Showing papers in "Innovations in Systems and Software Engineering in 2021"
TL;DR: In this article, the authors explore machine learning and deep learning approaches to understand the data sources and explore how these learnings can be provided to the requirements gathering team to bring objectivity in the conversations between the requirement gathering team and the customer's business team.
Abstract: Secured software development must employ a security mindset across software engineering practices. Software security must be considered during the requirements phase so that it is included throughout the development phase. Do the requirements gathering team get the proper input from the technical team? This paper unearths some of the data sources buried within software development phases and describes the potential approaches to understand them. Concepts such as machine learning and deep learning are explored to understand the data sources and explore how these learnings can be provided to the requirements gathering team. This knowledge system will help bring objectivity in the conversations between the requirements gathering team and the customer's business team. A literature review is also done to secure requirements management and identify the possible gaps in providing future research direction to enhance our understanding. Feature engineering in the landscape of software development is explored to understand the data sources. Experts offer their insight on the root cause of the lack of security focus in requirements gathering practices. The core theme is statistical modeling of all the software artifacts that hold information related to the software development life cycle. Strengthening of some traditional methods like threat modeling is also a key area explored. Subjectivity involved in these approaches can be made more objective.
21 citations
TL;DR: In this paper, an ensemble learning technique called Bootstrap aggregating has been proposed for software defect prediction object-oriented modules and it is evident that the proposed method outperformed well compared to other approaches.
Abstract: To ensure software quality, software defect prediction plays a prominent role for the software developers and practitioners. Software defect prediction can assist us with distinguishing software defect modules and enhance the software quality. In present days, many supervised machine learning algorithms have proved their efficacy to identify defective modules. However, those are limited to prove their major significance due to the limitations such as the adaptation of parameters with the environment and complexity. So, it is important to develop a key methodology to improve the efficiency of the prediction module. In this paper, an ensemble learning technique called Bootstrap aggregating has been proposed for software defect prediction object-oriented modules. The proposed method's accuracy, recall, precision, F-measure, and AUC-ROC efficiency were compared to those of many qualified machine learning algorithms. Simulation results and performance comparison are evident that the proposed method outperformed well compared to other approaches.
17 citations
TL;DR: The result shows that the clusters made by the algorithm based on PCA and K -Means are similar and the results are acceptable on the basis of feedback received from existing customers and satisfies the customers’ requirements based on the amount of money the customers want to spend while doing online shopping.
Abstract: Recommender system is a computer-based intelligent technique which facilitates the customers to fulfill their purchase requirements. In addition to this, it also helps retailers to manage the supply chain of their business and to develop different business strategies keeping in pace with the current market. Supply chain management (SCM) involves the streamlining of a business’s supply-side activities to remain competitive in the business landscape. Maximizing the customer value is another important activity of SCM to gain an advantage in the market. In this work, the K-Means clustering algorithm has been used for the effective segmentation of customers who have bought apparel items. PCA has been used for dimensionality reduction of different features of products and customers. The main focus of this work is to determine the different possible associations of customers in terms of brand, product, and price from their purchase habits. The result shows that the clusters made by the algorithm based on PCA and K-Means are similar and the results are acceptable on the basis of feedback received from existing customers and satisfies the customers’ requirements based on the amount of money (price range) the customers want to spend while doing online shopping. The features of products purchased by customers were combined together to generate a unique product key for business, and a model was prepared to segment products based on the volume of products sold and revenue generated, and the price of products sold and revenue generated. This work, in the long run, will help business houses to build a sustainable, profitable, and scalable e-commerce business. Environmental, social, and economic aspects are important to make e-commerce more sustainable for the benefit of the society.
17 citations
TL;DR: A gradient boosting regressor model is proposed as a robust approach to achieve the accurate estimation of software effort and significantly performs better than all regression models used in comparison with both the datasets.
