Search

Perhaps the most straightforward classifier in the arsenal or Machine Learning techniques is the Nearest Neighbour Classifier—classification is achieved by identifying the nearest neighbours to a query example and using those neighbours to determine the class of the query. This approach to classification is of particular importance, because issues of poor runtime performance is not such a problem these days with the computational power that is available. This article presents an overview of techniques for Nearest Neighbour classification focusing on: mechanisms for assessing similarity (distance), computational issues in identifying nearest neighbours, and mechanisms for reducing the dimension of the data. This article is the second edition of a paper previously published as a technical report [16]. Sections on similarity measures for time-series, retrieval speedup, and intrinsic dimensionality have been added. An Appendix is included, providing access to Python code for the key methods.

The recent boom of autonomous driving nowadays has made object detection in traffic scenes a hot topic of research. Designed to classify and locate instances in the image, this is a basic but challenging task in the computer vision field. With its powerful feature extraction abilities, which are vital for object detection, deep learning has expanded its application areas to this field during the past several years and thus achieved breakthroughs. However, even with such powerful approaches, traffic scenarios have their own specific challenges, such as real-time detection, changeable weather, and complex lighting conditions. This survey is dedicated to summarizing research and papers on applying deep learning to the transportation environment in recent years. More than 100 research papers are covered, and different aspects such as key generic object detection frameworks, categorized object detection applications in traffic scenario, evaluation metrics, and classified datasets are included. Some open research fields are also provided. We believe that it is the first survey focusing on deep learning-based object detection in traffic scenario.

Blockchain technology has been gaining visibility owing to its ability to enhance the security, reliability, and robustness of distributed systems. Several areas have benefited from research based on this technology, such as finance, remote sensing, data analysis, and healthcare. Data immutability, privacy, transparency, decentralization, and distributed ledgers are the main features that make blockchain an attractive technology. However, healthcare records that contain confidential patient data make this system very complicated because there is a risk of a privacy breach. This study aims to address research into the applications of the blockchain healthcare area. It sets out by discussing the management of medical information, as well as the sharing of medical records, image sharing, and log management. We also discuss papers that intersect with other areas, such as the Internet of Things, the management of information, tracking of drugs along their supply chain, and aspects of security and privacy. As we are aware that there are other surveys of blockchain in healthcare, we analyze and compare both the positive and negative aspects of their papers. Finally, we seek to examine the concepts of blockchain in the medical area, by assessing their benefits and drawbacks and thus giving guidance to other researchers in the area. Additionally, we summarize the methods used in healthcare per application area and show their pros and cons.

Human activity recognition is an important area of computer vision research. Its applications include surveillance systems, patient monitoring systems, and a variety of systems that involve interactions between persons and electronic devices such as human-computer interfaces. Most of these applications require an automated recognition of high-level activities, composed of multiple simple (or atomic) actions of persons. This article provides a detailed overview of various state-of-the-art research papers on human activity recognition. We discuss both the methodologies developed for simple human actions and those for high-level activities. An approach-based taxonomy is chosen that compares the advantages and limitations of each approach. Recognition methodologies for an analysis of the simple actions of a single person are first presented in the article. Space-time volume approaches and sequential approaches that represent and recognize activities directly from input images are discussed. Next, hierarchical recognition methodologies for high-level activities are presented and compared. Statistical approaches, syntactic approaches, and description-based approaches for hierarchical recognition are discussed in the article. In addition, we further discuss the papers on the recognition of human-object interactions and group activities. Public datasets designed for the evaluation of the recognition methodologies are illustrated in our article as well, comparing the methodologies' performances. This review will provide the impetus for future research in more productive areas.

Virtual machine consolidation has been a widely explored topic in recent years due to Cloud Data Centers’ effect on global energy consumption. Thus, academia and companies made efforts to achieve green computing, reducing energy consumption to minimize environmental impact. By consolidating Virtual Machines into a fewer number of Physical Machines, resource provisioning mechanisms can shutdown idle Physical Machines to reduce energy consumption and improve resource utilization. However, there is a tradeoff between reducing energy consumption while assuring the Quality of Service established on the Service Level Agreement. This work introduces a Systematic Literature Review of one year of advances in virtual machine consolidation. It provides a discussion on methods used in each step of the virtual machine consolidation, a classification of papers according to their contribution, and a quantitative and qualitative analysis of datasets, scenarios, and metrics.

Distributed data processing is becoming a reality. Businesses want to do it for many reasons, and they often must do it in order to stay competitive. While much of the infrastructure for distributed data processing is already there (e.g., modern network technology), a number of issues make distributed data processing still a complex undertaking: (1) distributed systems can become very large, involving thousands of heterogeneous sites including PCs and mainframe server machines; (2) the state of a distributed system changes rapidly because the load of sites varies over time and new sites are added to the system; (3) legacy systems need to be integrated—such legacy systems usually have not been designed for distributed data processing and now need to interact with other (modern) systems in a distributed environment. This paper presents the state of the art of query processing for distributed database and information systems. The paper presents the “textbook” architecture for distributed query processing and a series of techniques that are particularly useful for distributed database systems. These techniques include special join techniques, techniques to exploit intraquery paralleli sm, techniques to reduce communication costs, and techniques to exploit caching and replication of data. Furthermore, the paper discusses different kinds of distributed systems such as client-server, middleware (multitier), and heterogeneous database systems, and shows how query processing works in these systems.

