Code smells are structures in code that often harm its quality. Manually detecting code smells is challenging, so researchers proposed many automatic detectors. Traditional code smell detectors employ metric-based heuristics, but researchers have recently adopted a Machine-Learning (ML) based approach. This paper compares the performance of multiple ML-based code smell detection models against multiple metric-based heuristics for detection of God Class and Long Method code smells. We assess the effectiveness of different source code representations for ML: we evaluate the effectiveness of traditionally used code metrics against code embeddings (code2vec, code2seq, and CuBERT). This study is the first to evaluate the effectiveness of pre-trained neural source code embeddings for code smell detection to the best of our knowledge. This approach helped us leverage the power of transfer learning – our study is the first to explore whether the knowledge mined from code understanding models can be transferred to code smell detection. A secondary contribution of our research is the systematic evaluation of the effectiveness of code smell detection approaches on the same large-scale, manually labeled MLCQ dataset. Almost every study that proposes a detection approach tests this approach on the dataset unique for the study. Consequently, we cannot directly compare the reported performances to derive the best-performing approach.

With ever-greater reliance of the developed world on information and communication technologies, constructing secure software has become a top priority. To produce secure software, security activities need to be integrated throughout the software development lifecycle. One such activity is security design analysis (SDA), which identifies security requirements as early as the software design phase. While considered an important step in software development, the general opinion of information security subject matter experts and researchers is that SDA is challenging to learn and teach. Experimental evidence provided in literature confirms this claim. To help solve this, we have developed a framework for teaching SDA by utilizing case study analysis and the hybrid flipped classroom approach. We evaluate our framework by performing a comparative analysis between a group of students who attended labs generated using our framework and a group that participated in traditional labs. Our results show that labs created using our framework achieve better learning outcomes for SDA, as opposed to the traditional labs. Secondary contributions of our article include teaching materials, such as lab descriptions and a case study of a hospital information system to be used for SDA. We outline instructions for using our framework in different contexts, including university courses and corporate training programs. By using our proposed teaching framework, with our or any other case study, we believe that both students and employees can learn the craft of SDA more effectively.

We propose a novel semi-supervised learning algorithm, called IMCC, designed for co-training classifiers on single-view datasets. Our method runs the co-training algorithm for a predefined number of times, each time using a different random split of features. Thus, a set of diverse co-training classifiers is created. Each of these classifiers then labels each of the examples for which we want to determine the class label. In this way, each example for classification is assigned multiple labels. We then treat this as a problem of learning from inconsistent and unreliable annotators in a multi-annotator problem setting and estimate the single hidden true label for each example. In experimental results obtained on 25 benchmark datasets of various properties IMCC outperformed five considered alternative methods for co-training on single-view datasets, and resulted in a statistical tie with a Naive Bayes classifier trained using a much larger set of labeled examples.