Purpose

This paper follows a systematic literature review (SLR) method covering the published studies until March 2021. The authors have extracted the related studies from different online databases utilizing quality-assessment-criteria. In order to review high-quality studies, 32 papers have been chosen through the paper selection process. The selected papers have been categorized into three main groups, decision-making methods (17 papers), meta-heuristic methods (8 papers) and fuzzy-based methods (7 papers). The existing methods in each group have been examined based on important qualitative parameters, namely, time, cost, scalability, efficiency, availability and reliability.

Design/methodology/approach

Cloud computing is known as one of the superior technologies to perform large-scale and complex computing. With the growing tendency of network service users to utilize cloud computing, web service providers are encouraged to provide services with various functional and non-functional features and supply them in a service pool. In this regard, choosing the most appropriate services to fulfill users' requirements becomes a challenging problem. Since the problem of service selection in a cloud environment is known as a nondeterministic polynomial time (NP)-hard problem, many efforts have been made in recent years. Therefore, this paper aims to study and assess the existing service selection approaches in cloud computing.

Findings

The obtained results indicate that in decision-making methods, the assignment of proper weights to the criteria has a high impact on service ranking accuracy. Also, since service selection in cloud computing is known as an NP-hard problem, utilizing meta-heuristic algorithms to solve this problem offers interesting advantages compared to other approaches in discovering better solutions with less computational effort and moving quickly toward very good solutions. On the other hand, since fuzzy-based service selection approaches offer search results visually and cover quality of service (QoS) requirements of users, this kind of method is able to facilitate enhanced user experience.

Research limitations/implications

Although the current paper aimed to provide a comprehensive study, there were some limitations. Since the authors have applied some filters to select the studies, some effective works may have been ignored. Generally, this paper has focused on journal papers and some effective works published in conferences. Moreover, the works published in non-English formats have been excluded. To discover relevant studies, the authors have chosen Google Scholar as a popular electronic database. Although Google Scholar can offer the most valid approaches, some suitable papers may not be observed during the process of article selection.

Practical implications

The outcome of the current paper will be useful and valuable for scholars, and it can be a roadmap to help future researchers enrich and improve their innovations. By assessing the recent efforts in service selection in cloud computing and offering an up-to-date comparison of the discussed works, this paper can be a solid foundation for understanding the different aspects of service selection.

Originality/value

Although service selection approaches have essential impacts on cloud computing, there is still a lack of a detailed and comprehensive study about reviewing and assessing existing mechanisms in this field. Therefore, the current paper adopts a systematic method to cover this gap. The obtained results in this paper can help the researchers interested in the field of service selection. Generally, the authors have aimed to specify existing challenges, characterize the efficient efforts and suggest some directions for upcoming studies.

中文翻译：

---

面向云计算中的高效服务选择方法

目的

本文采用系统文献综述 (SLR) 方法，涵盖 2021 年 3 月之前已发表的研究。作者利用质量评估标准从不同的在线数据库中提取了相关研究。为了审查高质量的研究，通过论文选择过程选出了 32 篇论文。入选论文分为三大类，决策方法（17 篇论文）、元启发式方法（8 篇论文）和基于模糊的方法（7 篇论文）。基于重要的定性参数，即时间、成本、可扩展性、效率、可用性和可靠性，对每组中的现有方法进行了检查。

设计/方法/方法

云计算被誉为进行大规模复杂计算的优越技术之一。随着网络服务用户使用云计算的趋势越来越大，鼓励网络服务提供商提供具有各种功能性和非功能性特征的服务，并在服务池中提供。在这方面，选择最合适的服务来满足用户的需求成为一个具有挑战性的问题。由于云环境中的服务选择问题被称为非确定性多项式时间 (NP) 难题，因此近年来已经做出了许多努力。因此，本文旨在研究和评估云计算中现有的服务选择方法。

发现

获得的结果表明，在决策方法中，为标准分配适当的权重对服务排名准确性有很大影响。此外，由于云计算中的服务选择被称为 NP-hard 问题，因此与其他方法相比，使用元启发式算法来解决这个问题在以较少的计算工作量发现更好的解决方案和快速转向非常好的解决方案方面提供了有趣的优势。另一方面，由于基于模糊的服务选择方法可以直观地提供搜索结果并涵盖用户的服务质量（QoS）要求，因此这种方法能够促进增强的用户体验。

研究限制/影响

尽管当前的论文旨在提供全面的研究，但也存在一些局限性。由于作者应用了一些过滤器来选择研究，因此可能会忽略一些有效的作品。总的来说，这篇论文主要集中在期刊论文和一些在会议上发表的有效作品。此外，以非英文格式出版的作品已被排除在外。为了发现相关研究，作者选择了谷歌学术作为流行的电子数据库。虽然谷歌学术可以提供最有效的方法，但在选择文章的过程中可能没有观察到一些合适的论文。

实际影响

当前论文的成果将对学者有用和有价值，它可以成为帮助未来研究人员丰富和改进创新的路线图。通过评估最近在云计算服务选择方面的努力并提供所讨论工作的最新比较，本文可以为理解服务选择的不同方面奠定坚实的基础。

原创性/价值

尽管服务选择方法对云计算有着重要的影响，但仍然缺乏关于审查和评估该领域现有机制的详细而全面的研究。因此，本文采用系统的方法来弥补这一差距。本文获得的结果可以帮助对服务选择领域感兴趣的研究人员。总的来说，作者的目标是具体说明现有的挑战，描述有效的努力，并为即将进行的研究提出一些方向。