The management of electronic document collections is fundamentally different from the management of paper documents. The ephemeral nature of some electronic documents means that the document address (i.e., reference details of the document) can become incorrect some time after coming into use, resulting in references, such as index entries and hypertext links, failing to correctly address the document they describe. A classic case of invalidated references is on the World Wide Web—links that point to a named resource fail when the domain name, file name, or any other aspect of the addressed resource is changed, resulting in the well-known Error 404. Additionally, there are other errors which arise from changes to document collections. This paper surveys the strategies used both in World Wide Web software and other hypertext systems for managing the integrity of references and hence the integrity of links. Some strategies are preventative , not permitting errors to occur; others are corrective , discovering references errors and sometimes attempting to correct them; while the last strategy is adaptive , because references are calculated on a just-in-time basis, according the current state of the document collection.

This paper compares different indexing techniques proposed for supporting efficient access to temporal data. The comparison is based on a collection of important performance criteria, including the space consumed, update processing, and query time for representative queries. The comparison is based on worst-case analysis, hence no assumptions on data distribution or query frequencies are made. When a number of methods have the same asymptotic worst-case behavior, features in the methods that affect average case behavior are discussed. Additional criteria examined are the pagination of an index, the ability to cluster related data together, and the ability to efficiently separate old from current data (so that larger archival storage media such as write-once optical disks can be used). The purpose of the paper is to identify the difficult problems in accessing temporal data and describe how the different methods aim to solve them. A general lower bound for answering basic temporal queries is also introduced.

Fashion is the way we present ourselves to the world and has become one of the world's largest industries. Fashion, mainly conveyed by vision, has thus attracted much attention from computer vision researchers in recent years. Given the rapid development, this paper provides a comprehensive survey of more than 200 major fashion-related works covering four main aspects for enabling intelligent fashion: (1) Fashion detection includes landmark detection, fashion parsing, and item retrieval, (2) Fashion analysis contains attribute recognition, style learning, and popularity prediction, (3) Fashion synthesis involves style transfer, pose transformation, and physical simulation, and (4) Fashion recommendation comprises fashion compatibility, outfit matching, and hairstyle suggestion. For each task, the benchmark datasets and the evaluation protocols are summarized. Furthermore, we highlight promising directions for future research., Accepted by ACM Computing Surveys (2021). 39 pages including 2 pages of supplementary materials and 7 pages of reference

Recommender systems are software applications that help users to find items of interest in situations of information overload. Current research often assumes a one-shot interaction paradigm, where the users' preferences are estimated based on past observed behavior and where the presentation of a ranked list of suggestions is the main, one-directional form of user interaction. Conversational recommender systems (CRS) take a different approach and support a richer set of interactions. These interactions can, for example, help to improve the preference elicitation process or allow the user to ask questions about the recommendations and to give feedback. The interest in CRS has significantly increased in the past few years. This development is mainly due to the significant progress in the area of natural language processing, the emergence of new voice-controlled home assistants, and the increased use of chatbot technology. With this paper, we provide a detailed survey of existing approaches to conversational recommendation. We categorize these approaches in various dimensions, e.g., in terms of the supported user intents or the knowledge they use in the background. Moreover, we discuss technological approaches, review how CRS are evaluated, and finally identify a number of gaps that deserve more research in the future., Comment: 35 pages, 5 figures

With the rapid development of virtualization techniques, cloud data centers allow for cost effective, flexible, and customizable deployments of applications on virtualized infrastructure. Virtual machine (VM) placement aims to assign each virtual machine to a server in the cloud environment. VM Placement is of paramount importance to the design of cloud data centers. Typically, VM placement involves complex relations and multiple design factors as well as local policies that govern the assignment decisions. It also involves different constituents including cloud administrators and customers that might have disparate preferences while opting for a placement solution. Thus, it is often valuable to not only return an optimized solution to the VM placement problem but also a solution that reflects the given preferences of the constituents. In this paper, we provide a detailed review on the role of preferences in the recent literature on VM placement. We further discuss key challenges and identify possible research opportunities to better incorporate preferences within the context of VM placement., Comment: 40 pages, 5 figures, 6 tables

Interactive methods are useful decision-making tools for multiobjective optimization problems, because they allow a decision-maker to provide her/his preference information iteratively in a comfortable way at the same time as (s)he learns about all different aspects of the problem. A wide variety of interactive methods is nowadays available, and they differ from each other in both technical aspects and type of preference information employed. Therefore, assessing the performance of interactive methods can help users to choose the most appropriate one for a given problem. This is a challenging task, which has been tackled from different perspectives in the published literature. We present a bibliographic survey of papers where interactive multiobjective optimization methods have been assessed (either individually or compared to other methods). Besides other features, we collect information about the type of decision-maker involved (utility or value functions, artificial or human decision-maker), the type of preference information provided, and aspects of interactive methods that were somehow measured. Based on the survey and on our own experiences, we identify a series of desirable properties of interactive methods that we believe should be assessed.

Log messages have been used widely in many software systems for a variety of purposes during software development and field operation. There are two phases in software logging: log instrumentation and log management. Log instrumentation refers to the practice that developers insert logging code into source code to record runtime information. Log management refers to the practice that operators collect the generated log messages and conduct data analysis techniques to provide valuable insights of runtime behavior. There are many open source and commercial log management tools available. However, their effectiveness highly depends on the quality of the instrumented logging code, as log messages generated by high-quality logging code can greatly ease the process of various log analysis tasks (e.g., monitoring, failure diagnosis, and auditing). Hence, in this article, we conducted a systematic survey on state-of-the-art research on log instrumentation by studying 69 papers between 1997 and 2019. In particular, we have focused on the challenges and proposed solutions used in the three steps of log instrumentation: (1) logging approach; (2) logging utility integration; and (3) logging code composition. This survey will be useful to DevOps practitioners and researchers who are interested in software logging.

