Designing Intelligent Software System for software defect prediction (SDP) is an important challenge in the field of software engineering using machine learning algorithms as class-imbalance problems cause difficulties for classification of defective and non-defective modules so Imbalanced learning deals with this problem and also used by researchers, but unfortunately with inconsistent results. For this, we conducted a comprehensive experiment for designing intelligent software system using the effect of imbalanced learning on imbalance dataset metrics, type of classifier, input metrics and imbalanced learning method. The major requirement in designing Intelligent Software System for software defect prediction typically uses methods and frameworks which allow software engineers to focus on development activities in terms of defect-prone code, thereby improving software quality and making better use of resources. Many software defect prediction datasets, methods and frameworks are complex, thus a comprehensive picture of defect prediction research from development phase is missing. This research aims to identify and analyse the research trends, datasets, methods and frameworks used in software defect prediction research from 2014 to 2020.