networks when routing decisions are limited to mesh nodes, whereas it improves in sparse networks when mobile nodes also participate in routing. We also show that the effect of signal attenuation due to physical obstacles can potentially be parametrized in simulations using empirical real-world measurements.

The growing interest in wireless Vehicular Ad Hoc Networks (VANETs) has prompted greater research into simulation models that better reflect urban VANET deployments. Still, we lack a systematic understanding of the required level of simulation details in modeling various real-world urban constraints. In this work, we developed a series of simulation models that account for street layout, traffic rules, multilane roads, acceleration-deceleration, and RF attenuation due to obstacles. Using real and controlled synthetic maps, we evaluated the sensitivity of the simulation results toward these details. Our results indicate that the delivery ratio and packet delays in VANETs are more sensitive to the clustering effect of vehicles at intersections and their acceleration/deceleration. The VANET performance appears to be only marginally affected by the simulation of multiple lanes and careful synchronization at traffic signals. We also found that the performance in dense VANETs improves significantly when routing decisions are limited to a wireless backbone of mesh nodes, whereas in sparse VANETs, performance improves when vehicles also participate in ad hoc routing. Finally, through measurement and analysis of signal strengths around urban city blocks, we show that the effect of signal attenuation due to physical obstacles can potentially be parameterized in simulations. Our work provides a starting point for further understanding and development of more accurate VANET simulation model.