The purpose of a distributed file system (DFS) is to allow users of physically distributed computers to share data and storage resources by using a common file system. A typical configuration for a DFS is a collection of workstations and mainframes connected by a local area network (LAN). A DFS is implemented as part of the operating system of each of the connected computers. This paper establishes a viewpoint that emphasizes the dispersed structure and decentralization of both data and control in the design of such systems. It defines the concepts of transparency, fault tolerance, and scalability and discusses them in the context of DFSs. The paper claims that the principle of distributed operation is fundamental for a fault tolerant and scalable DFS design. It also presents alternatives for the semantics of sharing and methods for providing access to remote files. A survey of contemporary UNIX-based systems, namely, UNIX United, Locus, Sprite, Sun's Network File System, and ITC's Andrew, illustrates the concepts and demonstrates various implementations and design alternatives. Based on the assessment of these systems, the paper makes the point that a departure from the extending centralized file systems over a communication network is necessary to accomplish sound distributed file system design.

Database systems were a solution to the problem of shared access to heterogeneous files created by multiple autonomous applications in a centralized environment. To make data usage easier, the files were replaced by a globally integrated database. To a large extent, the idea was successful, and many databases are now accessible through local and long-haul networks. Unavoidably, users now need shared access to multiple autonomous databases. The question is what the corresponding methodology should be. Should one reapply the database approach to create globally integrated distributed database systems or should a new approach be introduced? We argue for a new approach to solving such data management system problems, called multidatabase or federated systems. These systems make databases interoperable, that is, usable without a globally integrated schema. They preserve the autonomy of each database yet support shared access. Systems of this type will be of major importance in the future. This paper first discusses why this is the case. Then, it presents methodologies for their design. It further shows that major commerical relational database systems are evolving toward multidatabase systems. The paper discusses their capabilities and limitations, presents and discusses a set of prototypes, and, finally, presents some current research issues.

In the past five years, distributed operating systems research has gone through a consolidation phase. On a large number of design issues there is now considerable consensus between different research groups. In this paper, an overview of recent research in distributed systems is given. In turn, the paper discusses overall system structure, protection issues, file system designs, problems and solutions for fault tolerance and a mechanism that is rapidly becoming very important for efficient distributed systems design: hints. An attempt was made to provide sufficient references to interesting research projects for the reader to find material for more detailed study.

Recommender systems have become a popular and effective means to manage the ever-increasing amount of multimedia content available today and to help users discover interesting new items. Today’s recommender systems suggest items of various media types, including audio, text, visual (images), and videos. In fact, scientific research related to the analysis of multimedia content has made possible effective content-based recommender systems capable of suggesting items based on an analysis of the features extracted from the item itself. The aim of this survey is to present a thorough review of the state-of-the-art of recommender systems that leverage multimedia content, by classifying the reviewed papers with respect to their media type, the techniques employed to extract and represent their content features, and the recommendation algorithm. Moreover, for each media type, we discuss various domains in which multimedia content plays a key role in human decision-making and is therefore considered in the recommendation process. Examples of the identified domains include fashion, tourism, food, media streaming, and e-commerce.

Cryptographic algorithms have been used not only to create robust ciphertexts but also to generate cryptograms that, contrary to the classic goal of cryptography, are meant to be broken. These cryptograms, generally called puzzles, require the use of a certain amount of resources to be solved, hence introducing a cost that is often regarded as a time delay---though it could involve other metrics as well, such as bandwidth. These powerful features have made puzzles the core of many security protocols, acquiring increasing importance in the IT security landscape. The concept of a puzzle has subsequently been extended to other types of schemes that do not use cryptographic functions, such as CAPTCHAs, which are used to discriminate humans from machines. Overall, puzzles have experienced a renewed interest with the advent of Bitcoin, which uses a CPU-intensive puzzle as proof of work. In this paper, we provide a comprehensive study of the most important puzzle construction schemes available in the literature, categorizing them according to several attributes, such as resource type, verification type, and applications. We have redefined the term puzzle by collecting and integrating the scattered notions used in different works, to cover all the existing applications. Moreover, we provide an overview of the possible applications, identifying key requirements and different design approaches. Finally, we highlight the features and limitations of each approach, providing a useful guide for the future development of new puzzle schemes., This article has been accepted for publication in ACM Computing Surveys

The past decade has witnessed great success of deep learning technology in many disciplines, especially in computer vision and image processing. However, deep learning-based video coding remains in its infancy. This paper reviews the representative works about using deep learning for image/video coding, which has been an actively developing research area since the year of 2015. We divide the related works into two categories: new coding schemes that are built primarily upon deep networks (deep schemes), and deep network-based coding tools (deep tools) that shall be used within traditional coding schemes or together with traditional coding tools. For deep schemes, pixel probability modeling and auto-encoder are the two approaches, that can be viewed as predictive coding scheme and transform coding scheme, respectively. For deep tools, there have been several proposed techniques using deep learning to perform intra-picture prediction, inter-picture prediction, cross-channel prediction, probability distribution prediction, transform, post- or in-loop filtering, down- and up-sampling, as well as encoding optimizations. In the hope of advocating the research of deep learning-based video coding, we present a case study of our developed prototype video codec, namely Deep Learning Video Coding (DLVC). DLVC features two deep tools that are both based on convolutional neural network (CNN), namely CNN-based in-loop filter (CNN-ILF) and CNN-based block adaptive resolution coding (CNN-BARC). Both tools help improve the compression efficiency by a significant margin. With the two deep tools as well as other non-deep coding tools, DLVC is able to achieve on average 39.6\% and 33.0\% bits saving than HEVC, under random-access and low-delay configurations, respectively. The source code of DLVC has been released for future researches.

