基于PAR 的排序算法自动生成研究

排序是计算机学科中的一类特殊问题,其算法设计策略的灵活性使得求解算法更具多样性.基于形式化方法PAR(partition-and-recur),研究了排序算法的自动生成问题.刻画了排序问题的代数性质,形式化构建了排序算法领域的泛型类型构件和算法构件,建立了排序领域特定语言和算法生成形式化模型,以参数替换的方式自动生成了一组排序算法,包括快速排序、堆排序、Shell排序等典型的已知算法以及增量选择排序等若干未见于现有文献的算法,并在程序生成系统中予以了实现.通过上层框架研究和底层构件支持,显著提高了特定领域算法的开发效率和可靠性.     

向量加法系统验证问题研究综述

Petri网是形式化验证领域最重要的模型之一,具有重要的理论和应用价值.从验证算法分析的角度Petri网可以被等价地抽象为"向量加法系统".在对向量加法模型的研究中人们又发展了一些重要的扩展模型.本文对近些年来国内外学者在向量加法系统验证领域取得的成果进行了系统总结.首先,给出了向量加法系统及几个关键验证问题的形式化定义,并重点总结了一般向量加法系统模型上可达性问题的最新研究进展和关键技术;接着,总结了当限定模型的维度为固定值时相关研究进展,重点给出了2维情况的核心定理;随后,介绍了几个重要扩展模型,并总结了这些模型上验证问题研究的最新进展.在每一部分都对未来研究方向及可能面临的挑战进行了展望.     

基于用户浏览行为的个性化推荐研究综述

研究如何充分利用海量用户浏览行为数据,构建更加精确的推荐算法和模型,以提高推荐系统性能,是目前个性化推荐领域研究的热点.针对这些问题,首先对用户的浏览行为进行了简要概括表述,给出了基于浏览行为推荐系统的总体框架,回顾总结了基于用户浏览行为的推荐系统的发展历程.对其关键技术和单一浏览行为量化方法与混合浏览行为量化方法进行总结、对比和分析.最后讨论了结合多源异构数据的浏览行为推荐的最新成果,总结了该领域未来研究难点和发展趋势.     

协同过滤推荐研究综述

推荐技术是目前在很多领域中广泛使用的技术之一.而协同过滤推荐算法是应用在推荐技术中很成功的算法.主要介绍了协同过滤推荐技术,总结了当前推荐算法的传统方法、改进算法以及性能评测方法.同时,分析了协同过滤推荐算法中的问题以及相应的解决办法.最后阐述了协同过滤推荐系统中仍需解决的问题和未来可能的发展方向.     

协同过滤中相似度算法研究进展

推荐算法通过历史数据发现用户的兴趣偏好,在数据资源中寻找用户的偏好信息,并对用户进行推荐。目前,推荐系统中的协同过滤算法在各领域应用广泛,由于数据稀疏性和冷启动,使得推荐质量有所下降,为提升推荐精度,有学者从相似度方向进行研究。总结了推荐系统中最广泛使用的协同过滤算法,以及推荐系统中常用的传统相似度算法;对比分析了基于Pearson相关系数的相似度、余弦相似度、修正的余弦相似度等的适用场景;从冷启动和数据稀疏等方面分析了相似度的研究现状,研究表明通过混合相似度计算用户相似性,提高了推荐质量。最后,总结了相关文献在改进后存在推荐效率低、复杂度增高的问题,在提高推荐精度和推荐效率方面对相似度改进进行了展望。     

基于关联规则算法的推荐方法研究综述

目前传统推荐算法已经被广泛应用于各类推荐系统,然而在推荐过程中仍然存在着无法处理非结构化数据、数据潜在关系发现困难、数据稀疏和冷启动等问题。关联规则技术的出现有效缓解了这些问题,推荐效率也因此得到提高。将关联规则技术的特殊属性与推荐算法进行高质量的结合成为推荐领域的研究热点。通过综述关联规则技术与数据的不同关联规则类别在传统推荐算法中的应用,对传统算法在推荐过程中的优缺点进行了归纳阐述。针对基于关联规则推荐算法的研究进行总结,并对其未来的发展趋势进行展望。     

推荐系统综述

随着网络应用的不断发展,网络资源呈指数型增长,信息过载现象日益严重,如何高效获取符合需求的资源成为困扰人们的问题之一。推荐系统能对海量信息进行有效过滤,为用户推荐符合其需求的资源。对推荐系统的研究现状进行详细介绍,包括基于内容的推荐、协同过滤推荐和混合推荐这三种传统推荐方式,并重点分析了基于卷积神经网络（CNN）、深度神经网络（DNN）、循环神经网络（RNN）和图神经网络（GNN）这四种常见的深度学习推荐模型的研究进展;归纳整理了推荐领域常用的数据集,同时分析对比了传统推荐算法和基于深度学习的推荐算法的差异。最后,总结了实际应用中具有代表性的推荐模型,讨论了推荐系统面临的挑战和未来的研究方向。     

社会化推荐系统研究

近年来,社会化推荐系统已成为推荐系统研究领域较为活跃的研究方向之一.如何利用用户社会属性信息缓解推荐系统中数据稀疏性和冷启动问题、提高推荐系统的性能,成为社会化推荐系统的主要任务.对最近几年社会化推荐系统的研究进展进行综述,对信任推理算法、推荐关键技术及其应用进展进行前沿概括、比较和分析.最后,对社会化推荐系统中有待深入研究的难点、热点及发展趋势进行展望.     

