软件度量技术在SSMBSS中的应用

上海软件度量基准体系(SSMBSS)针对上海市、全国甚至世界范围内的软件项目或产品开发过程中的各种数据进行收集、整合和分析以得到软件行业的各种度量指标,从而更有效的为软件企业的项目管理及政府部门的决策提供支持.因此度量内容的选取和对度量数据的分析是其中关键的一个环节,论文将对该环节进行具体分析,以期得到合理的度量内容和指示器.     

基于复杂度对开源软件Linux和FreeBSD内核的分析

开源软件内核进化分析是一个较新的研究方向,研究大型自由和开源软件进化可以帮助估计软件体系结构的稳定性,可以较准确地预报软件进化趋势和预报软件进化结果.选取Linux和FreeBSD的多个版本的内核进行数据的收集和处理.采用低层次的度量尺度,利用分析工具SLOCOUNT和CCCC,对提取的内核数据进行复杂度分析,得出开源自由软件进化的结论,并提出了快速测算大型开源软件进化趋势和复杂性的方法.     

支持数据挖掘的装备软件测评管理系统

设计并实现了一种可以支持测评基准数据挖掘的装备软件测评管理系统(AESTM)。AESTM为B/S结构,采用DCOM构件集成的方式开发,支持测试组的协同工作,可自动生成装备软件测评过程中的各类文档,可在装备软件定型测评项目过程中收集各类度量信息,并利用数据挖掘手段产生装备软件的基准数据。     

基于联合编程技术的线轮廓度评价软件开发

换挡毂线轮廓度测量评价难度大,单一的系统工程软件难以实时高效地实现线轮廓度评价的任务需求.采用.NET面向对象编程接口技术,利用Matlab开发线轮廓度评价算法,LabView进行实时数据的采集分析,实现联合编程,开发了线轮廓度评价软件.联合编程技术加速实现了从控制、采集、分析等任务的集成到实际部署应用的进程,提高了开发效率.     

一个基于GDM方法的集成软件过程度量模型

为满足软件企业不同角色的度量需求,基于GDM方法提出了一个面向企业应用的集成软件过程度量模型,模型包含组织结构定义、角色职责定义、GDM分析结果、业务目标和度量目标评价函数等基础元素,定义了元素间的关联关系.模型支持项目度量元收集范围的确立,维护了组织内数据加工及数据分析流程,可以为角色定制度量分析视图,并辅助确定过程改善活动的目标过程.基于此模型,开发了企业集成软件过程度量系统,在系统支持下,企业度量活动的自动化程度及效率得到了提高.     

软件众包参与度影响因素分析及预测模型

随着"众包"这种商业模式的快速发展, 越来越多的互联网公司选择以"众包"的形式发布软件任务. 然而, 软件任务因其高门槛、高复杂度、长周期等特性, 面临着严重的低参与度问题. 本文结合全球最大的软件众包平台TopCoder的数据, 对软件众包任务的参与度进行研究. 首先, 使用多元回归分析了影响软件众包参与度的因素; 接着, 综合数据挖掘领域的多种分类预测算法, 探讨软件众包参与度的预测模型. 希望通过本实证研究, 为发包方、众包平台降低软件众包的低参与风险提供参考.     

基于AHP与云模型的软件过程绩效评价方法

为改进软件质量和提高组织软件过程能力,提出了一种基于AHP与云模型的软件过程绩效评价方法.在对组织过程目标分析的前提下,建立起过程-子过程-度量项的映射关系.通过采集组织的历史数据,以项目测量和产品测量为基础,建立过程基线,运用AHP方法定量的确定出各层次对目标的权重值.基于此,结合云模型分析项目测量和产品测量数据,对过程绩效进行定性评价,持续性量化的改进软件过程能力.同时还给出了该方法的实际应用,应用结果表明了该评价方法的可行性.     

计算机图形设计软件教学改革的实践

通过对计算机图形设计软件教学现状及教学特点的分析,将研究性教学应用到计算机图形设计软件的教学过程,通过创设情境、任务驱动及成果展示评价,启发学生的求知欲,有效地培养学生的创新精神和创新能力.     