Extended Berkeley Packet Filter (eBPF) is an instruction set and an execution environment inside the Linux kernel. It enables modification, interaction and kernel programmability at runtime. eBPF can be used to program the eXpress Data Path (XDP), a kernel network layer that processes packets closer to the NIC for fast packet processing. Developers can write programs in C or P4 languages and then compile to eBPF instructions, which can be processed by the kernel or by programmable devices (e.g. SmartNICs). Since its introduction in 2014, eBPF has been rapidly adopted by major companies such as Facebook, Cloudflare, and Netronome. Use cases include network monitoring, network traffic manipulation, load balancing, and system profiling. This work aims to present eBPF to an inexpert audience, covering the main theoretical and fundamental aspects of eBPF and XDP, as well as introducing the reader to simple examples to give insight into the general operation and use of both technologies., All code in this paper was tested using kernel version 5.0. GitHub with step-by-step instructions on how to compile, load and run each example shown throughout this text, including a VM with all tools and dependencies necessary to develop eBPF programs are available on https://github.com/racyusdelanoo/bpf-tutorial.

Video description is the automatic generation of natural language sentences that describe the contents of a given video. It has applications in human-robot interaction, helping the visually impaired and video subtitling. The past few years have seen a surge of research in this area due to the unprecedented success of deep learning in computer vision and natural language processing. Numerous methods, datasets and evaluation metrics have been proposed in the literature, calling the need for a comprehensive survey to focus research efforts in this flourishing new direction. This paper fills the gap by surveying the state of the art approaches with a focus on deep learning models; comparing benchmark datasets in terms of their domains, number of classes, and repository size; and identifying the pros and cons of various evaluation metrics like SPICE, CIDEr, ROUGE, BLEU, METEOR, and WMD. Classical video description approaches combined subject, object and verb detection with template based language models to generate sentences. However, the release of large datasets revealed that these methods can not cope with the diversity in unconstrained open domain videos. Classical approaches were followed by a very short era of statistical methods which were soon replaced with deep learning, the current state of the art in video description. Our survey shows that despite the fast-paced developments, video description research is still in its infancy due to the following reasons. Analysis of video description models is challenging because it is difficult to ascertain the contributions, towards accuracy or errors, of the visual features and the adopted language model in the final description. Existing datasets neither contain adequate visual diversity nor complexity of linguistic structures. Finally, current evaluation metrics ..., Comment: Accepted by ACM Computing Surveys

Anomaly detection has numerous applications in diverse fields. For example, it has been widely used for discovering network intrusions and malicious events. It has also been used in numerous other applications such as identifying medical malpractice or credit fraud. Detection of anomalies in quantitative data has received a considerable attention in the literature and has a venerable history. By contrast, and despite the widespread availability use of categorical data in practice, anomaly detection in categorical data has received relatively little attention as compared to quantitative data. This is because detection of anomalies in categorical data is a challenging problem. Some anomaly detection techniques depend on identifying a representative pattern then measuring distances between objects and this pattern. Objects that are far from this pattern are declared as anomalies. However, identifying patterns and measuring distances are not easy in categorical data compared with quantitative data. Fortunately, several papers focussing on the detection of anomalies in categorical data have been published in the recent literature. In this article, we provide a comprehensive review of the research on the anomaly detection problem in categorical data. Previous review articles focus on either the statistics literature or the machine learning and computer science literature. This review article combines both literatures. We review 36 methods for the detection of anomalies in categorical data in both literatures and classify them into 12 different categories based on the conceptual definition of anomalies they use. For each approach, we survey anomaly detection methods, and then show the similarities and differences among them. We emphasize two important issues, the number of parameters each method requires and its time complexity. The first issue is critical, because the performance of these methods are sensitive to the choice of these parameters. The time complexity is also very important in real applications especially in big data applications. We report the time complexity if it is reported by the authors of the methods. If it is not, then we derive it ourselves and report it in this article. In addition, we discuss the common problems and the future directions of the anomaly detection in categorical data.

The use of machine learning (ML) in smart building applications is reviewed in this article. We split existing solutions into two main classes: occupant-centric versus energy/devices-centric. The first class groups solutions that use ML for aspects related to the occupants, including (1) occupancy estimation and identification, (2) activity recognition, and (3) estimating preferences and behavior. The second class groups solutions that use ML to estimate aspects related either to energy or devices. They are divided into three categories: (1) energy profiling and demand estimation, (2) appliances profiling and fault detection, and (3) inference on sensors. Solutions in each category are presented, discussed, and compared; open perspectives and research trends are discussed as well. Compared to related state-of-the-art survey papers, the contribution herein is to provide a comprehensive and holistic review from the ML perspectives rather than architectural and technical aspects of existing building management systems. This is by considering all types of ML tools, buildings, and several categories of applications, and by structuring the taxonomy accordingly. The article ends with a summary discussion of the presented works, with focus on lessons learned, challenges, open and future directions of research in this field. © ACM, 2019. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published here, https://doi.org/10.1145/3311950

The world is undergoing an unprecedented technological transformation, evolving into a state where ubiquitous Internet-enabled “things” will be able to generate and share large amounts of security- and privacy-sensitive data. To cope with the security threats that are thus foreseeable, system designers can find in Arm TrustZone hardware technology a most valuable resource. TrustZone is a System-on-Chip and CPU system-wide security solution, available on today’s Arm application processors and present in the new generation Arm microcontrollers, which are expected to dominate the market of smart “things.” Although this technology has remained relatively underground since its inception in 2004, over the past years, numerous initiatives have significantly advanced the state of the art involving Arm TrustZone. Motivated by this revival of interest, this paper presents an in-depth study of TrustZone technology. We provide a comprehensive survey of relevant work from academia and industry, presenting existing systems into two main areas, namely, Trusted Execution Environments and hardware-assisted virtualization. Furthermore, we analyze the most relevant weaknesses of existing systems and propose new research directions within the realm of tiniest devices and the Internet of Things, which we believe to have potential to yield high-impact contributions in the future.

