You are on page 1of 14

1

Systematic Literature Review

Name

Institution Affiliation

Date
2

Abstract

Over the years, with the dramatic growth of new cloud providers, applications, services,

management platforms, data, etc., we have seen new solutions to manage such vast, shared, and

diverse services and resources. As a result, challenges to interoperability, portability, security,

discovery, selection, negotiation, cloud service, and resource details can often occur. In this

sense, semantic web technologies that are highly capable of cloud computing have proven to be

an effective way to overcome these challenges. This paper examines and explores the role of

Earth Web Technologies in the cloud from various authors. It looks and evaluates different

perspectives, architecture, and frameworks based on eight major research questions. At the end

of the review, research opportunities will be discussed as a roadmap. Keywords: cloud

computing, semantic web technologies, otology;

Introduction

Cloud computing has become a new technology model that aims to significantly reduce costs and

focus on providing computing resources over the Internet based on an easy-to-use payment

model. However, this technology's rapid development has reached a level of complexity, mainly

due to its vast, shared, and diverse range of services and resources. More precisely, the diversity

of cloud services and resources and how they are managed and accessed provide different and

non-standard interfaces for cloud actors and trainers and management. Besides, the descriptions

provided for cloud resources and services are ambiguous because they use different meanings to

address the same concept.

In this unavoidable situation, inconsistencies may arise, disparities between cloud

services may occur and still require human intervention. They were thereby leading to severe
3

challenges promoting cloud adoption. These challenges often include cloud resources and service

descriptions, cloud security, Cloud Services Innovation, Selection, and Intervention. The lack of

economics indicates an essential factor contributing enormously to the emergence of such

challenges and the urgent need for general explanation to avoid cloud services and resources and

cloud resources and services. Logic processes have been organized to verify various related

activities automatically. This may be the right starting point for easy adoption and make it easier

for machines to understand and read the resources of areas where websites have attracted more

attention to increasing aim.

The best literature on computing has already promoted these technologies to solve the

problems mentioned above. Most of them can be implemented with appropriate changes in cloud

services and resources. Although these components define specific features compared to

traditional web services, some challenges limit the application of these solutions in real-life

cloud areas. That situation in the cloud is an example of service innovation and choice. Such new

solutions have to deal with problematic interruptions and dynamic intervention support due to

cloud resources and services' dynamic nature and speed. It is hoped that this review will provide

researchers with a comprehensive overview of new ways to apply semantic technologies in cloud

computing and to analyze specific research solutions.

RESEARCH METHODOLOGY

We select the Systematic Literary Review (SLR) to assess the contribution of semantic web technologies

to improving cloud computing. To date, there has been a systematic review of the cloud, published in

2012, focusing solely on the role of oncology. To conduct this review, we go to a rigorous method for

identifying, evaluating and describing available research. According to Kitchen [38], an SLR has three

main stages: reporting plans, reviews and reviews. In the following subsections, each step is described.
4

Step 1: Review plan

At this stage the review protocol is developed based on the objectives, research questions, search

strategy, inclusion and exclusion criteria.

Objective and research questions.

Current research finds and maintains collaboration with Earth Web Technologies to adapt to cloud

computing. To do this, we are going to analyze publications (RQ) reviewed through research questions

with the aim of obtaining a comprehensive overview of semantic-based solutions: RQ 1. What are the

main objectives for implementing technology in the cloud? RQ 2. What is Semantic Web Technologies?

RQ3. What is the degree of perfection of each particular model? RQ 4. Do the authors suggest based on

whether their designs existed from the beginning or were they inspired by the criteria already defined in

this area? RQ 5. Which solution can be considered as a cloud model? RQ 6. What are the development

tools used to complete each solution? RQ 7. What are the evaluation methods used to assess the

efficiency of each solution? Solutions Need to verify cloud services (CS) or emulation tools (ST)? RQ 8.

What problems should each solution solve? Search strategy. We review Google Scholar, IEEE Explorer,

ACM Digital Library, Elsevier Scopus, Web of Science and arXiv.org for presenting voices at SCCC, Cloud

and SCGRD conferences for review. As search terms, we have selected the following: Cloud Computing,

Semantic Web Technologies and Oncology. These terms were combined to find related studies: "cloud

computing", "semantic web technologies"; "Cloud Computing", "Oncology". Admission and exemption

criteria. A review protocol should be included to evaluate each initial study and set exception criteria.

