Research On Task Assignment Technology For Crowdsourcing Software Development

Crowdsourcing is a distributed outsourcing technology to an indeterminate and usually outsized crowd of people in an open-call format.It has attracted significant attentions from industry and academia.When employing crowdsourcing to accomplish software development tasks,crowdsourced software development(CSD)faces the serious challenges,such as assigning,sorting,and searching for suitable developers with respect to specific tasks.In this digital&competitive era,the majority of development tasks are assigned by bidding or competition.Consequently,as a result,the time required to search for a suitable task based on personal preference has wasted a significant amount of human effort,although many CSD developers would not compete for the tasks.Crowdsourced software development has provided a new method for the software industry to produce quality software based on an open-call format.Selecting an appropriate task to develop(developer-end)or evaluate(platform-end)is one of the primary problems in this type of open-call format.Assigning an improper task to an improper crowdsource(CS)developer does not only decrease the quality of the software deliverables but also causes overburden on both the platform and the developers.Selection of a suitable task from the extensively available large set of tasks is an intricate job for the developers in crowdsourcing software development.Additionally,it is also a tiring and a time-consuming job for the platform to evaluate thousands of tasks submitted by developers.Previous studies stated that managerial and technical aspects have prime importance in bringing success for software development projects,however,these two aspects can be more effective and conducive if combined with human aspects.To address the task assignment problem which best matches with the developers to task and vice versa,sorting the tasks based on the developers' human characteristics may increase task relevancy for developers,which can accelerate efficiency and lessen complexity of task assignment.The main objective of this study is to provide a novel solution which could be used to overcome serious challenges of CSD platform and to minimize the complexities of crowdsourcing task assignment and evaluation.In this regard,an algorithm has been proposed based on Myers-Briggs Type Indicator(MBTI)personality test and tested on controlled experiments.Results revealed that it is a more appropriate technique to measure the human personal preference and skills of CSD workers.Total 133 participants were involved in this study.The findings of studies suggested a novel solution to the developer end and also platform end to select and assign the task according to the personality type of the CS developers.This dissertation provides following contributions to the task assignment technology.1.This dissertation has developed a conceptual framework for task assignment model based on personality types.This model provided a basic structure for CSD workers to find suitable tasks and also a platform to assign the task directly.This will also match their personality and tasks.Because personality is an internal force which influences the behavior of developers.Consequently,this research presented a personality-based Task Assignment Model(TAM)from a developers point of view.Moreover,it also provides a solution to the platform to assign a task to CSD workers according to their personality types directly.2.This dissertation established an OPTAA(Optimum Task Assignment Algorithm)in crowdsourcing software development.It was based on the observation that CSD workers have chosen those tasks in which they felt comfortable and preferred according to their personality.For example,the personality type ISTJ is more attractive in developing but not interested in debugging task.Furthermore,the developers belonging to ISTJ,INTJ,ISFP,ESTP,ENTJ personalities have more inclination towards developing,ISFJ,ISTP,ISFP,INTP,ESFP personalities types are more inclined towards designing,INFJ,INFP,ESFJ,ENFJ personalities types are more inclined towards debugging.The experimental results confirmed the validity of this proposed algorithm.It provides an efficient solution to the platform and CSD worker to select and assign the task according to their personal preference.3.This study has also put forward a task sorting method in order to measure the impact of personality in task selection for CSD.This method was used to observe the significant relationship between task and personality type of a CSD worker,based on the important characteristics of a task i.e.money,time and type.An experimental approach was applied to a student data set belongs to Software engineering students from the University of Sindh(UOS)in Jamshoro,Pakistan.Myers-Briggs Type Indicator(MBTI)instrument was used to collect the personality types.The task information was composed of the selected tasks undertaken by the participants and their results.To extract the patterns from the collected data,several efficient techniques were used,for instance,cross tabulation and frequency techniques were applied to identify the common behaviors within the data.A weighted degree centrality(WDC)metric was applied to identify the most suitable node(i.e.,personality type for the tasks).Experimental results demonstrate that personality has a significant relationship with task selection.For instance,developers with intuitive(N)and feeling(F)personality traits are primarily focused on the time duration of a project.