Departmental Bulletin Paper 深層学習を用いた要求分類の関する研究

安齋, 太地

It is important that requirements are clearly defined in software development, because they greatly affect the final correctness and the quality of software. To support producing higher-quality requirements, we propose an approach to classify requirements into functional requirements and non-functional requirements using the deep learning technology. We convert each request description sentence into a vector using Doc2Vec and classify it with a convolution neural network.Key Words : Requirement Engineering, Deep Learning, Functional / Non-functional requirements

Number of accesses :  

Other information