To apply Data Mining for Classification of Crowd sourced Software Requirements

Soonh Taj,Qasim Arain,Imran Memon,Asma Zubedi
DOI: https://doi.org/10.1145/3328833.3328837
2019-04-09
Abstract:Now a day's main focus of developers is to build quality software that works according to customer needs and for this reason it is necessary to gather right requirements as requirement elicitation is the critical step that impacts on the success of software project as misinterpreted requirements leads to the failure of software project. By keeping this in mind a research is carried out on improving requirements elicitation process and automating the process of classifying requirements. In this research, a model is proposed which will help in this scenario for requirements elicitation and requirement classification. This paper presents a model in which crowd sourcing approach is used so that customers, end users, stakeholders, developers and software engineers can make active participation for requirement elicitation process and requirements gathered using crowdsourcing approach are used by model for classification process i.e. classification of requirements into functional and non-functional requirements. For the proof of proposed model a case study is conducted. Results of case study provided the usefulness and efficiency of proposed model for classification of crowd sourced software requirements.
What problem does this paper attempt to address?