教学课程资源库 更多>>

关键词： Object-oriented software   Clone refactoring   Source code attributes   Unit test cases   Metrics   Relationships   Linear regression   Machine learning algorithms  

摘要： This paper aims at empirically measuring the effect of clone refactoring on the size of unit test cases in object-oriented software. We investigated various research questions related to the: (1) impact of clone refactoring on source code attributes (particularly size, complexity and coupling) that are related to testability of classes, (2) impact of clone refactoring on the size of unit test cases, (3) correlations between the variations observed after clone refactoring in both source code attributes and the size of unit test cases and (4) variations after clone refactoring in the source code attributes that are more associated with the size of unit test cases. We used different metrics to quantify the considered source code attributes and the size of unit test cases. To investigate the research questions, and develop predictive and explanatory models, we used various data analysis and modeling techniques, particularly linear regression analysis and five machine learning algorithms (C4.5, KNN, Naive Bayes, Random Forest and Support Vector Machine). We conducted an empirical study using data collected from two open-source Java software systems (ANT and ARCHIVA) that have been clone refactored. Overall, the paper contributions can be summarized as: (1) the results revealed that there is a strong and positive correlation between code clone refactoring and reduction in the size of unit test cases, (2) we showed how code quality attributes that are related to testability of classes are significantly improved when clones are refactored, (3) we observed that the size of unit test cases can be significantly reduced when clone refactoring is applied, and (4) complexity/size measures are commonly associated with the variations of the size of unit test cases when compared to coupling.

关键词： Synergistic degradation   Selective pressure   Phenolic compounds   Coal pyrolysis wastewater   Microbial community  

摘要： This study illustrated synergistic degradation of phenolic compounds by LAC-AS process via the insight into succession of microbial community under selective pressure. The results demonstrated that high phenols exhibited toxicity pressure to single AS process by eliminating non-tolerate bacteria, inducing vicious circulation by intermediates (catechol, nitrate, etc.) accumulation. However, LAC exerted another selective pressure and facilitated positive bio-community succession of moving biological bed reactor (MBBR). Firstly, it created rich microenvironments for diverse bacteria and promoted resilient adsorption for phenols with the assistance of biodegradation. Secondly, LAC enriched facultative bacteria, which developed multiple degradation paths on phenols and nitrogen based on multifunctional genes, counteracting the toxicity pressure. Specifically, phenols were degraded by the combination of anaerobic hydrolysis and oxidation, while conventional and shortcut nitrification-denitrification (SND) and nitrogen fixation all participated in nitrogen removal, achieving high removal of COD (93.49%), Tph (93.74%), TN (92.20%) and NH4+-N (93.20%) under the highest phenols.

关键词： Regression testing   Test case prioritization   State machine diagram   Test scenarios  

摘要： The exponential growth in the complexity of product requirements has raised the importance of software testing. Software change management is one of the major challenge and regression testing provides a solution in this regard. Regresion testing is performed to find out the effect of the modified component on the other components present in the application. Here, a behavioural model i.e. an UML State Machine Diagram (SMD) is used to model the system requirements. Further, an intermediate State Machine Graph (SMG) is constructed from SMD. The test scenarios are generated by identifying the paths those are linearly independent of each other. These test scenarios are validated for code coverage using dynamic program analysis technique. Then, by traversing the SMG, nodes are identified those got affected due to the different change carried away in the software. A repository is maintained to store the information about affected nodes periodically. Next time, during regression testing, pattern of nodes those are frequently affected, are identified from the stored data and these patternshelps in test case prioritization. Two more criterias i.e. functional priority set by the client and number of times the user accessed the user interface are also added up. Further, it calculates the total priority value of a test scenarios and test-scenarios are arranged in depending upon their total priority value. Moelling dynamic nature of the application as well as applying mining on the historical data makes the proposed approach more efficient when compared with the related works.

关键词： 课程资源云平台   智慧教室   电子教室   Github群体资源  

摘要： 随着社会信息化的提高,软件行业在信息技术中所占的比重远远高于其他行业,软件相关专业已经成为新技术、新经济、新产业发展的基石。软件工程课程被誉为软件开发的兵法,在软件类专业中具有举足轻重的地位。我们针对软件工程教学中存在的问题,提出"四维一体"的教学方法,将课程资源云平台、智慧教室、电子教室、Github群体资源统筹应用于软件工程课程教学中,改变了传统的教学方法和教学模式,克服了教学中时间、空间的局限性,提高了学生学习的积极性和主动性,取得了较好的教学效果。

关键词： 软件工程专业   能力培养   实践教学  

摘要： 实践教学在软件工程专业人才培养中具有重要地位,是培养具有合作竞争意识、探究精神、创新思维和实践能力的创新型高素质软件人才的重要环节。论文分析并总结了时代背景下软件工程专业实践教学尤其是实验教学的现状,结合行业企业对软件人才培养质量的要求,从顶层软件工程专业人才的培养计划制定、实验教学环节的教学内容和教学方法、实践教学体系师资培养等多个层面对软件工程专业实践教学改革进行了研究。