Premium
Cloud service discovery method: A framework for automatic derivation of cloud marketplace and cloud intelligence to assist consumers in finding cloud services
Author(s) -
Alkalbani Asma Musabah,
Hussain Walayat
Publication year - 2021
Publication title -
international journal of communication systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.344
H-Index - 49
eISSN - 1099-1131
pISSN - 1074-5351
DOI - 10.1002/dac.4780
Subject(s) - cloud computing , computer science , ontology , cloud testing , the internet , service (business) , cloud computing security , world wide web , data mining , information retrieval , philosophy , economy , epistemology , economics , operating system
Summary The increase in the number of cloud services advertisements, needs for cloud services marketplace to enable significant interaction with cloud consumers. Majority of the existing literature has focused on developing algorithms (such as matching algorithms) and assumed the availability of cloud service information. Furthermore, little attention is given to the efficient discovery of cloud services over the internet. Existing approaches unable to describe a user‐friendly method of harvesting related cloud services from the web. Moreover, the existing literature lacks a comprehensive ontology to represent cloud services and a registry for cloud services publication and discovery. The incomplete information prevents discovering accurate services and deriving intelligence from cloud reviews data. The paper presents a framework for automatic derivation of cloud marketplace and cloud intelligence (ADCM&CI) that assist cloud consumers for an effective and efficient cloud service discovery. The framework depends on the capabilities of the Harvester as a Service (HaaS) crawler that provides a user‐friendly interface to extract real‐time cloud dataset. The paper used Protégé OWL a domain‐specific ontology to extract meaningful data from a semi‐structured repository and transform to SaaS ads attribute. The framework conducts sentimental analysis to excerpt the polarity of reviews that assist potential consumers in service selection. The paper considers three measures—precision, recall and F Score as a benchmark and evaluates the accuracy of the proposed approach using machine learning methods—SVM, KNN, Decision Tree and Naïve Bayes algorithms. Through experiments, we validate and demonstrate the suitability of the proposed framework for an effective and efficient cloud service discovery.