Legacy encryption systems depend on sharing a key (public or private) among the peers involved in exchanging an encrypted message. However, this approach poses privacy concerns. Especially with popular cloud services, the control over the privacy of the sensitive data is lost. Even when the keys are not shared, the encrypted material is shared with a third party that does not necessarily need to access the content. Moreover, untrusted servers, providers, and cloud operators can keep identifying elements of users long after users end the relationship with the services. Indeed, Homomorphic Encryption (HE), a special kind of encryption scheme, can address these concerns as it allows any third party to operate on the encrypted data without decrypting it in advance. Although this extremely useful feature of the HE scheme has been known for over 30 years, the first plausible and achievable Fully Homomorphic Encryption (FHE) scheme, which allows any computable function to perform on the encrypted data, was introduced by Craig Gentry in 2009. Even though this was a major achievement, different implementations so far demonstrated that FHE still needs to be improved significantly to be practical on every platform. First, we present the basics of HE and the details of the well-known Partially Homomorphic Encryption (PHE) and Somewhat Homomorphic Encryption (SWHE), which are important pillars of achieving FHE. Then, the main FHE families, which have become the base for the other follow-up FHE schemes are presented. Furthermore, the implementations and recent improvements in Gentry-type FHE schemes are also surveyed. Finally, further research directions are discussed. This survey is intended to give a clear knowledge and foundation to researchers and practitioners interested in knowing, applying, as well as extending the state of the art HE, PHE, SWHE, and FHE systems., Comment: - Updated. (October 6, 2017) - This paper is an early draft of the survey that is being submitted to ACM CSUR and has been uploaded to arXiv for feedback from stakeholders

The goal of this paper is to illustrate how frameworks and patterns address complexities that arise in the design and implementation of high-performance distributed software systems. These complexities are both inherent (e.g., latency reduction and throughput preservation), and accidental (e.g., the continuous reinvention of key concepts and components). This paper explains how complexities occurring in the development of high-performance Web servers can be alleviated with the use of design patterns and object-oriented application frameworks. These techniques were applied to the development our high-performance adaptive Web server framework, JAWS. JAWS exemplifies how a framework can remain flexible without sacrificing performance. 1 Applying Patterns and Frameworks to Web Servers Developers of Web servers strive to build fast, scalable, and configurable systems. This paper describes some common pitfalls encountered by these developers and how to avoid these pitfalls. Common pitfalls include (1) coping with tedious and error-prone low-level programming details, (2) lack of portability, and (3) the complexity of navigating the wide range of server design alternatives. By carefully utilizing patterns and frameworks, these hazards can be avoided, by allowing developers to leverage reuse of design and code. 1.1 Common Pitfalls of Developing Web Server Software Web servers perform the following tasks: connection establishment, service initialization, event demultiplexing, event handler dispatching, interprocess communication, memory management and file caching, static and dynamic component configuration, concurrency, synchronization, and persistence. In most Web servers, these tasks are implemented in an ad hoc manner using low-level native OS application programming interfaces (APIs), such as Win32 or UNIX/POSIX, which are written in C. Unfortunately, native OS APIs are not an effective way to develop Web servers or other types of communication middleware and applications [1]. The following are common pitfalls associated with the use of native OS APIs: Excessive low-level details: Building Web servers with native OS APIs requires developers to have intimate knowledge of low-level OS details. Developers must carefully track which error codes are returned by each system call and handle these OS-specific problems in their servers. Such details divert attention from the broader, more strategic issues, such as protocol semantics and server structure. For example, UNIX developers who use the wait system call must distinguish between return errors due to no child processes being present and errors from signal interrupts. In the latter case, the wait must be reissued. Reinvention of incompatible programming abstractions: A common remedy for the excessive level of detail with OS APIs is to define higher-level programming abstractions. For instance, many Web servers create a file cache to avoid accessing the filesystem for each client request. However, these types of abstractions are often rediscovered and reinvented independently by each developer or project. This ad hoc devel

This paper presents a summary of the results of several relatively large studies which attempted statistical analysis of audit trails created by learners accessing information in typical hypermedia or hypertext learning environments, and interpreted them in relation to learner characteristics and study tasks. Significant differences in the information access strategy, amount of information accessed, student estimates of achievement and knowledge outcome were observed between learners in these studies. This paper concluded that some learners may be systematically disadvantaged where support for (or the delivery of) the curriculum depends on hypermedia, such as via a networked learning environment delivered passively over the WWW. It is suggested that the audit tools available from the WWW provide an opportunity to develop multi-discipline evaluation mechanisms which may enable researchers to provide learners with standard "learning profiles" with which to reflect on their own learning effectiveness when using hypermedia educational materials.

The vulnerability of face recognition systems to presentation attacks (also known as direct attacks or spoof attacks) has received a great deal of interest from the biometric community. The rapid evolution of face recognition systems into real-time applications has raised new concerns about their ability to resist presentation attacks, particularly in unattended application scenarios such as automated border control. The goal of a presentation attack is to subvert the face recognition system by presenting a facial biometric artifact. Popular face biometric artifacts include a printed photo, the electronic display of a facial photo, replaying video using an electronic display, and 3D face masks. These have demonstrated a high security risk for state-of-the-art face recognition systems. However, several presentation attack detection (PAD) algorithms (also known as countermeasures or antispoofing methods) have been proposed that can automatically detect and mitigate such targeted attacks. The goal of this survey is to present a systematic overview of the existing work on face presentation attack detection that has been carried out. This paper describes the various aspects of face presentation attacks, including different types of face artifacts, state-of-the-art PAD algorithms and an overview of the respective research labs working in this domain, vulnerability assessments and performance evaluation metrics, the outcomes of competitions, the availability of public databases for benchmarking new PAD algorithms in a reproducible manner, and finally a summary of the relevant international standardization in this field. Furthermore, we discuss the open challenges and future work that need to be addressed in this evolving field of biometrics.