Abstract: The immense increase in software technology has resulted in the convolution of software projects. Software effort estimation is fundamental to commence any software project and inaccurate estimation may lead to several complications and setbacks for present and future projects. Several techniques have been following for ages of the software effort estimation. As the application of software is extensively increased in its size and complexity, the traditional methods aren’t adequate to meet the requirements. To achieve the accurate estimation of software effort, in this paper, a gradient boosting regressor model is proposed as a robust approach. The performance is compared with regression models such as stochastic gradient descent, K-nearest neighbor, decision tree, bagging regressor, random forest regressor, Ada-boost regressor, and gradient boosting regressor by employing COCOMO’81 containing 63 projects and CHINA of 499 projects. The regression models are evaluated by the evaluation metrics such as MAE, MSE, RMSE, and R2. From the results, it is evident that the gradient boosting regressor model is performing well by obtaining an accuracy of 98% with COCOMO’81 and 93% with CHINA dataset. The proposed method significantly performs better than all regression models used in comparison with both the datasets.
13 citations
TL;DR: A comprehensive survey of CPDP considering the latest work along with its SWOT analysis is presented in this article, where the authors present the current state of progress and the future prospects of cross-project defect prediction.
Abstract: Software fault prediction (SFP) refers to the process of identifying (or predicting) faulty modules based on its characteristics/software metrics. SFP can be done either using the same project data in both the training and testing phase i.e. within project defect prediction or using a different one, as done in cross-project defect prediction (CPDP). Previous works show that contemporary research in this field is progressing towards CPDP. To present the current state of progress and the future prospects of CPDP, this article presents a comprehensive survey of CPDP considering the latest work along with its SWOT analysis. This survey is targeted to present the novice researchers, academicians, and practitioners with the alphas and omegas of this contemporary challenging field. We have also carried a qualitative and quantitative evaluation of CPDP w.r.t some of the targeted research questions. A total of 34 significant primary CPDP studies published from 2008 to 2019 were selected. Both qualitative and quantitative data are extracted from each study. The collected data is then consolidated and analyzed to present a comprehensive report showing the current state of the art, along with the answers to the targeted research questions and finally the CPDP SWOT analysis. We observed that there exists a big scope for performance improvement in CPDP. Integration of feature engineering, exploration with different process metrics, hyperparameter tuning, class imbalance handling in CPDP setting are some of the ways identified for bringing enhancement in CPDP performance. Apart from this, we would like to conclude that there is a strong need to investigate Precision over the Recall and model’s validity in terms of effort/cost-effectiveness.
9 citations
TL;DR: Results showed that proposed metaheuristics algorithms-based weighted ensembles of hybrid search-based algorithms are good and outperformed machine learning- based algorithms and their ensemble for prediction of software effort.
Abstract: Software effort estimation is an essential task for software organizations to allocate resources efficiently during the development of software and to negotiate with customers. Various machine learning techniques-based models are used to predict the efforts required for the development of software products. These models utilize the past data of software projects to predict the efforts. In the present work, software efforts are estimated with the weighted ensemble of hybrid search-based algorithms. Weighted ensembles are created with the help of metaheuristic algorithms like firefly algorithm, black hole optimization, and genetic algorithm. The weighted ensemble of hybrid search-based algorithms based on metaheuristics algorithms is compared with some well-known machine learning algorithms and machine learning-based ensemble techniques. All the experiments were performed on three datasets obtained from PROMISE repository. Experiments were performed in R programming language using RKEEL and MetaheuristicsOpt r packages. Obtained results showed that proposed metaheuristics algorithms-based weighted ensembles of hybrid search-based algorithms are good and outperformed machine learning-based algorithms and their ensembles for prediction of software effort.
8 citations
TL;DR: In this paper, the proposed model is designed with 2 convolution layers and 3 dense layers and examined the module with 5 datasets including three benchmark datasets, namely CASIA, UBIRIS, MMU, random dataset, and the live video.