The criteria were carefully selected before completing the search to minimize practice. These criteria are

as follows: - Only complete and lengthy publications in English from journals and conferences published

from 2008 to 2016 will be considered. - Publications that do not mention at least one search order by
5

title or abstraction are automatically excluded. - Old publications that have never been quoted are

excluded from the review.

Step 2: Perform a review

During this time, we will follow a few steps: Search for practice; Select related studies and perform data

extraction and data synthesis. Search for study. The search strategy defined above was implemented in

September 2015. Through this we have recovered a total of 400 publications. However, these results are

not significant as they appear in some different databases. After filtering them, we completed the

publications of 100 candidates for the next step. Study selection. We determine which publications meet

inclusion criteria by analyzing their titles, summaries, and keywords. Initially, 72 publications followed

defined inclusion criteria. Therefore, after reading the full text, only 36 publications were carefully

selected. Each publication was examined to gather information on the eight questions selected.

III. REPORTING THE REVIEW

In this section, we will use the eight key questions described above to evaluate each solution. The first

question RQ1 option is very useful because it allows you to capture selected publications into four main

categories: Cloud Resources and Services Description, Upgrade, Cloud Services Selection, Interaction,

Cloud Cloud Interoperability, Cloud Security. The answer to RQ2 is to illustrate the semantic web

methods used by researchers. OQL, RDF / RDF, LED, SPARQL, RIF, SWRL, etc. For RQ3, we use P to

denote the partial perfection level and C to denote the absolute perfection. To provide feedback on

RQ4, we cite designs or criteria on which authors rely to refer to their designs. The answer to RQ5 in

each model is BP (BPSS stands for BPA, S stands for P and PAS stands for IAS). The answer to RQ6 refers

to the development tools that the authors used to implement their solutions. These tools can be

semantic APIs (Xena, Cartridges, etc.) or programming languages (Java, J2EE, etc.). Refers to cloud

services or simulators to provide feedback on RQ7. Finally, in order to answer the RQ8 final research
6

question, we need to categorize the topics that we can investigate. In fact, the reviewed documents

show evidence from six categories: Cloud Computing Basics (CCB: Cloud Services (CS), Resources (RES),

Service Models (SM), Deployment Models (DM), Technology (TRA): SLA Properties (FL) , Non-Functional

Properties, Cloud Actors (CA), Cloud Computing Characters (CCH: Resilience (ML) Properties (NT)) (NFP)

Business Associates (BRA): Pricing Models (PM), Governance Inspection (G) Angles (SRA).

Description of cloud resources and services

Studies in this category use semantic methods to describe and classify cloud services and resources. One

of the first attempts at this topic was suggested in [8], which has not been officially defined or

implemented, showing the main layers of the cloud. [7] and [10] describe a semantic model that

describes aspects of cloud governance. However, he did not provide any information on its

implementation or evaluation. [2] A three-layer semantic orientation model has been proposed, all of

which represent the oncology group. It is formalized in OWL, OWLS and SPARQL languages and inherits

many components from various programs such as NIST and OCCI. Bellini et al [4] refer to cloud oncology,

which allows you to define cloud service aspects at all levels to support cloud resource management.

The classification of cloud services and resources did not attract the attention of researchers, only two

papers were selected. De Martino et al [3] refer to the oncology that classifies cloud services and virtual

devices at pass and SAS levels, with automated creation and customer support when selected.

In addition, Ning et al introduced a semantic approach to solving the problems of multi-granularity

resource virtualization in cloud preparation. As described in Table I, most solutions are typically based

on OWL, SPARQL, and SWRL, which are partial in most cases, but complete in some cases. It is very clear

that the authors want to rely on existing models or are motivated by pre-defined criteria to represent

their designs. For RQ5, most policies represent three main cloud models: IAS, PA and SAS. 70% of jobs
7

have not developed or practically verified the tools needed to address them. Finally, CCBs and TRAs have

attracted a lot of attention from researchers in the area.