基于强化学习的推荐研究综述

推荐系统致力于从海量数据中为用户寻找并自动推荐有价值的信息和服务,可有效解决信息过载问题,成为大数据时代一种重要的信息技术.但推荐系统的数据稀疏性、冷启动和可解释性等问题,仍是制约推荐系统广泛应用的关键技术难点.强化学习是一种交互学习技术,该方法通过与用户交互并获得反馈来实时捕捉其兴趣漂移,从而动态地建模用户偏好,可以较好地解决传统推荐系统面临的经典关键问题.强化学习已成为近年来推荐系统领域的研究热点.文中从综述的角度,首先在简要回顾推荐系统和强化学习的基础上,分析了强化学习对推荐系统的提升思路,对近年来基于强化学习的推荐研究进行了梳理与总结,并分别对传统强化学习推荐和深度强化学习推荐的研究情况进行总结;在此基础上,重点总结了近年来强化学习推荐研究的若干前沿,以及其应用研究情况.最后,对强化学习在推荐系统中应用的未来发展趋势进行分析与展望.     

安全强化学习算法及其在CPS智能控制中的应用

信息物理系统(cyber-physicalsystem,CPS)的安全控制器设计是一个热门研究方向,现有基于形式化方法的安全控制器设计存在过度依赖模型、可扩展性差等问题.基于深度强化学习的智能控制可处理高维非线性复杂系统和不确定性系统,正成为非常有前景的CPS控制技术,但是缺乏对安全性的保障.针对强化学习控制在安全性方面的不足,围绕一个工业油泵控制系统典型案例,开展安全强化学习算法和智能控制应用研究.首先,形式化了工业油泵控制的安全强化学习问题,搭建了工业油泵仿真环境;随后,通过设计输出层结构和激活函数,构造了神经网络形式的油泵控制器,使得油泵开关时间的线性不等式约束得到满足;最后,为了更好地权衡安全性和最优性控制目标,基于增广拉格朗日乘子法设计实现了新型安全强化学习算法.在工业油泵案例上的对比实验表明,该算法生成的控制器在安全性和最优性上均超越了现有同类算法.在进一步评估中,所生成神经网络控制器以90%的概率通过了严格形式化验证;同时,与理论最优控制器相比实现了低至2%的最优目标值损失.所提方法有望推广至更多应用场景,实例研究的方案有望为安全智能控制和形式化验证领域其他学者提供借鉴.     

Analyzing of research patterns based on a temporal tracking and assessing model

Scientific research works conducted by researchers spread all over the world in every research field, which are hard to be tracked and quantified. Although there are many research works focused on scientific community discovery and researcher profiling, it is still a big challenge to track the research patterns and assess the research development for an individual researcher or a research group over time. In this study, we seek to model researchers’ scientific activities and quantify their outcome during their research career. A temporal tracking and assessing model is introduced to represent the research development and quantify the scientific outcome for both an individual and a group along the time. Based on our model, a research topic analyzing approach is developed to extract the topics covered by a research group for the research pattern analysis. Furthermore, a latent research pattern discovering approach is proposed to depict how a research group’s research works contributed by its members are discovered and visualized. The effectiveness of our approach is evaluated based on a real academic dataset.     

Interpretation,intervention, and reduction in the organizational laboratory: a framework for in-context information system research

Understanding how technical artefacts are created and used within organizations is a central aspect of the IS research discipline. The conduct of research in an organizational setting is thus a major issue for the IS community. A research framework for in-context IS research is presented and used to position purified and hybrid forms of research method. From the framework theoretical support for an `action case' research method is presented. Two IS research cases are presented and analyzed using the IS research framework, leading to a practice-based rationale for an action case method. Characteristics of the action case method, a hybrid of interpretation and intervention, are described. Learning at three levels of analysis—concrete, general, and meta—is proposed and used as a way of reflecting on IS research methods and IS research frameworks.     

代码坏味研究综述

代码坏味(code smells)是低质量的急需重构的代码片段.代码坏味是软件工程领域的一个研究热点,并且相关研究方向众多、时间跨度大、研究成果丰富.为梳理相关研究思路和研究成果、分析研究热点并预判未来研究方向,对1990年至2020年6月间发表的代码坏味相关的339篇论文进行了系统地分析和归类,对代码坏味的发展趋势进行了分析与统计,量化揭示了相关研究的主流与热点.揭示了学术界关注的关键代码坏味,并研究了工业界与学术界的关注点的差异及其影响.     

