Automatic Software Defect Assignment Method Based on Deep Learning Processing Algorithm

Due to the maturity of information technology and the change of market demand, the software structure is more complex, the reports of software project defects are increasing, and the personnel and time costs of software project management are becoming more and more expensive. Therefore, more automatic and semi-automatic defect report dispatch algorithms are proposed to replace the original manual dispatch. This paper aims to study the automatic software defect assignment method based on deep learning processing algorithm. Based on the analysis of deep learning processing algorithm, software defect classification and defect report life cycle, an automatic defect assignment method based on deep learning is proposed. Experimental verification was carried out on the two large data sets of large open source software Eclipse and Moilla. The results show that the effect of adding two features at the same time is greatly improved.