International Journal of Software Engineering and Knowledge Engineering

Nowadays, one of the biggest challenges in banking sector, certainly, is assessment of the client’s creditworthiness. In order to improve the decision-making process and risk management, banks resort to using data mining techniques for hidden patterns recognition within a wide data. The main objective of this study is to build a high-performance customized credit scoring model. The model named Reliable client is based on Bank’s real dataset and originally built by applying four different classification algorithms: decision tree (DT), naive Bayes (NB), generalized linear model (GLM) and support vector machine (SVM). Since it showed the greatest results, but also seemed as the most appropriate algorithm, the adopted model is based on GLM algorithm. The results of this model are presented based on many performance measures that showed great predictive confidence and accuracy, but we also demonstrated significant impact of data pre-processing on model performance. Statistical analysis of the model identified the most significant parameters on the model outcome. In the end, created credit scoring model was evaluated using another set of real data of the same Bank.

Extracting the usability requirements are crucial during requirements review and requirements validation for different purposes. Usability requirements are hard to be determined until the real user has experienced the software. It is even more challenging when these usability requirements are documented in natural language, which has an inconsistent and unrestricted structure. Automated requirements extraction has been widely studied to facilitate the process of requirements checking. Nevertheless, the accuracy of requirements extraction method still can be improved. Thus, this paper has presented the usability keywords repository that followed the ISO 9126 and ISO 25010 usability category and has gone through the expert validation process. The usability requirement extraction method is moreover enhanced with extra procedures in conforming the usability requirement statement. First, each statement in the requirement document is checked if there is a keyword usability, whereby the keywords used must match with the usability keyword repository. In order to ensure that the selected statement is a usability requirement, the corresponding usability keyword position should be after the fixed auxiliary verbs. The performance of the improved usability requirements extraction method is then evaluated using precision, recall and accuracy.

For software testing, it is well known that the architecture of a software system can be utilized to enhance testability, fault detection and error locating. However, how much and what effects architecture-based software testing has on software testing have been rarely studied. Thus, this paper undertakes case study investigation of the effects of architecture-based software testing specifically with respect to fault detection and error locating. Through comparing the outcomes with the conventional testing approaches that are not based on test architectures, we confirm the effectiveness of architecture-based software testing with respect to fault detection and error locating. The case studies show that using test architecture can improve fault detection rate by 44.1%–88.5% and reduce error locating time by 3%–65.2%, compared to the conventional testing that does not rely on test architecture. With regard to error locating, the scope of relevant components or statements was narrowed by leveraging test architecture for approximately 77% of the detected faults. We also show that architecture-based testing could provide a means of defining an exact oracle or oracles with range values. This study shows by way of case studies the extent to which architecture-based software testing can facilitate detecting certain types of faults and locating the errors that cause such faults. In addition, we discuss the contributing factors of architecture-based software testing which enable such enhancement in fault detection and error locating.

In this paper, we introduce a framework for knowledge reuse and deduction in mechanical products design and development. The proposed system effectively exploits the capitalized and inferred knowledge. To this end, we settled up an ontology dealing with the design process of mechanical products such as “the car”. The ontology-based framework is supported by a software tool that brings an automatic and personalized assistance to correspondent actors using the deduction process. Indeed, the systems provides the relevant knowledge to the suitable users in order to facilitate their professional tasks considering their roles and collaboration. Experimental results have demonstrated the effectiveness of reusing knowledge during product development lifecycle.

Logs play an important role in the maintenance of large-scale systems. The number of logs which indicate normal (normal logs) differs greatly from the number of logs that indicate anomalies (abnormal logs), and the two types of logs have certain differences. To automatically obtain faults by K-Nearest Neighbor (KNN) algorithm, an outlier detection method with high accuracy, is an effective way to detect anomalies from logs. However, logs have the characteristics of large scale and very uneven samples, which will affect the results of KNN algorithm on log-based anomaly detection. Thus, we propose an improved KNN algorithm-based method which uses the existing mean-shift clustering algorithm to efficiently select the training set from massive logs. Then we assign different weights to samples with different distances, which reduces the negative effect of unbalanced distribution of the log samples on the accuracy of KNN algorithm. By comparing experiments on log sets from five supercomputers, the results show that the method we proposed can be effectively applied to log-based anomaly detection, and the accuracy, recall rate and F measure with our method are higher than those of traditional keyword search method.

With the aggrandizement scale of software system, the number of test cases has grown explosively. Test case prioritization (TCP) has been widely used in software testing to effectively improve testing efficiency. However, traditional TCP methods are mostly based on software code and they are difficult to apply to model-based testing. Moreover, existing model-based TCP techniques often do not take the likely distribution of faults into consideration, yet software faults are not often equally distributed in the system, and test cases that cover more fault prone modules are more likely to reveal faults so that they should be run with a higher priority. Therefore, in this paper, we provide a TCP approach based on Hidden Markov Model (HMM), to detect faults as earlier as possible and reduce the cost of modification. This approach consists of the following main parts: (1) transforming the Unified Modeling Language (UML) sequence diagram to HMM; (2) estimating the fault urgency according to fault priority and probability; (3) estimating the fault severity by analyzing the weight of the state in the HMM; (4) generating test case priority from fault urgency and fault severity, then prioritizing test case. The proposed approach is implemented on unmanned aerial vehicles (UAV) flight control system to perform TCP. The experimental results show that our proposed TCP approach can effectively enhance the probability of earlier fault detection and improve the efficiency and stability as compared to other prioritization techniques, such as original prioritization, random prioritization, additional prioritization and EPS-UML.