Cloud service upgrade, selection and intervention

This section provides a variety of studies using semantic technologies to find, select or communicate

with the cloud service. Considerable effort has been made to select and innovate different solutions for

the cloud service type (SASS, PA, IAS). De Modica et al. [6] Describes a semantic search based on a set of

oncology that allows us to describe the resource features offered by cloud providers and the needs

expressed by users. In [A], Goshi et al relied on gynecology to define the key elements and relationships

that are developing in the interaction, selection, innovation and structure of cloud services. The cloud

storage service prototype built on the Nilgiris platform confirms this reference. Deng et al. [16] officially

introduced a gynecological approach to model service offerings and related approaches, which ensures

system integrity. Darren et al [17] proposed the Chaplin-based, business-integrated cloud service

ontology model, which allows the service feature to capture cloud concept descriptions and their

interactions at multiple service segments and capture levels. The service recommendation system model

was developed as KB using this model. Kong et al. [39] and Tahman et al. [18] Cloud oncology-based

innovation systems have been suggested to determine similarity with services. In addition, a service

restructuring approach applicable to SAS and PA levels is suggested [19]. It focuses on service granularity

and semantic service selection with the goal of ensuring faster cloud error recovery. Aims to address

cloud services at the IAS level, Dost‌sardi et al. Oncology based framework is suggested for CloudPick to

facilitate cross-cloud expansion. Cloud services are enriched with semantics to automatically increase

search accuracy and memory. George et al solved the overall service selection problem in the multi-

cloud by suggesting a brokerbed approach using two gynecological models to illustrate overall service

requirements and provider submissions.


8

Somasundaram et al. . At SASS level, Engen et al [40] proposed an OWL-S based semantic cloud service

detection and selection system. More specifically, they defined an algorithm that facilitates match-

making and service scoring, as well as complex SWRL rules that allow compatibility criteria. Garcia et al.

[22] A semantic-based architecture has been proposed to automate and improve cloud services. Dimitris

et al [23] proposed a semantic framework based on the definition of business rules for web services and

software applications to support providers in the cloud service provisioning process. Khaim et al. [24] An

Earth Web-based approach was used to automate the QoS of the service. As noted in Table II, most

policies focus on controlling their knowledge of OWL, SPARQL and SWRL, although only a few DLs are

supported. However, in most cases, specific models are partially developed. In addition, most

approaches rely on models already developed in the area to develop their own models. For RQ5, most

policies and all models are viewed at once. In contrast, Sass focuses on the IAS model. Unfortunately, we

have revealed that there are some approaches that suggest and evaluate tools for solutions. CCB and

TRA represent the broader features studied in this section. C-Cloud interoperability and portability,

identified as one of the biggest barriers to cloud computing, can be addressed through lock-in,

interoperable and portable cloud services. The studies in this section show how the use of semantics can

differentiate between different cloud providers and achieve portability. Attempts to address

interoperability issues in cloud computing at the SASA and PA levels are in the context of the Mosaic

Project [35], which uses an open source cloud application programming interface. . Cloud based

applications. It is based primarily on general oncology, also known as mosaic oncology, which allows to

describe resources and resources. As far as we know, this gynecology is the complete gynecology

prescribed so far. So it can create a candidate for reuse through future contributions. In terms of

portability and interoperability at the SAS level, De Martino et al [26] provided a semantic

representation. There is an interconnected OWL Oncology set that describes cloud services, models,

virtual devices and their structure. Automated logic uses SPARQL queries, SWRL rules, and DLs to specify
9

similarities between services, tools, and models to improve this representation. Resoy et al. . There is a

connection between the actors and the constituencies.