A multi-method, holistic strategy for researching critical success factors in IT projects

In the field of information systems (IS) there is an observable trend towards the use of multi-method research. Using different research methods allows for the cross-validation of data obtained via multiple approaches, with the potential to increase the robustness of research results. Such a multi-method approach is applicable to a comprehensive research agenda on critical success factors, an agenda that needs to take into account not only the identification, but also the analysis and management of critical success factors. The goal of this article is to contribute new knowledge on how to carry out research on critical success factors in IS projects using a multi-method approach. For this purpose, two research projects are presented, each a variation of the research design customized to particular circumstances. First, there is an outline of the research approach taken for a critical success factor research project in the field of portal implementation, with discussion of the strengths and weaknesses of the project. Taking into consideration these experiences, the research approach of a similar critical success factor research project in the field of offshore software development is then described. Finally, recommendations for using the multi-method research approach in critical success factor research are presented.     

Joint software research between industry and academia

A model for industry-sponsored academic software research tha helps the sponsor gain tangible results quickly but respects the university's research mission is presented. Central to the model is joint effort to achieve a win-win situation for all participants. The key is the relationship between the university's principal investigator and the company's visiting scholar. The principal investigator initiates research contracts and is responsible for research results but, because he or she is an academic, may not have an industrial perspective. The company dispatches a technical representative, called the visiting scholar, to participate directly in the daily research. A research structure, relations with the sponsoring company, issues regarding environmental compatibility and funding amount and duration, and a reporting structure are all spelled out in the model. Legal recommendations are made with respect to confidentiality, copyrights and patents, export controls, and other issues. Other considerations that arise when writing a research contract are discussed     

基于角色的工作流模型的研究

工作流是近年来软件工程领域的一个热点研究问题，而基于角色的研究也是计算机领域里的热点研究问题。本文就是结合这个两个热点问题来研究，阐述了工作流和基于角色的基本概念和原理，并结合具体项目，给出了一个基于角色的工作流模型。     

基于角色的工作流模型的研究

工作流是近年来软件工程领域的一个热点研究问题,而基于角色的研究也是计算机领域里的热点研究问题。本文就是结合这个两个热点问题来研究,阐述了工作流和基于角色的基本概念和原理,并结合具体项目,给出了一个基于角色的工作流模型。     

Research in online learning environments: Priorities and methodologies

Due to increasing demand for education and training in the information age, online learning and teaching is becoming a necessity in our future. However, lack of research goals to understand impact of online learning environments on students is a problem in research on online learning environments. We identified four main research goals to pursue in online learning environments based on their impact on learner achievement, engagement, and retention (opposite of attrition). Those goals are (a) enhancing learner engagement & collaboration, (b) promoting effective facilitation, (c) developing assessment techniques, and (d) designing faculty development programs. Current promising work in those areas is presented. Four methods that are common in the instructional technology literature are recommended to pursue those goals. Formative research and developmental research are relevant for all four. Although appropriate for any of the goals, experimental research is a better fit for goals b and c, and activity theory is useful for goals a and b.     

RoboCup: Today and tomorrow—What we have learned

RoboCup is an increasingly successful attempt to promote the full integration of AI and robotics research. The most prominent feature of RoboCup is that it provides the researchers with the opportunity to demonstrate their research results as a form of competition in a dynamically changing hostile environment, defined as the international standard game definition, which the gamut of intelligent robotics research issues are naturally involved. This article describes what we have learned from the past RoboCup activities, mainly the first and the second RoboCups, and overviews the future perspectives of RoboCup in the next century. First, the issue on what and why RoboCup is addressed, and a wide range of research issues are explained. Next, the current leagues are introduced and the research achievements are reviewed from a viewpoint of system architecture. Some of these achievements are included in this special issue. Finally, prospects for future activities are discussed.     

Too close for comfort? Distance and engagement in interpretive information systems research

In this paper, we seek to explore the implications of the assumptions underlying interpretivism for the preferred relationship between the researcher and the research phenomenon. The growing interest in interpretive information systems research has drawn attention to the need to gain access to the interpretations of social actors. Various data-gathering methods are available to the researcher to achieve such access. These may be seen as located on a spectrum in terms of the degree of engagement between the researcher and the research subject. While engagement is not without its drawbacks from a research perspective, it may be argued that it enables good access to the sorts of data that interpretive researchers are seeking. Most of the research reported in the information systems (IS) field, however, has adopted relatively distant methods by which the researcher avoids intervention in the research context. Some of the characteristics of more engaged forms of research are illustrated through a discussion of a participant observation study of executive information systems development. This research has highlighted a number of issues that can be seen to be common to other data-gathering methods. Reasons for the neglect of engaged data-gathering methods are explored, and it is suggested that consideration of the issues it raises may contribute to more reflexive interpretive IS research practice.