The security research community has invested significant effort in improving the security of Android applications over the past half decade. This effort has addressed a wide range of problems and resulted in the creation of many tools for application analysis. In this article, we perform the first systematization of Android security research that analyzes applications, characterizing the work published in more than 17 top venues since 2010. We categorize each paper by the types of problems they solve, highlight areas that have received the most attention, and note whether tools were ever publicly released for each effort. Of the released tools, we then evaluate a representative sample to determine how well application developers can apply the results of our community’s efforts to improve their products. We find not only that significant work remains to be done in terms of research coverage but also that the tools suffer from significant issues ranging from lack of maintenance to the inability to produce functional output for applications with known vulnerabilities. We close by offering suggestions on how the community can more successfully move forward.

Cloud log forensics (CLF) mitigates the investigation process by identifying the malicious behavior of attackers through profound cloud log analysis. However, the accessibility attributes of cloud logs obstruct accomplishment of the goal to investigate cloud logs for various susceptibilities. Accessibility involves the issues of cloud log access, selection of proper cloud log file, cloud log data integrity, and trustworthiness of cloud logs. Therefore, forensic investigators of cloud log files are dependent on cloud service providers (CSPs) to get access of different cloud logs. Accessing cloud logs from outside the cloud without depending on the CSP is a challenging research area, whereas the increase in cloud attacks has increased the need for CLF to investigate the malicious activities of attackers. This paper reviews the state of the art of CLF and highlights different challenges and issues involved in investigating cloud log data. The logging mode, the importance of CLF, and cloud log-as-a-service are introduced. Moreover, case studies related to CLF are explained to highlight the practical implementation of cloud log investigation for analyzing malicious behaviors. The CLF security requirements, vulnerability points, and challenges are identified to tolerate different cloud log susceptibilities. We identify and introduce challenges and future directions to highlight open research areas of CLF for motivating investigators, academicians, and researchers to investigate them.

The following paper by Kernighan and Plauger is one of three chosen from the December 1974 issue of ACM Computing Surveys. It's considerably shorter and less theoretical than the companion papers by Knuth [Paper 20] and Wirth [Paper 13]; indeed, it's less theoretical and more down to earth than most of the papers in this book! For that reason, many readers will view it as a breath of fresh air, and will rejoice at the presence of real FORTRAN examples and real PL/I examples. The paper is largely excerpted from Kernighan and Plauger's first book, The Elements of Programming Style, providing a nice, concise, 21-page summary that you can read in less than an hour. One of the themes of this paper is that structured programming is, in a sense, a secondary issue; the primary concern of programming, according to the authors, is style. The elements of programming style consist of such things as expression (organizing individual statements so that they read clearly), structure (organizing larger blocks of code so that the program "hangs together"), robustness (writing code that can "defend itself against bad data from the outside world"), and, finally, efficiency. As I've said, there are examples to illustrate these elements of programming style --- examples that are "real," from the kind of programs that one would expect to find in an actual scientific or business-oriented EDP shop. Indeed, the examples are real, but in a very special sense: Kernighan and Plauger have taken all of their examples verbatim from other programming textbooks. Although the examples don't include any COBOL or assembler code, the enterprising reader can generalize from the FORTRAN examples so as to apply the lessons to his own work. There is one other theme in this paper, one that I think is particularly important in these days of elegant programming languages like ALGOL and PASCAL. Rather than trying to restate Kernighan and Plauger's point, let me quote them directly: " . . . many people try to excuse badly written programs by blaming inadequacies of the language that must be used. We have seen repeatedly that even Fortran can be tamed with proper discipline. The presence of bad features is not an invitation to use them, nor is .the absence of good features an excuse to avoid simulating them as cleanly as possible. Good languages are nice, but not vital." FORTRAN and COBOL programmers, take heed!

The rising abuse of computers and increasing threat to personal privacy through data banks have stimulated much interest in the technical safeguards for data. There are four kinds of safeguards, each related to but distinct from the others. Access controls regulate which users may enter the system and subsequently which data sets an active user may read or write. Flow controls regulate the dissemination of values among the data sets accessible to a user. Inference controls protect statistical databases by preventing questioners from deducing confidential information by posing carefully designed sequences of statistical queries and correlating the responses. Statistical data banks are much less secure than most people believe. Data encryption attempts to prevent unauthorized disclosure of confidential information in transit or in storage. This paper describes the general nature of controls of each type, the kinds of problems they can and cannot solve, and their inherent limitations and weaknesses. The paper is intended for a general audience with little background in the area.

This paper presents a systematic approach to matching categories of query language interfaces with the requirements of certain user types. The method is based on a trend model of query language development on the dimensions of functional capabilities and usability. From the trend model the following are derived: a classification scheme for query languages, a criterion hierarchy for query language evaluation, a comprehensive classification scheme of query language users and their requirements, and preliminary recommendations for allocating language classes to user types. The method integrates the results of existing human factors studies and provides a structured framework for future research in this area. Current and expected developments are exemplified by the description of "new generation" database query languages. In a practical query language selection problem, the results of this paper can be used for preselecting suitable query language types; the final selection decision will also depend on organization-specific factors, such as the available database management system, hardware and software strategies, and financial system costs.

