Empirical research is playing a significant role in software engineering (SE), and it has been applied to evaluate software artifacts and technologies. There have been a great number of empirical research articles published recently. There is also a large research community in empirical software engineering (ESE). In this paper, we identify both the overall landscape and detailed implementations of ESE, and investigate frequently applied empirical methods, targeted research purposes, used data sources, and applied data processing approaches and tools in ESE. The aim is to identify new trends and obtain interesting observations of empirical software engineering across different sub-fields of software engineering. We conduct a mapping study on 538 selected articles from January 2013 to November 2017, with four research questions. We observe that the trend of applying empirical methods in software engineering is continuously increasing and the most commonly applied methods are experiment, case study and survey. Moreover, open source projects are the most frequently used data sources. We also observe that most of researchers have paid attention to the validity and the possibility to replicate their studies. These observations are carefully analyzed and presented as carefully designed diagrams. We also reveal shortcomings and demanded knowledge/strategies in ESE and propose recommendations for researchers.

Activity-oriented context-aware (AOCA) applications are representative in pervasive computing. These applications recognize daily-life human activities, perceive the environment status related to the activities, and react to ensure the smooth performance of the activities. Existing research proposed a specific light-weight, incremental method to support the development of such applications; however it is not easy to learn and use. This paper aims to facilitate the development of such applications and improve the productivity of developers. We propose AocML, a textual domain-specific language which provides a high-level abstraction of AOCA applications. Specifically, we first show the software model of AOCA applications and the abstract syntax of AocML. Then, we introduce the concrete syntax of AocML. We also implement the tools for AocML, including the development environment as well as the generation of Java code and ontology specification. Moreover, we use a case study and evaluation to demonstrate the advantages of AocML.

In this work, we present DUSM (Disciplined Use Cases with Screen Mockups), a novel method for describing and refining requirements specifications based on disciplined use cases and screen mockups. Disciplined use cases are characterized by a quite stringent template to prevent common mistakes, and to increase the quality of the specifications. Use cases descriptions are formulated in a structured natural language, which allows to reach a good level of precision, avoiding the need for further notations and complex models. Screen mockups are precisely associated with the steps of the use cases scenarios and they present the corresponding GUIs (graphical user interfaces) as seen by the human actors before/after the steps executions, improving the comprehension and the expression of the non-functional requirements on the user interface. DUSM has been proposed and fine-tuned during several editions of a software engineering course at the University of Genova. Then, by means of a series of case studies and experiments, we validated the method and evaluated:1) its effectiveness in improving the comprehension and, in general, the quality of the produced requirements specification, and 2) its applicability in the industry, where the method has been found useful and not particularly onerous.

Modern compilers apply various code transformation algorithms to improve the quality of the target code. However, a complex problem is to determine which transformation algorithms must be utilized. This is difficult because of three reasons:a number of transformation algorithms, various combination possibilities, and several configuration possibilities. Over the last few years, various intelligent systems were presented in the literature. The goal of these systems is to search for transformation algorithms and thus apply them to a certain program. This paper proposes a flexible, low-cost and intelligent system capable of identifying transformation algorithms for an input program, considering the program's specific features. This system is flexible for parameterization selection and has a low-computational cost. In addition, it has the capability to maximize the exploration of available computational resources. The system was implemented under the Low Level Virtual Machine infrastructure and the results indicate that it is capable of exceeding, up to 21.36%, performance reached by other systems. In addition, it achieved an average improvement of up to 17.72% over the most aggressive compiler optimization level of the Low Level Virtual Machine infrastructure.