Abstract: Biometric applications are very sensitive to the process because of its complexity in presenting unstructured input to the processing. The existing applications of image processing are based on the implementation of different programing segments such as image acquisition, segmentation, extraction, and final output. The proposed model is designed with 2 convolution layers and 3 dense layers. We examined the module with 5 datasets including 3 benchmark datasets, namely CASIA, UBIRIS, MMU, random dataset, and the live video. We calculated the FPR, FNR, Precision, Recall, and accuracy of each dataset. The calculated accuracy of CASIA using the proposed system is 82.8%, for UBIRIS is 86%, MMU is 84%, and the random dataset is 84%. On live video with low resolution, calculated accuracy is 72.4%. The proposed system achieved better accuracy compared to existing state-of-the-art systems.
8 citations
TL;DR: An energy efficient street lighting framework is proposed in this paper to reduce energy consumption obtained from the street lights with minimum mean square error and shows an improvement over existing works.
Abstract: An energy efficient street lighting framework is proposed in this paper to reduce energy consumption obtained from the street lights. It is determined for various possible inter-distances offered by International Commission on Illumination. An ANN model is approached to obtain such reduced energy consumption for various traffic volumes on the road with minimum mean square error. The results of the proposed approach show an improvement over existing works.
7 citations
TL;DR: The development of a BCI-based gaming application with a mode of interactive media that takes advantage of the EEG signal gathered from the hardware used and discussed the scope of enhancing cognitive control capabilities in various psychiatric disorders in an interactive way with the help of the developed application.
Abstract: Assessing an individual’s attentional capabilities in an interactive way with the help of EEG-based signal and utilizing it to enhance cognitive control in obsessive–compulsive disorder is the primary focus of this paper. To realize this objective, BCI technology is used by studying the EEG signals emitted by the brain and processing the data in real time. In this paper, we presented the development of a BCI-based gaming application with a mode of interactive media that takes advantage of the EEG signal gathered from the hardware used. Further, we discussed the scope of enhancing cognitive control capabilities in various psychiatric disorders in an interactive way with the help of the developed application.
7 citations
TL;DR: In this paper, a sequential ensemble model is proposed to predict software faults, and the proposed model is also implemented on the 8 datasets taken from PROMISE and ECLIPSE repository.
Abstract: Unlike several other engineering disciplines, software engineering lacks well-defined research strategies. However, with the exponential rise in automation, the demand for software has observed an enormous elevation. Simultaneously, it necessitates having zero failures in the software modules to maximize the availability and optimize the maintenance cost. This has attracted many researchers to try their hand in formalizing the strategies for testing of software. Numerous researchers have suggested various models in this context. The authors in this paper present a sequential ensemble model to predict software faults. The employment of ensemble modeling in software fault prediction is motivated by its competence in various domains. The proposed model is also implemented on the 8 datasets taken from PROMISE and ECLIPSE repository. The proposed model's performance is evaluated using various error metrics, viz. average absolute error, average relative error, and prediction. The obtained results are encouraging and thus establish the competence of the proposed model.
7 citations
TL;DR: Wang et al. as discussed by the authors presented the regression testing using well-known algorithms, genetic algorithm, particle swarm optimization, a relatively new nature-inspired approach, gravitational search algorithm, and its hybrid with particle swarm optimizer algorithm.
Abstract: A software needs to be updated to survive in the customers’ ever-changing demands and the competitive market. The modifications may produce undesirable changes that require retesting, known as regression testing, before releasing it in the public domain. This retesting cost increases with the growth of the software test suite. Thus, regression testing is divided into three techniques: test case prioritization, selection, and minimization to reduce costs and efforts. The efficiency and effectiveness of these techniques are further enhanced with the help of optimization techniques. Therefore, we present the regression testing using well-known algorithms, genetic algorithm, particle swarm optimization, a relatively new nature-inspired approach, gravitational search algorithm, and its hybrid with particle swarm optimization algorithm. Furthermore, we propose a tri-level regression testing, i.e., it performs all the three methods in succession. Nature-inspired algorithms prioritize the test cases on code coverage criteria. It is followed by selecting the modification-revealing test cases based on the proposed adaptive test case selection approach. The last step consists of the removal of redundant test cases. The hybrid algorithm performed well for the average percentage of statement coverage, and the efficiency of genetic algorithm and particle swarm optimization is better comparatively. The proposed test case selection method can select at least 75% modification-revealing test cases using nature-inspired algorithms. Additionally, it minimizes the test suite with full statement coverage and almost negligible fault coverage loss. Overall, the simulation results show that the proposed hybrid technique outperformed the other algorithms.