软件过程中同行评审的应用与度量

在软件过程中采用同行评审可以及早、高效地发现软件缺陷,从而广泛、深入且有效地吸收和应用软件过程信息,同时评价和提升软件过程能力水平.为了在软件过程中更加有效地应用同行评审,以CMM/CMMI为基础,详细分析了软件过程中同行评审的方法,并介绍了同行评审的流程.在此基础上,对如何建立和实施同行评审的度量进行了研究和分析,并提出了利用同行评审的度量数据建立组织同行评审过程能力基线的方法.     

一种新型圆弧插补算法

在现代计算机数控(CNC)系统中,插补算法大都采用计算机的程序软件来实现.目前应用最普遍的插补算法有基准脉冲法和数据采样法.     

An intelligent and improved density and distance-based clustering approach for industrial survey data classification

Engineering Asset Management (EAM) emphasizes on achieving sustainable business outcomes and competitive advantages by applying systematic and risk-based processes to decisions concerning an organization's physical assets. Nowadays, there is no specific method to evaluate performance of EAM and lack of benchmark to rank performance. To fill this gap, an improved density and distance-based clustering approach is proposed. The proposed approach is intelligent and efficient. It has largely simplified the current evaluating method so that the commitment in resources for manual data analyzing and performance ranking can be significantly reduced. Moreover, the proposed approach provides a basis on benchmarking for measuring and ranking the performance in Engineering Asset Management (EAM). Additionally, by using the intelligent approach, companies can avoid to pay expensive consultant fees for inviting external consultancy company to provide the necessary EAM auditing and performance benchmarking.     

Designing a knowledge-based system for benchmarking: A DEA approach

First developed by Xerox in 1979, benchmarking provides measurement and comparison to improve processes and achieve higher performance. Benchmarking has proven a powerful tool for total quality management and process improvement. Successful benchmarking implementation is based on an effective benchmarking tool. To effectively implement benchmarking processes, this work proposes an integrated framework for the benchmarking tool and knowledge-based system using the data envelopment analysis (DEA) method, and then develops an intellectual benchmarking knowledge-based system (BKBS) for benchmarking, performance evaluation and process improvement. Accordingly, this work illustrates how the benchmarking knowledge-based system (BKBS) is implemented in a medical center. This system can help determine the particular benchmarking partners to evaluate the relative efficiency and fill the gaps between the benchmarking partners in the healthcare industry. Finally, the intellectual benchmarking knowledge-based system offers a very fast way to implement the benchmarking processes.     

Repeated results analysis for middleware regression benchmarking

The paper outlines the concept of regression benchmarking as a variant of regression testing focused at detecting performance regressions. Applying the regression benchmarking in the area of middleware development, the paper explains how the regression benchmarking differs from middleware benchmarking in general, and shows on real-world examples why the existing benchmarks do not give results sufficient for regression benchmarking. Considering two broad groups of benchmarks based on their complexity, novel techniques are proposed for the repeated analysis of results for the purpose of detecting performance regressions.     

Semantic BMS: Allowing usage of building automation data in facility benchmarking

Facility benchmarking and evaluation of facility performance are the crucial tasks in reaching efficient, economical and sustainable facility operation. Modern buildings are equipped with building automation systems (BAS) that contain vast numbers of various sensors that can be utilised in performance assessment. However, such systems lack convenient tools for data inspection, which limits their use in building performance and efficiency analysis and benchmarking especially on large sites. The paper presents a middleware layer designed to enrich BAS data with additional semantic information. As a semantic model, an adaptation of the Semantic Sensor Network (SSN) ontology for the field of building operation analysis is used. The middleware provides convenient interfaces for querying the model. The proposed system provides the facility managers with a convenient way to use the BAS data for benchmarking and decision support.     

基于压缩感知的鲁棒性人脸表情识别

为了有效提高噪声背景下的人脸表情识别性能,提出一种基于压缩感知的鲁棒性人脸表情识别方法.先通过对腐蚀的测试样本表情图像进行稀疏表示,再利用压缩感知理论寻求其最稀疏的解,然后采用求得的最稀疏解信息实现人脸表情的分类.在标准的Cohn-Kanade表情数据库的实验测试结果表明,该方法取得的人脸表情识别性能优于最近邻法、支持向量机以及最近邻子空间法.可见,该方法用于人脸表情识别,识别效果较好,鲁棒性较高.     

