An Ontology-Based Model for Task Recommendation in Crowdsourced Software Engineering Environment

Assigning tasks to appropriate people in crowd-based software engineering has always been anarduous task for a manager because the volume of the crowd to which a task is assigned is increasing day by day,and the process of selecting which task should be allocated to which person is becoming difficult. In order to maketask recommendations more effective and less time-consuming, much research has been done. However, theproblem is that such research does not contain the required implementation strategy and current recommendationtechniques in crowdsourcing use workers’ history and mining methods, which cannot relate workers with tasks andother workers. This research aims to incorporate semantic capabilities in task allocation in a crowdsourced softwareengineering environment so that machines could guide crowdsourcing managers in choosing appropriate crowdworkers for a task to be solved. In order to solve this problem and allow the machine to prescribe which resource ismore suitable for a task semantically, an ontology-based recommendation technique is proposed. Ontologies relateinformation in a way that the machine can understand its semantics. These ontologies, along with data, create aknowledge base upon which semantic web works. Two ontologies, one of the crowds to which the task must beassigned and one of the tasks under consideration, have been developed to incorporate semantic capabilities in thecrowdsourced task allocation process. These ontologies were designed keeping in view various factors obtainedthrough analyzing different researches done to improve the task allocation in crowdsourcing. Then these ontologieswere developed using RDF and OWL languages, and, in the end, these ontologies were tested using SPARQL. Theperformance and accuracy of the results returned by queries were also measured to know the efficiency of thisapproach in task allocation in a crowdsourcing environment. This approach can save time and provide moreefficiency in a manager responsible for assigning tasks to different persons in a crowd-based software engineeringenvironment.