TL;DR: The authors present their own approach to iris-based human identity recognition with DFFT components selected with principal component analysis algorithm, and show that satisfactory results can be obtained with the proposed method.
Abstract: One of the most important modules of computer systems is the one that is responsible for user safety. It was proven that simple passwords and logins cannot guarantee high efficiency and are easy to obtain by the hackers. The well-known alternative is identity recognition based on biometrics. In recent years, more interest was observed in iris as a biometrics trait. It was caused due to high efficiency and accuracy guaranteed by this measurable feature. The consequences of such interest are observable in the literature. There are multiple, diversified approaches proposed by different authors. However, neither of them uses discrete fast Fourier transform (DFFT) components to describe iris sample. In this work, the authors present their own approach to iris-based human identity recognition with DFFT components selected with principal component analysis algorithm. For classification, three algorithms were used—k-nearest neighbors, support vector machines and artificial neural networks. Performed tests have shown that satisfactory results can be obtained with the proposed method.
TL;DR: In this paper, the capability of skyline is extended to work with multiple dimensions and to search the multiple interesting points from the given search space. But, they restrict the computational complexity within a fixed upper bound.
Abstract: Skyline is a technique in database management system for multi-criterion decision making based on dominance analysis. Skyline overcomes the limitation of relational databases by handling the criteria that are inversely proportional to each other. Traditional skyline operation is conceptualized over two dimensions only, and it finds out single interesting point. In this paper we extend the capability of skyline to work with multiple dimensions and to search the multiple interesting points from the given search space. The work furthermore ranks skyline points with respect to the multiple interesting points. However, we restrict the computational complexity within a fixed upper bound. Skyline is commonly applied on tourism industries, and we consider two different case studies from this domain and execute the proposed methodology over the real-life data. Comparative study is given based on different parameters, and statistical analysis is also performed to illustrate the efficacy of the proposed method over the existing methods.
TL;DR: A modified Fisher’s test-based statistical method that makes use of test execution results as well as statement coverage information to determine the suspiciousness of each executable statement is proposed, which returns a rank list of statements based on their suspiciousity of containing a fault.
Abstract: For effective fault localization, we propose a modified Fisher’s test-based statistical method that makes use of test execution results as well as statement coverage information to determine the suspiciousness of each executable statement. Our technique returns a rank list of statements based on their suspiciousness of containing a fault. We also discuss an extension to our proposed approach for localizing programs with multiple faults. This involves partitioning the failed test cases into clusters such that they target different faults. Our experimental studies show that on an average, our proposed fault localization technique requires examination of 37.09% less code than existing techniques for localizing faults.
TL;DR: This study tackles the detection of refactoring documentation as binary classification problem by relying on text-mining, natural language preprocessing, and supervised machine learning techniques, and designs a tool to overcome the limitation of the manual process.
Abstract: Refactoring is the art of improving the internal structure of a program without altering its external behavior, and it is an important task when it comes to software maintainability. While existing studies have focused on the detection of refactoring operations by mining software repositories, little was done to understand how developers document their refactoring activities. Therefore, there is recent trend trying to detect developers documentation of refactoring, by manually analyzing their internal and external software documentation. However, these techniques are limited by their manual process, which hinders their scalability. Hence, in this study, we tackle the detection of refactoring documentation as binary classification problem. We focus on the automatic detection of refactoring activities in commit messages by relying on text-mining, natural language preprocessing, and supervised machine learning techniques. We design our tool to overcome the limitation of the manual process, previously proposed by existing studies, through exploring the transformation of commit messages into features that are used to train various models. For our evaluation, we use and compare five different binary classification algorithms, and we test the effectiveness of these models using an existing dataset of manually curated messages that are known to be documenting refactoring activities in the source code. The experiments are carried out with different data sizes and number of bits. As per our results, the combination of Chi-Squared with Bayes point machine and Fisher score with Bayes point machine could be the most efficient when it comes to automatically identifying refactoring text patterns in commit messages, with an accuracy of 0.96, and an F-score of 0.96.