Seva Semantics collects profiles from the cloud broker through the OWL-S editor, service process model

and various functions that provide semantic intervention through the generation of the service

foundation component. Kernan et al [2] specifically proposed a semantic framework that formalizes and

integrates the details of the BPAS configuration model to differentiate between BPAS providers. Due to

differences in PA levels, Lotus et al. It targets the growth of immigration or economic inequality by

defining three levels: financial institutions, financial types and the scope of financial conflict. Get

interactive at the IAS level, Di Martino et al. [29] Semantic architecture refers to the semantic nature of

intercloud communication. It relies on inter oncology oncology to meaningfully describe entities in

intercoded federated environments and to implement SPARQL queries to determine resource

configurations and exchanges. . It is used to synchronize the various data resources being developed by

the cloud management engine. Manno et al. [31] The Oncology Bed Resource introduced a framework

for integrated cloud infrastructure to manage the life cycle and distinguish between different clouds and

autonomous clouds. The proposed framework is based on the distribution and temporary KB involved in

the three oncology cases. Implemented the proposed framework developed as software. However, it is

not yet used in real world scenes. [32], an oncology-based system architecture proposed with the goal

of creating an inter-Pacific scheduler for integrated clouds working with multiple IAS. SPARQL is used as

a powerful way to define complex similarities between sources. Azarak et al [33] introduced a multi-

agent approach to allow oncology-based semantic resources. Earth Logic and Rule engines have been

proposed with the aim of correcting differences with different resources. So it is easy to consolidate

multiple resources from different providers. Finally, Moran et al. [12] The field RIF language is used to

control the behavior of applications mounted on multiple clouds. Therefore, it has proven to be a

valuable candidate to support such a goal through its flexibility and ability to avoid portability and
10

interoperability issues. As mentioned in Table III, most policies have concluded that their models are

symbolic of economics. However, most of them suggested partial models. The authors reused most of

the existing models for their own reference. Compared to other levels, IAAS is faster in this segment. In

addition, many tools have been implemented for different devices, many of which are not yet publicly

available but have not yet been evaluated using real cloud services.

A. Cloud security

In this section, we found in only four studies how the use of semantics improves cloud security. At the

SASS level, [12] and [13] worked to provide meaningful access control systems. . The theoretically

proposed approach assesses how the access system works through a case study. However, Choi et al. .

Connect. It helps identify the exact source of infection (GEN) and research capabilities (SPARQL).

However, this has not been implemented or verified. In contrast, Barnabas et al [14] focused on the IAS

level, which provides a meaningful understanding of gynecological use. Specifies SWRL rules, OWL

classes, basic information models, and authorization details stored in the KB.

The Authority uses the rational process of prescribing new knowledge to make decisions and eliminate

inequality in KB. In addition, this policy provides for multi-rental support and federal capability among

tenants through trusted relationships expressed in the SWRL Terms. Its implementation depends on the

structure to be placed on the IAS membrane, where a sample is tested for open oak stock. Finally, the

syntax demonstrates its limitations in expressing security policies, de Modica et al. [15] An

understanding has been developed based on gynecology that defines the key security aspects of policies

for cloud resources. Views suggested. Matchmaker engines, customer support capabilities and ability to

meet provider security requirements / match standards are also indicated. Both consumption cases

partially confirmed the effect. With the exception of the first study, all policies were formalized based on

economics as described in Table IV. However, many people suggest a partial model to show the
11

effectiveness of their ideas, which should not improve the treatment of such problems in practice.

Additionally, Paa levels are ignored and SaaS and IaaS levels are emphasized. Here, the focus is on cloud

security issues. Access control, security policy, etc.

IV. Rejection and Research Roadmap

There has been a lot of focus over the last five years on the use of semantic technologies to improve the

adoption of cloud computing. This research shows that all the selected approaches can be divided into

four main categories. Using the four tables already quoted and the descriptions of each category, we can

observe the following information about eight major research questions. About RQ1, researchers have a

great focus on cloud service discovery, selection and interaction, cloud interoperability and portability,

cloud resources and service details and finally cloud security. Since RQ2, most policies are based on the

semantic web standard languages OWL, SparkQL and SWRL. In the case of RQ3, the reviewed approach,

in most cases, provided a partial model with a brief description. In the case of RQ4, we believe that

many of the policies are based on existing models. Rarely, authors like to imagine their designs from

scratch, so they encounter errors in developing a new model. In the case of RQ5, most policies for the

first and second category are the same for all models. In contrast, it focuses on a specific pattern for

other categories. Regarding RQ6, we commented that some approaches using semantic programming

APIs and programming languages have developed tools aimed at implementing their solutions.