Semantic data models have emerged from a requirement for more expressive conceptual data models. Current generation data models lack direct support for relationships, data abstraction, inheritance, constraints, unstructured objects, and the dynamic properties of an application. Although the need for data models with richer semantics is widely recognized, no single approach has won general acceptance. This paper describes the generic properties of semantic data models and presents a representative selection of models that have been proposed since the mid-1970s. In addition to explaining the features of the individual models, guidelines are offered for the comparison of models. The paper concludes with a discussion of future directions in the area of conceptual data modeling.

Much research has recently been devoted to "multikey" searching problems. In this paper the partmular multlkey problem of range searching Is investigated and a number of data structures that have been proposed as solutions to this problem are surveyed. The purposes of this paper are to bring together a collection of widely scattered results, to acquaint the reader with the structures currently avadable for solving the particular problem of range searching, and to display a set of general methods for attacking multikey searching problems.

The alpha-beta algorithm forms the basis of many programs that search game trees. A number of methods have been designed to improve the utility of the sequential version of this algorithm, especially for use in game-playing programs. These enhancements are based on the observation that alpha beta is most effective when the best move in each position is considered early in the search. Trees that have this so-called strong ordering property are not only of practical importance but possess characteristics that can be exploited in both sequential and parallel environments. This paper draws upon experiences gained during the development of programs which search chess game trees. Over the past decade major enhancements of the alpha beta algorithm have been developed by people building game-playing programs, and many of these methods will be surveyed and compared here. The balance of the paper contains a study of contemporary methods for searching chess game trees in parallel, using an arbitrary number of independent processors. To make efficient use of these processors, one must have a clear understanding of the basic properties of the trees actually traversed when alpha-beta cutoffs occur. This paper provides such insights and concludes with a brief description of amore » refinement to a standard parallel search algorithm for this problem. 33 references.« less

In this paper, a terminological framework is provided for describing different transactionoriented recovery schemes for database systems in a conceptual rather than an implementation-dependent way. By introducing the terms materialized database, propagation strategy, and checkpoint, we obtain a means for classifying arbitrary implementations from a unified viewpoint. This is complemented by a classification scheme for logging techniques, which are precisely defined by using the other terms. It is shown that these criteria are related to all relevant questions such as speed and scope of recovery and amount of redundant information required. The primary purpose of this paper, however, is to establish an adequate and precise terminology for a topic in which the confusion of concepts and implementational aspects still imposes a lot of problems.

Concurrent logic languages are high-level programming languages for parallel and distributed systems that offer a wide range of both known and novel concurrent programming techniques. Being logic programming languages, they preserve many advantages of the abstract logic programming model, including the logical reading of programs and computations, the convenience of representing data structures with logical terms and manipulating them using unification, and the amenability to metaprogramming. Operationally, their model of computation consists of a dynamic set of concurrent processes, communicating by instantiating shared logical variables, synchronizing by waiting for variables to be instantiated, and making nondeterministic choices, possibly based on the availability of values of variables. This paper surveys the family of concurrent logic programming languages within a uniform operational framework. It demonstrates the expressive power of even the simplest language in the family and investigates how varying the basic synchronization and control constructs affect the expressiveness and efficiency of the resulting languages. In addition, the paper reports on techniques for sequential and parallel implementation of languages in this family, mentions their applications to date, and relates these languages to the abstract logic programming model, to the programming language PROLOG, and to other concurrent computational models and programming languages.

This paper reviews empirical studies concerning the readability of text from computer screens. The review focuses on the form and physical attributes of complex, realistic displays of text material. Most studies comparing paper and computer screen readability show that screens are less readable than paper. There are many factors that could affect the readability of computer screens. The factors explored in this review are the features of characters, the formatting of the screen, the contrast and color of the characters and background, and dynamic aspects of the screen. Numerous areas for future research are pinpointed.

This paper is a study of the unique problems posed by the use of computers by composers and performers of music. The paper begins with a presentation of the basic concepts involved in the musical interaction with computer devices, followed by a detailed discussion of three musical tasks: music manuscript preparation, music language interfaces for composition, and real-time performance interaction. Fundamental design principles are exposed through an examination of several early computer music systems, especially the Structured Sound Synthesis Project. A survey of numerous systems, based on the following categories, is presented: compositions and synthesis languages, graphics score editing, performance instruments, digital audio processing tools, and computer-aided instruction in music systems. An extensive reference list is provided for further study in the field.

Explanation-based learning (EBL) is a technique by which an intelligent system can learn by observing examples. EBL systems are characterized by the ability to create justified generalizations from single training instances. They are also distinguished by their reliance on background knowledge of the domain under study. Although EBL is usually viewed as a method for performing generalization, it can be viewed in other ways as well. In particular, EBL can be seen as a method that performs four different learning tasks: generalization, chunking, operationalization, and analogy. This paper provides a general introduction to the field of explanation-based learning. Considerable emphasis is placed on showing how EBL combines the four learning tasks mentioned above. The paper begins with a presentation of an intuitive example of the EBL technique. Subsequently EBL is placed in its historical context and the relation between EBL and other areas of machine learning is described. The major part of this paper is a survey of selected EBL programs, which have been chosen to show how EBL manifests each of the four learning tasks. Attempts to formalize the EBL technique are also briefly discussed. The paper concludes with a discussion of the limitations of EBL and the major open questions in the field.

Object-oriented programming is becoming a popular approach to the construction of complex software systems. Benefits of object orientation include support for modular design, code sharing, and extensibility. In order to make the most of these advantages, a type theory for objects and their interactions should be developed to aid checking and controlled derivation of programs and to support early binding of code bodies for efficiency. As a step in this direction, this paper surveys a number of existing type theories and examines the manner and extent to which these theories are able to represent the ideas found in object-oriented programming. Of primary interest are the models provided by type theories for abstract data types and inheritance, and the major portion of this paper is devoted to these topics. Code fragments illustrative of the various approaches are provided and discussed. The introduction provides an overview of object-oriented programming and types in programming languages; the summary provides a comparative evaluation of the reviewed typing systems, along with suggestions for future work.