TL;DR: The proposed model achieves non-threshold secret sharing using ElGamal and Paillier systems and is a zero-knowledge proof of the identification model also.
Abstract: This paper proposes a fully homomorphic computational model for secret sharing. The backbone of the proposed model is Chinese remainder theorem. The proposed model achieves non-threshold secret sharing. The homomorphism has been achieved using ElGamal and Paillier systems. Cryptographic hash function has been used for the identification of the true shareholders. The model identifies the legitimate shareholders without revealing their secret information. Thus, the model is a zero-knowledge proof of the identification model also. Further, the model regenerates the secret in the homomorphic domain. The efficiency and security of the model have also been analyzed.
TL;DR: A metaheuristic discrete genetic-learning-enabled particle swarm optimization algorithm combined with a trustworthiness-heuristic-based local search strategy has been proposed for targeted positive influence maximization in CRNs and the existing spread estimation function has been replaced by a computationally efficient positive influence spread estimationfunction.
Abstract: A consumer review network (CRN) is a social network among the members of a consumer review website where the relationships are formed based on the mutual ratings of the existing reviews of the participating members. The relationship is positive (negative) i.e. trust-based (distrust-based) when most of the reviews are given high (low) ratings. In a CRN, the consumers may not be interested in all product categories. The influence maximization in such a network demands seed set selection in such a way that the number of influenced consumers (with interest in the advertised product category) will be maximum. Formally, this is referred to as the targeted influence maximization (TIM) problem. Moreover, as the CRN is treated as a signed social network, the polarity of the social relationships impacts the influence propagation. As per the present state of the art, none of the existing solutions for TIM have considered the network as a signed one and are thus not suitable for CRN. In this paper, a metaheuristic discrete genetic-learning-enabled particle swarm optimization algorithm combined with a trustworthiness-heuristic-based local search strategy has been proposed for targeted positive influence maximization in CRNs. The existing spread estimation function has been replaced by a computationally efficient positive influence spread estimation function. The experiment has been conducted on two real-life CRNs and compared with the existing notable algorithm for necessary validation of the proposed solution.
TL;DR: With a pre-computation complexity, the proposed approach can provide significant resistance to SPA, DPA and HODPA attacks against modular exponentiation-based cryptosystems.
Abstract: This paper presents a secured computation of modular exponentiation to resist higher-order differential power analysis (HODPA) attacks in asymmetric cryptosystems like RSA. HODPA attacks can be resisted by segmenting secret sensitive data and its intermediate values into multiple shares. In modular exponentiation-based cryptosystems, the exponent plays a significant part in the secret key. We have used inner product with differential evolution algorithm to segment the exponent into multiple shares. Using entropy-based nearest neighbor algorithm, we have randomly computed independent modular exponentiation to resist SPA and DPA attacks. Analysis was done on 1024, 1536 and 2048 bit RSA. With a pre-computation complexity, the proposed approach can provide significant resistance to SPA, DPA and HODPA attacks against modular exponentiation-based cryptosystems.
TL;DR: In this article, a hybrid genetic algorithm-based firefly mating algorithm was proposed to solve Sudoku instances with a greater success rate for easy, medium, and hard difficulty level puzzles.