An approach to assessing virtual environments for synchronous and remote collaborative design

This paper considers applying novel Virtual Environments (VEs) in collaborative product design, focusing on reviewing activities. Companies are usually anchored to commercial ICT tools, which are mature and reliable. However, two main problems emerge: the difficulty in selecting the most suitable tools for specific purposes and the complexity in evaluating the impact that using technology has on design collaboration. The present work aims to face both aspects by proposing a structured benchmarking method based on expert judgements and defining a set of benchmarking weights based on experimental tests. The method considers both human–human interaction and teamwork-related aspects. A subsequent evaluation protocol considering both process efficiency and human–human interaction allows a closed-loop verification process. Pilot projects evaluate different technologies, and the benchmarking weights are verified and adjusted for more reliable system assessment. This paper focuses on synchronous and remote design review activities: three different tools have been compared according to expert judgements. The two best performing tools have been implemented as pilot projects within real industrial chains. Design collaboration has been assessed by considering both process performance and human–human interaction quality, as well as benchmarking results validated by indicating some corrective actions. The final benchmarking weights can thus be further adopted for an agile system benchmark in synchronous and remote design. The main findings suggest defining both an innovative process to verify the expert benchmark reliability and a trusty benchmarking method to evaluate tools for synchronous and remote design without experimental testing. Furthermore, the proposed method has a general validity and can be properly set for different collaborative dimensions.     

A collaborative benchmarking framework for multibody system dynamics

Despite the importance given to the computational efficiency of multibody system (MBS) simulation tools, there is a lack of standard benchmarks to measure the performance of these kinds of numerical simulations. This works proposes a collaborative benchmarking framework to measure and compare the performance of different MBS simulation methods. The framework is made up of two main components: (a) an on-line repository of test problems with reference solutions and standardized procedures to measure computational efficiency and (b) a prototype implementation of a collaborative web-based application to collect, organize and share information about performance results in an intuitive and graphical form. The proposed benchmarking framework has been tested to evaluate the performance of a commercial MBS simulation software, and it proved to be an effective tool to collect and analyze information about the numerous factors which affect the computational efficiency of dynamic simulations of multibody systems.     

软件健壮性的包裹测试

软件的健壮性正变得越来越重要,目前还缺乏准确的、可重复利用的方法来衡量系统的健壮性。健壮性基准程序就是为了解决这一问题而提出的,它由一系列健壮性测试案例组成。本文以Linux为例,简要介绍了当今操作系统健壮性基准程序测试的主要方法,对用于分析测试结果的维度模型进行了分析,提出了软件健壮性的包裹测试方法,并用实例说明了其有效性。     

Unlabelled extra data do not always mean extra performance for semi-supervised fault prediction

Abstract: This research focused on investigating and benchmarking several high performance classifiers called J48, random forests, naive Bayes, KStar and artificial immune recognition systems for software fault prediction with limited fault data. We also studied a recent semi-supervised classification algorithm called YATSI (Yet Another Two Stage Idea) and each classifier has been used in the first stage of YATSI. YATSI is a meta algorithm which allows different classifiers to be applied in the first stage. Furthermore, we proposed a semi-supervised classification algorithm which applies the artificial immune systems paradigm. Experimental results showed that YATSI does not always improve the performance of naive Bayes when unlabelled data are used together with labelled data. According to experiments we performed, the naive Bayes algorithm is the best choice to build a semi-supervised fault prediction model for small data sets and YATSI may improve the performance of naive Bayes for large data sets. In addition, the YATSI algorithm improved the performance of all the classifiers except naive Bayes on all the data sets.     

基于时空稀疏模型的穿戴式心电信号压缩感知方法

基于时空稀疏模型,提出一种穿戴式心电信号的压缩感知方法,利用信号的时间相关性和空间相关性,来实现心电信号的重构.同时,还提出了一种"分-合"式字典学习算法,通过利用心电信号内在的聚类结构,对训练样本进行字典学习,从而构造出符合心电信号特点的字典,并对其进行稀疏表示.从而进一步提高了心电信号的重构性能.为了验证提出的心电信号压缩感知方法的有效性,采用OSET数据库中的心电数据,将其与其他两种基准算法进行了对比.仿真实验结果表明,所提出的心电信号压缩感知方法能有效地提高心电信号重构的质量.