Thanks to the emerging 3D integration technology, The multiprocessor system on chips (MPSoCs) can now integrate more IP cores on chip with improved energy efficiency. However, several severe challenges also rise up for 3D ICs due to the die-stacking architecture. Among them, power supply noise becomes a big concern. In the paper, we investigate power supply noise (PSN) interactions among different cores and tiers and show that PSN variations largely depend on task assignments. On the other hand, high integration density incurs a severe thermal issue on 3D ICs. In the paper, we propose a novel task scheduling framework considering both the PSN and the thermal issue. It mainly consists of three parts. First, we extract current stimuli of running tasks by analyzing their power traces derived from architecture level simulations. Second, we develop an efficient power delivery network (PDN) solver to evaluate PSN magnitudes efficiently. Third, we propose a heuristic algorithm to solve the formulated task scheduling problem. Compared with the state-of-the-art task assignment algorithm, the proposed method can reduce PSN by 12% on a 2×2×2 3D MPSoCs and by 14% on a 3×3×3 3D MPSoCs. The end-to-end task execution time also improves as much as 5.5% and 7.8% respectively due to the suppressed PSN.

In the dark silicon era, many independent components of many-core processors are becoming voluntarily inactive due to the constraint of power consumption on a chip. However, to keep network connectivity, the on-chip interconnection must still be kept activated and wastes considerable energy to avoid the isolation of these inactive components, harming the energy-proportionality of the whole processor chip. In this paper, we propose a novel design to provide more energyproportional on-chip connection without damaging the network connectivity. To achieve this goal, we redesign the router architecture. The new architecture, DimRouter, supports three modes:normal, dark and dim. In the dim mode, only part of the router is active and provides flexible connection while the dark mode puts all router elements in the asleep state. Moreover, to maximize the number of dark routers, we also propose a reconfiguration algorithm based on degree-constrained Steiner Tree. The evaluation result under synthetic traffic shows that the new design can reduce the energy consumption up to 85% compared with the common design. For real application traffic, the new design can also save average 46% energy consumption with 4% performance improvement.

Integrated circuit (IC) camouflaging technique has been applied as a countermeasure against reverse engineering (RE). However, its effectiveness is threatened by a boolean satisfiability (SAT) based de-camouflaging attack, which is able to restore the camouflaged circuit within only minutes. As a defense to the SAT-based de-camouflaging attack, a brand new camouflaging strategy (called CamoPerturb) has been proposed recently, which perturbs one minterm by changing one gate's functionality and then restores the perturbed circuit with a separated camouflaged block, achieving good resistance against the SAT-based attack. In this paper, we analyze the security vulnerabilities of CamoPerturb by illustrating the mechanism of minterm perturbation induced by gate replacement, then propose an attack to restore the changed gate's functionality, and recover the camouflaged circuit. The attack algorithm is facilitated by sensitization and implication principles in automatic test pattern generation (ATPG) techniques. Experimental results demonstrate that our method is able to restore the camouflaged circuits with very little time consumption.

Topic modeling is a mainstream and effective technology to deal with text data, with wide applications in text analysis, natural language, personalized recommendation, computer vision, etc. Among all the known topic models, supervised Latent Dirichlet Allocation (sLDA) is acknowledged as a popular and competitive supervised topic model. However, the gradual increase of the scale of datasets makes sLDA more and more inefficient and time-consuming, and limits its applications in a very narrow range. To solve it, a parallel online sLDA, named PO-sLDA (Parallel and Online sLDA), is proposed in this study. It uses the stochastic variational inference as the learning method to make the training procedure more rapid and efficient, and a parallel computing mechanism implemented via the MapReduce framework is proposed to promote the capacity of cloud computing and big data processing. The online training capacity supported by PO-sLDA expands the application scope of this approach, making it instrumental for real-life applications with high real-time demand. The validation using two datasets with different sizes shows that the proposed approach has the comparative accuracy as the sLDA and can efficiently accelerate the training procedure. Moreover, its good convergence and online training capacity make it lucrative for the large-scale text data analyzing and processing.

Time intervals are often associated with tuples to represent their valid time in temporal relations, where overlap join is crucial for various kinds of queries. Many existing overlap join algorithms use indices based on tree structures such as quad-tree, B⁺-tree and interval tree. These algorithms usually have high CPU cost since deep path traversals are unavoidable, which makes them not so competitive as data-partition or plane-sweep based algorithms. This paper proposes an efficient overlap join algorithm based on a new two-layer flat index named as Overlap Interval Inverted Index (i.e., O2i Index). It uses an array to record the end points of intervals and approximates the nesting structures of intervals via two functions in the first layer, and the second layer uses inverted lists to trace all intervals satisfying the approximated nesting structures. With the help of the new index, the join algorithm only visits the must-be-scanned lists and skips all others. Analyses and experiments on both real and synthetic datasets show that the proposed algorithm is as competitive as the state-of-the-art algorithms.