Abstract: Sudoku is an NP-complete-based mathematical puzzle, which has enormous applications in the domains of steganography, visual cryptography, DNA computing, and so on. Therefore, solving Sudoku effectively can bring revolution in various fields. Several heuristics are there to solve this interesting structure. One of the heuristics, genetic algorithm, is used by many researchers to solve Sudoku successfully, but they face various problems. Genetic algorithm has so many lacunas, and to overcome these, we have hybridised it in a novel way. In this paper, we have developed a hybrid genetic algorithm-based firefly mating algorithm, which can solve Sudoku instances with a greater success rate for easy, medium, and hard difficulty level puzzles. Our proposed method has controlled “getting stuck in local optima”, considering a smaller population and lesser generation.
TL;DR: Proposed parallel version of the steganography algorithm has been tested on five test samples of images for scalability analysis and results indicate significant speed up as compared to the sequential implementation of the technique.
Abstract: Concealing secret information in an image so that any perceptible evidence of the image alteration is insignificant, is known as image steganography. Image steganography can be implemented with either spatial or transform domain techniques. Spatial domain-based algorithms, generally the most widely used ones, refer to the process of embedding the secret information in the least significant bit positions of the cover image pixels. This paper proposes a chaotic tent map-based bit embedding as a novel steganography algorithm with a multicore implementation. The potential reasons for using chaotic maps in image steganography are sensitivity of these functions to initial conditions and control parameters. The computational complexity of the sequential least significant bit algorithm is known to be O(n). Hence, time complexity of the encryption/decryption algorithm is also a very important aspect. With the advantages offered by multicore processors, the proposed steganography algorithm can now be explicitly parallelized using the OpenMP API. As a pre-embedding operation, the quality of the randomness of the chaotic number sequences is tested with a NIST cryptographic test suite. The quality of the stego image is validated with statistical parameters such as structural similarity index (SSIM), mean square error (MSE) and peak signal-to-noise ratio (PSNR). Moreover, exploiting data parallelism inherent in the algorithm, multicore implementation of the algorithm with OpenMP API has also been reported. Proposed parallel version of the technique has been tested on five test samples of images for scalability analysis and results indicate significant speed up as compared to the sequential implementation of the technique.
TL;DR: A novel hybrid data mining technique consisting of Clustering and Modified Apriori Algorithm that results in improved efficiency and reliability of Software Defect Prediction is proposed that works by reducing the number of association rules generated.
Abstract: Software quality has been the important area of interest for decades now in the IT sector and software firms. Defect prediction gives the tester the pointers as to where the bugs will most likely be hidden in the software product. Identifying and reporting the defect probe areas is the main job of software defect prediction techniques. Early detection of software defects during Software Development Life Cycle could lead to a reduction in cost of development, time involved in further testing activities and rework effort post-production and maintenance phase, thus resulting in more reliable software. Software metrics can be used for developing the defect prediction models. Several data mining techniques can be applied on the available open-source software datasets. These datasets are extracted from software programs. Such datasets made publicly available by National Aeronautics and Space Administration for their various softwares have been extensively used in software engineering-related research activities. These datasets contain information on associated Software Metrics at module level. The proposed idea is a novel hybrid data mining technique consisting of Clustering and Modified Apriori Algorithm that results in improved efficiency and reliability of Software Defect Prediction. This technique works by reducing the number of association rules generated. The results are achieved by using interestingness measure called spread. The paper also does a comparative analysis of the results obtained from the novel technique with the existing hybrid technique of Clustering and Apriori.
TL;DR: The present paper gives an insight into the optimal set of filters in CNN model that gives the maximum overall accuracy of the classifier system.
Abstract: The convolutional neural network (CNN) has brought about a drastic change in the field of image processing and pattern recognition. The filters of CNN model correspond to the activation maps that extract features from the input images. Thus, the number of filters and filter size are of significant importance to learning and recognition accuracy of CNN model-based systems such as the biometric-based person authentication system. The present paper proposes to analyze the impact of varying the number of filters of CNN models on the accuracy of the biometric-based single classifiers using human face, fingerprint and iris for person identification, and also biometric-based super-classification using both bagging- and programming-based boosting methods. The present paper gives an insight to the optimal set of filters in CNN model that gives the maximum overall accuracy of the classifier system.