The purpose of this paper is to show that logic provides a convenient formalism for studying classical database problems. There are two main parts to the paper, devoted respectively to conventional databases and deductive databases. In the first part, we focus on query languages, integrity modeling and maintenance, query optimization, and data dependencies. The second part deals mainly with the representation and manipulation of deduced facts and incomplete information. Categories and Subject Descriptors: H.2.1 [Database Management]: Logical Design— data models; H.2.3 [Database Management]: Languages— query languages; H.2.4 [Database Management]: Systems— query processing General Terms: Deductive Databases, Indefinite Data, Logic and Databases, Null Values, Relational Databases

Hypervideos and interactive multimedia presentations allow the creation of fully interactive and enriched video. It is possible to organize video scenes in a nonlinear way. Additional information can be added to the video ranging from short descriptions to images and more videos. Hypervideos are video-based but also provide navigation between video scenes and additional multimedia elements. Interactive multimedia presentations consist of different media with a temporal and spatial synchronization that can be navigated via hyperlinks. Their creation and description requires description formats, multimedia models, and standards—as well as players. Specialized authoring tools with advanced editing functions allow authors to manage all media files, link and arrange them to an overall presentation, and keep an overview during the whole process. They considerably simplify the creation process compared to writing and editing description documents in simple text editors. Data formats need features that describe interactivity and nonlinear navigation while maintaining temporal and spatial synchronization. Players should be easy to use with extended feature sets keeping elements synchronized. In this article, we analyzed more than 400 papers for relevant work in this field. From the findings we discovered a set of trends and unsolved problems, and propose directions for future research.

Web service is a widely used implementation technique under the paradigm of Service-Oriented Architecture (SOA). A service-based system is subjected to continuous evolution and regression testing is required to check whether new faults have been introduced. Based on the current scientific work of web service regression testing, this survey aims to identify gaps in current research and suggests some promising areas for further study. To this end, we performed a broad automatic search on publications in the selected electronic databases published from 2000 to 2013. Through our careful review and manual screening, a total of 30 papers have been selected as primary studies for answering our research questions. We presented a qualitative analysis of the findings, including stakeholders, challenges, standards, techniques, and validations employed in these primary studies. Our main results include the following: (1) Service integrator is the key stakeholder that largely impacts how regression testing is performed. (2) Challenges of cost and autonomy issues have been studied heavily. However, more emphasis should be put on the other challenges, such as test timing, dynamics, privacy, quota constraints, and concurrency issues. (3) Orchestration-based services have been largely studied, while little attention has been paid to either choreography-based services or semantic-based services. (4) An appreciable amount of web service regression testing techniques have been proposed, including 48 test case prioritization techniques, 10 test selection techniques, two test suite minimization techniques, and another collaborative technique. (5) Many regression test techniques have not been theoretically proven or experimentally analyzed, which limits their application in large-scale systems. We believe that our survey has identified gaps in current research work and reveals new insights for the future work.

International audience; The adequacy of a programming language to a given software project or application domain is often considered a key factor of success in software development and engineering, even though little theoretical or practical information is readily available to help make an informed decision. In this paper, we address a particular version of this issue by comparing the adequacy of generalpurpose synchronous programming languages to more domain-specific languages (DSL) in the field of computer music. More precisely, we implemented and tested the same lookup table oscillator example program, one of the most classical algorithms for sound synthesis, using a selection of significant synchronous programming languages, half of which designed as specific music languages -Csound, Pure Data, SuperCollider, ChucK, Faust - and the other half being general synchronous formalisms - Signal, Lustre, Esterel, Lucid Synchrone and C with the OpenMP Stream Extension (Matlab/Octave is used for the initial specification). The advantages of both approaches are discussed, providing practical insights to both software developers and language designers regarding the choice of programming language styles when tackling audio applications.

Networking research often relies on simulation in order to test and evaluate new ideas. An important requirement of this process is that results must be reproducible so that other researchers can replicate, validate, and extend existing work. We look at the landscape of simulators for research in peer-to-peer (P2P) networks by conducting a survey of a combined total of over 280 papers from before and after 2007 (the year of the last survey in this area), and comment on the large quantity of research using bespoke, closed-source simulators. We propose a set of criteria that P2P simulators should meet, and poll the P2P research community for their agreement. We aim to drive the community towards performing their experiments on simulators that allow for others to validate their results.

The goal is to survey dependability modeling and analysis of software and systems specified with UML, with focus on reliability, availability, maintainability, and safety (RAMS). From the literature published in the last decade, 33 approaches presented in 43 papers were identified. They are evaluated according to three sets of criteria regarding UML modeling issues, addressed dependability characteristics, and quality assessment of the surveyed approaches. The survey shows that more works are devoted to reliability and safety, fewer to availability and maintainability, and none to integrity. Many methods support early life-cycle phases (from requirements to design). More research is needed for tool development to automate the derivation of analysis models and to give feedback to designers.

Trust management systems are frameworks for authorization in modern distributed systems, allowing remotely accessible resources to be protected by providers. By allowing providers to specify policy, and access requesters to possess certain access rights, trust management automates the process of determining whether access should be allowed on the basis of policy, rights, and an authorization semantics. In this paper we survey modern state-of-the-art in trust management authorization, focusing on features of policy and rights languages that provide the necessary expressiveness for modern practice. We characterize systems in light of a generic structure that takes into account components of practical implementations. We emphasize systems that have a formal foundation, since security properties of them can be rigorously guaranteed. Underlying formalisms are reviewed to provide necessary background.