Numerous applications of recommender systems can provide us a tool to understand users. A group recommender reflects the analysis of multiple users' behavior, and aims to provide each user of the group with the things they involve according to users' preferences. Currently, most of the existing group recommenders ignore the interaction among the users. However, in the course of group activities, the interactive preferences will dramatically affect the success of recommenders. The problem becomes even more challenging when some unknown preferences of users are partly influenced by other users in the group. An interaction-based method named GRIP (Group Recommender Based on Interactive Preference) is presented which can use group activity history information and recommender post-rating feedback mechanism to generate interactive preference parameters. To evaluate the performance of the proposed method, it is compared with traditional collaborative filtering on the MovieLens dataset. The results indicate the superiority of the GRIP recommender for multi-users regarding both validity and accuracy.

Packet classification has been studied for decades; it classifies packets into specific flows based on a given rule set. As software-defined network was proposed, a recent trend of packet classification is to scale the five-tuple model to multi-tuple. In general, packet classification on multiple fields is a complex problem. Although most existing softwarebased algorithms have been proved extraordinary in practice, they are only suitable for the classic five-tuple model and difficult to be scaled up. Meanwhile, hardware-specific solutions are inflexible and expensive, and some of them are power consuming. In this paper, we propose a universal multi-dimensional packet classification approach for multi-core systems. In our approach, novel data structures and four decomposition-based algorithms are designed to optimize the classification and updating of rules. For multi-field rules, a rule set is cut into several parts according to the number of fields. Each part works independently. In this way, the fields are searched in parallel and all the partial results are merged together at last. To demonstrate the feasibility of our approach, we implement a prototype and evaluate its throughput and latency. Experimental results show that our approach achieves a 40% higher throughput than that of other decomposed-based algorithms and a 43% lower latency of rule incremental update than that of the other algorithms on average. Furthermore, our approach saves 39% memory consumption on average and has a good scalability.

The state estimation plays an irreplaceable role in many real applications since it lays the foundation for decision-making and control. This paper studies the multi-sensor estimation problem for a contention-based unreliable wireless network. At each time step, no more than one sensor can communicate with the base station due to the potential contention and collision. In addition, data packets may be lost during transmission since wireless channels are unreliable. A novel packet arrival model is proposed which simultaneously takes into account the above two issues. Two scenarios of wireless sensor networks (WSNs) are considered:the sensors transmit the raw measurements directly and the sensors send the local estimation instead. Based on the obtained packet arrival model, necessary and sufficient stability conditions of the estimation at the base station side are provided for both network scenarios. In particular, all offered stability conditions are expressed by simple inequalities in terms of the packet arrival rates and the spectral radius of the system matrix. Their relationships with existing related results are also discussed. Finally, the proposed results are demonstrated by simulation examples and an environment monitoring prototype system.

To understand the content of ultrasound images more conveniently and more quickly, in this paper, we propose a coarse-to-fine ultrasound image captioning ensemble model, which can automatically generate the annotation text that is composed of relevant n-grams to describe the disease information in the ultrasound images. First, the organs in the ultrasound images are detected by the coarse classification model. Second, the ultrasound images are encoded by the corresponding fine-grained classification model according to the organ labels. Finally, we input the encoding vectors to the language generation model, and the language generation model generates automatically annotation text to describe the disease information in the ultrasound images. In our experiments, the encoding model can obtain the high accuracy rate in the ultrasound image recognition. And the language generation model can automatically generate high-quality annotation text. In practical applications, the coarse-to-fine ultrasound image captioning ensemble model can help patients and doctors obtain the well understanding of the contents of ultrasound images.