TL;DR: In this article, a set of MAPE design patterns for decentralized control in self-adaptive systems is proposed and an approach for composing them using a UML profile is presented.
Abstract: IoT systems are required to manage themselves to changes regarding their internal and external contexts. So, adaptability is a very important aspect in IoT software systems. The MAPE (Monitoring, Analysis, Planning, Execution) control loop model, inspired from the autonomic nervous system, has been identified as a crucial element for realizing self-adaptation in software systems. In fact, software design patterns provide architects and developers with reusable software elements helping them to master building complex software systems including several interconnected components. Complex self-adaptive systems require several architectural patterns in their design which leads to the need of architectural pattern composition. In this paper, we focus in modeling adaptability in IoT systems through a set of MAPE design patterns for decentralized control in self-adaptive systems and we propose an approach for composing them using a UML profile. Then, we propose formalizing the composition process using the Event-B method. In addition, we propose verifying adaptation properties based on the resulting formal specification. We illustrate our approach by modeling structural and behavioral features of the hybrid pattern resulting from the composition of two MAPE patterns and applied to the fall-detection ambient assisting living system for elderly people.
TL;DR: A CNN-based deep learning framework, named as STDNet: Script-Type detection Network, was developed to detect single-/mixed-script images and was compared to a state-of-the-art deep learning techniques and handcrafted feature-based methodologies where the proposed approach obtained a better performance.
Abstract: Script identification serves as a guide to the detection of the text of the scene through optical character recognition (OCR). But this is not a principal concern for the OCR engine. Until script identification, it is important to identify the script-type because today the text of the scene in natural images does not consist only of a single script, rather mixed-script words at character level are very often encountered. These words are also used in various ways, such as signboards, t-shirt graffiti, hoardings, and banners and often written in artistic way. In this work, a CNN-based deep learning framework, named as STDNet: Script-Type detection Network, was developed to detect single-/mixed-script images. To determine the feasibility of the system presented, tests were also undertaken with an outlier which is composed of a wide range of single scripts. Experiments were performed with over 20K images and 99.53% highest accuracy was reached. This approach was compared to a state-of-the-art deep learning techniques and handcrafted feature-based methodologies where the proposed approach obtained a better performance.
TL;DR: This paper presents simplified MDE implementations of Petri-nets applying Java, QVT, Kermeta and fUML that were experimented in order to debug a safety-critical system and summarises the lessons learned from this study.
Abstract: One of the promising techniques to address the dependability of a system is to apply, at early design stages, domain-specific languages (DSLs) with execution semantics. Indeed, an executable DSL would not only represent the expected system’s structure, but it is intended to itself behave as the system should run. In order to make executable DSLs a powerful asset in the development of safety-critical systems, not only a rigorous development process is required but the domain expert should also have confidence in the execution semantics provided by the DSL developer. To this aim, we recently developed the Meeduse tool and showed how to bridge the gap between MDE and a proof-based formal approach. In this work, we apply our approach to the Petri-net DSL and we present MeeNET, a proved Petri-net designer and animator powered by Meeduse. MeeNET is built on top of PNML (Petri-Net Markup Language), the international standard ISO/IEC 15909 for Petri-nets, and provides underlying formal static and dynamic semantics that are verified by automated reasoning tools. This paper first presents simplified MDE implementations of Petri-nets applying Java, QVT, Kermeta and fUML that we experimented in order to debug a safety-critical system and summarises the lessons learned from this study. Then, it provides formal alternatives, based on the B method and process algebra, which are well-established techniques allowing interactive animation on the one hand and reasoning about the behaviour correctness, on the other hand.
TL;DR: In this paper, the authors propose a risk management framework for incremental software development (ISD) processes that provides an estimate of risk exposure for the project when functional features are frozen while ignoring the associations with non-functional requirements.