However, with the exception of the Mosaic Project [31], none of the policies provide any software

applications or tools already available for training. As with RQ7, we believe that the evaluation of

policies in practice is partial or incomplete. Some policies test their solution using an actual cloud service

and in most cases rely on the environment or representative examples they emulate. With regard to the

RQ8 question, it is clear that much attention has been paid by CCBs and Tros researchers. Although

multi-rental and resilience represent the most important features of the cloud area, they have been
12

overlooked by many approaches. Barnabas et al., In their work [14], considered multi-rent to be

intended to clarify the trust relationship between tenants.

In addition, there are certain procedures for capturing and reporting business goods. The cloud

governance field is a growing area of research and learning that illustrates the small interest gained in

governance. Finally, security issues are still not included in the research community. In addition to our

research questions, we have identified research opportunities for more sophisticated and innovative

semi-based solutions for cloud practitioners to gain more flexibility and scalability. From first class

studies, we observed that the majority of approaches do not follow the scientific process of forming

their semantic models. The design of such a model should follow a simple process, viz. Identify ideas and

relationships and formally describe them as logical theories to define and prove their level of expression

about rational language. In addition, current approaches may not be consistent with the growing design

approaches of their models. To automatically generate and expedite the information provided in the

cloud. Services or resources. The software available with the aim of providing such facilities is very useful

for trainers in this field. Similarly, some existing models also need to be partially redesigned. In fact,

some of them have many meanings and often they cause very functional overlap problems. In a similar

vein, no approach shows that they used the least specific methods to examine the structure of their

specimens. Based on the number of exaggerated ideas, most of them are not useful compared to those

in the cloud. In fact, all of these statistics apply to other categories. For the second category, it is

important to upgrade existing search systems to referral systems, allowing users to return their services,

freely stating their preferences or retrieving information. In addition, the study of the feasibility of

providing a search service, including an opinion mining engine that allows for an emotional classification

of services, represents an important area of future research. The third category requires further

research on the interoperability and portability of PA systems, as most current models and frameworks

focus on cloud computing only at the IAS and SAS levels. Research in this area is still very immature.
13

With regard to the languages used, current approaches to semantic selection, search and management

use SPARQL in many cases, making it very difficult for non-specialist users to use. Developers or cloud

users are asked to include a user-friendly heart application based on a natural language interface to

embed in their platform, which makes query systems simpler and more understandable to learners.

Although DL provides a positive transition between expression and scalability, very few support it. In a

similar vein, all approaches ignore RIF capabilities, because it is immediately clear that a single language

does not cover all known models that use rules for cognitive representation. From our observation, RIF

refers to a candidate who has the ability to articulate terms and ensure interchangeability between

them, as there are disadvantages and capabilities that reduce portability and interoperability between

cloud systems across different platforms. Different rules use languages.

As is the general opinion that touches all categories, existing semi-oriented models take into

account interactions with more general and extended models, the cloud layer and other layers. .

In conclusion, there is no guarantee that most policies are based on partial models that do not

implement tools in line with solutions. Furthermore, the immaturity of the evaluation of current

land solutions is frustrating in some cases, as clouds also need a specific model to accurately

mimic the environment, such as a provider such as Amazon or Microsoft References. Multi-

rental and resilience represent key features of cloud computing. However, the challenges they

face in cloud services or resources are still problematic. Currently, we are monitoring support for

such features. Based on the use of semantic technology models, initial approaches have been put

forward for management approaches that support multi-rental and elastic properties. However,

learning more about BRA, GRA and SRA is an open path to future research.

V. Conclusion
14

Cloud computing industry development has faced many challenges related to cloud service and

resource characteristics, differences, portability, security, selection, innovation and discussion

onions. As a direct result, semantic technologies have been suggested as an appropriate way to

overcome these potential challenges for cloud computing. In this review, we provide a

comprehensive overview of the application of semantic technologies in cloud computing based

on a comprehensive analysis of eight research questions on approaches, architecture and

frameworks taken from 36 standard papers published in 2008–2016. The successes of this review

shed light on the current state of semantic technologies in cloud computing and clarify the

research roadmap for the research questions raised in this review. In addition, we plan to further

explore these studies by defining technical features such as integration experience, QoS

enhancement and customer satisfaction, and to adopt such technology and show added value in

the cloud.

You might also like