点击数: 更新日期: 2022-09-05
中文题目：采用云计算进行软件测试的影响因素关联分析：一种两阶段的ISM-ANN方法
论文题目：Analyzing the interactions among factors affecting cloud adoption for software testing: a two-stage ISM-ANN approach
录用期刊：Soft Computing, 2020, V. 26, Issue 16（JCR Q2）
作者列表：
1）Sikandar Ali，中国石油大学（北京），信息科学与工程学院
2）Samad Baseer, Department of Computer System Engineering, University of Engineering and Technology, Peshawar 25000, Pakistan
3）Irshad Ahmed Abbasi, Department of Computer Science, Faculty of Science and Arts at Belqarn, University of Bisha, Bisha 61985, Saudi Arabia
4）Bader Alouffi and Wael Alosaimi, Department of Computer Science, College of Computers and Information Technology, Taif University, P. O. Box 11099, Taif 21944, Saudi Arabia
5）Jiwei Huang*，中国石油大学（北京），信息科学与工程学院
摘要:
To reduce costs and improve organizational efficiency, the adoption of innovative services such as Cloud services is the current trend in today’s highly competitive global business venture. The aim of the study is to guide the software development organization for Cloud-based testing adoption. Therefore, the objective is to develop a two-stage Interpretive Structural Model (ISM) and Artificial Neural Network (ANN)-based approach, for analyzing the factors influencing cloud adoption for software testing. This study first identifies the determinants and predictors of Cloud adoption for software testing through systematic literature (SLR) and empirical survey. Based on the collected data, an ANN was incorporated to weight the nonlinear effect of the predictors. Then, based on the results of empirical survey; a panel of ten experts was selected, to explore the multifaceted interrelationships among the influential factors (IFs) through ISM. To provide a concise understanding of the facts, Cross-Impact Matrix Multiplication Applied to the Classification (MICMAC) was used for factors classification. To achieve our objective, through SLR this study identifies 70 IFs. To offer a brief understanding of the issue, we distributed the identified IFs into ten predictors and analyzed their nonlinear effect on the predictors through ANN. Finally, the key 44 factors, identified through panel review, were priorities through ISM and were distributed into four Quadrants using MICMAC approach. Some studies in the form of survey have been conducted to examine the IFs affecting CCA. However, no attempt was made to explore the multifaceted interrelationships among them. This study concludes that software testing should be carried out in the Cloud.
背景与动机:
Some studies have been conducted to examine the barriers to SOP formation such as Tuten and Urban, Susarla, Verner et al., Chou and Pramudawardhani, Aundhe and Mathew, Kinnula et al., Ren et al., and Abdullah and Verner. However, no attempt was made to explore the multifaceted interrelationships among them. Further, Piltan and Sowlati considered partnership formation as a multi-criteria decision making (MCDM) problem. Therefore, unlike other researchers, we consider the SOP formation problem as an MCDM problem. ISM approach is an application of MCDM that explains the complex pattern of associations by incorporating simple notations of graph theory. Therefore, to bridge these gaps, this study implements the ISM approach to reconnoiter the interrelationships amongst the barriers.
设计与实现:
For the present study, we have developed the proposed research model by incorporating empirical research methodologies
using a combination of quantitative and qualitative research methods for data gathering, and a two-step ISM-ANN technique
for data analysis as shown in Fig. 1 explained in the subsequent sub-sections.
Phase#1: In the first phase, literature was reviewed via SLR in order to extract critical influential factors (CIFs).
Phase#2: In the second phase, to analyze the nonlinear relationship through ANN, a survey in the form of an online
questionnaire was carried out.
Phase#3: In the third phase, to analyze the structural association among the IFs, expert panel review was conducted.
Figure 1: Research methodology
理论模型:
To offer a brief understanding of the issue, we distributed the identified IFs into ten predictors as shown in Fig. 2. Using the collected data as a base, artificial neural networks (ANNs) were incorporated to weight the nonlinear effect of the predictors.
In this study, the items to weigh factors and questions for items are borrowed from the available literature on cloud adoption. The scale used in this study to measure various factors and predictors is attached. The splendid way to guarantee content validity is to choose elements from the available firmly established scale. For that reason, to sanction the content validity of the scales, we adapted and modify items from the earlier published research in the domain. As a result of SLR, we have recognized ten predictors and seventy factors as shown in Fig. 2. All factors and predictors were measured via 113 items using a questionnaire survey.
Figure 2: A model illustrating the classification of the identified cloud adoption factors into relevant predictors
实验验证:
Fig. 3 shows the Results obtained through artificial neural Network.
Figure 3. Proposed artificial neural network architecture
Sensitivity analysis of the NN model is performed to get the relative rank of the independent variables in predicting the adoption of CC. The summary of the normalized importance of independent variables are shown in Fig. 4.
Figure 4. Importance and normalized importance for sensitivity analysis
通讯作者简介:
黄霁崴，博士，教授，博士生导师，石油数据挖掘北京市重点实验室主任，中国石油大学（北京）计算机科学与技术系主任。2015年度北京市优秀人才，2018年度中国石油大学（北京）优秀青年学者，2020年度北京市科技新星。分别在2009年和2014年于清华大学计算机科学与技术系获得工学学士和工学博士学位，2012-2013年国家公派赴美国佐治亚理工学院联合培养。研究方向包括：系统性能评价和优化、随机模型理论和应用、服务质量测量与保障技术、服务计算和物联网等。担任中国计算机学会（CCF）服务计算专委会委员，CCF高级会员，IEEE、ACM会员。已主持国家自然科学基金、北京市自然科学基金等科研项目13项，在国内外著名期刊和会议发表论文五十余篇，出版学术专著1部，获得国家发明专利5项、软件著作权3项，担任多个国际顶级期刊和知名会议审稿人。联系方式：huangjw@cup.edu.cn。