Abstract: In incremental software development (ISD) functionalities are delivered incrementally and requirements keep on evolving across iterations. The requirements evolution involves the addition of new dependencies and conflicts among functional and non-functional requirements along with changes in priorities and dependency weights. This, in turn, demands refactoring the order of development of system components to minimize the impact of these changes. Neglecting the non-functional constraints in the software development process exposes it to risks that may accumulate across several iterations. In this research work, we propose a risk management framework for ISD processes that provides an estimate of risk exposure for the project when functional features are frozen while ignoring the associations with non-functional requirements. Our framework proposes suitable risk reduction strategies that work in tandem with the risk assessment module. We also provide a tool interface for our risk management framework.
TL;DR: In this article, the authors present two constrained permutation-based test scenario generation approaches, namely the level permutation and DFS-level permutation for concurrent activity diagrams, which restrict the exponential size to a reasonable size of test scenarios.
Abstract: Concurrency in application systems can be designed and visualized using concurrent activity diagrams. Such diagrams are useful to design concurrency test scenarios for testing. However, the number of test scenarios inside a fork-join construct could be exponential in size. The commonly used permutation technique generates all possible test scenarios, but it is exponential in size. Existing UML graph theoretic-based approaches generate a few test scenarios for concurrency testing. But they do not consider the full functionality of concurrent activity diagrams. In this work, we present two constrained permutation-based test scenario generation approaches, namely the level permutation and DFS level permutation for concurrent activity diagrams. These approaches restrict the exponential size to a reasonable size of test scenarios. It is achieved by generating a subset of permutations at different levels. The generated test scenarios are sufficient to uncover most concurrency errors like synchronization, data-race, and deadlocks. The proposed technique improves interleaving activity path coverage up to 35% compared to the existing approaches.
TL;DR: The first approach retrofits industrial control systems with autonomic properties that will allow them to automatically detect and recover from cyberattacks and other failures through the use of microservices that reconfigure the systems dynamically during attacks or failures.
Abstract: Many cyber physical systems have little or no cybersecurity mechanisms due to their limited computing capabilities or their history of running on isolated networks. As these systems have become interconnected and connected to corporate networks, they have become more vulnerable to cyberattacks. Providing cyber physical systems with autonomic properties will allow them to become more self-aware and react in near real time to attacks and failures. Testing these systems for their susceptibility to intelligent attacks is also needed to provide assurance of their resilience. This paper describes two approaches to providing assurances to cyber physical systems. The first approach retrofits industrial control systems with autonomic properties that will allow them to automatically detect and recover from cyberattacks and other failures through the use of microservices that reconfigure the systems dynamically during attacks or failures. The second approach uses intelligent agents in a modeling and simulation framework to test the resiliency of autonomous unmanned aerial systems. Agents are orchestrated using a range of algorithms and subjected to stressful environments to measure the efficiency and safety of their operations in a simulate multi-UAS air-traffic control problem.
TL;DR: In this paper, the authors proposed a Bayesian hierarchical technique for obtaining the optimal route to the most feasible parking lot, which is based on conflicting objectives, and hence, the problem belongs to the domain of multi-objective optimization.
Abstract: Discovering an optimal route to the most feasible parking lot has been a matter of apprehension for any driver. Selecting the most optimal route to the most favorable lot aggravates further during peak hours of the day and at congested places. This leads to a considerable wastage of resources specifically time and fuel. This work proposes a Bayesian hierarchical technique for obtaining this optimal route. The route selection is based on conflicting objectives, and hence, the problem belongs to the domain of multi-objective optimization. A probabilistic data-driven method has been used to overcome the inherent problem of weight selection in the popular weighted sum technique. The weights of these conflicting objectives have been refined using a Bayesian hierarchical model based on multinomial and Dirichlet prior. Genetic algorithm has been used to obtain optimal solutions. Simulated data have been used to obtain routes which are in close agreement with real-life situations. Statistical analyses have shown the superiority of the weights obtained using the proposed algorithm based on Bayesian technique over the existing frequentist technique.