Program Behavior Research Articles

An AI powered system call analysis with bag of word approaches for the detection of intrusions and malware in Australian Defence Force Academy and virtual machine monitor malware attack data set

AbstractThis study propose the use of AI enabled machine learning algorithms with the Bag‐of‐Word (BoW) methods for the detection of intrusions by analysing the system call patterns. Host based Intrusion Detection System can make use of system call patterns to differentiate between normal and anomalous program behaviours. First, the system call patterns are pre‐processed with different approaches like BoW, BoW with Boolean value, BoW with Probability value and BoW with TF‐IDF. Next machine learning algorithms are used to evaluate the performance of classifier models. We used J48 (C4.5), Random Forrest, RIPPER, KNN, SVM, and NaiveBayes ML algorithms. This process was carried out on ADFA‐LD and on our proposed virtual machine monitor (VMM) malware attack data set for analysis. The proposed work is evaluated based on detection accuracy and false alarm rate metrics. Random Forrest algorithm performs better compared with other ML algorithms in terms of intrusion detection accuracy and false alarm rate on ADFA and VMM malware data set. The proposed data set provide better results compared with ADFA‐LD analysed using ML algorithms. The classifier model trained with ADFA and VMM malware system call data sets may do predictive analytics in detecting security issues for Industry 4.0 systems.

Forming‐Free Resistive Switching Memory Crosspoint Arrays for In‐Memory Machine Learning

In‐memory computing (IMC) with crosspoint arrays of resistive switching memory (RRAM) has gained wide attention for accelerating machine learning, data analysis, and deep neural networks. By IMC, matrix‐vector multiplication (MVM) can be executed in the crosspoint array in just one step, thus accelerating a broad range of tasks in machine learning and data analytics. However, a key issue for RRAM crosspoint arrays is the forming operation of the memories which limits the stability and accuracy of the conductance state in the memory device. In this work, a hardware implementation of crosspoint array of forming‐free devices for fast, energy‐efficient accelerators of MVM is reported. RRAM devices with a 1.5 nm‐thick HfO2 layer show an initial low resistance without forming and an analogue‐mode programming behavior for high‐accuracy IMC. Accurate hardware MVM is demonstrated by experimental eigenvalue/eigenvector calculation according to the power‐iteration algorithm, with a fast convergence within about ten iterations to the correct solution. Deflation technique and principal component analysis (PCA) enable the classification of the Iris dataset with 98% accuracy compared with floating‐point implementation. These results support forming‐free crosspoint arrays for accelerating advanced machine learning with IMC.

Complexity-guided container replacement synthesis

Containers, such as lists and maps, are fundamental data structures in modern programming languages. However, improper choice of container types may lead to significant performance issues. This paper presents Cres, an approach that automatically synthesizes container replacements to improve runtime performance. The synthesis algorithm works with static analysis techniques to identify how containers are utilized in the program, and attempts to select a method with lower time complexity for each container method call. Our approach can preserve program behavior and seize the opportunity of reducing execution time effectively for general inputs. We implement Cres and evaluate it on 12 real-world Java projects. It is shown that Cres synthesizes container replacements for the projects with 384.2 KLoC in 14 minutes and discovers six categories of container replacements, which can achieve an average performance improvement of 8.1%.

Fidelity to Program Specification of the National Health Service Digital Diabetes Prevention Program Behavior Change Technique Content and Underpinning Theory: Document Analysis.

BackgroundThe National Health Service (NHS) Diabetes Prevention Program is a behavior change intervention for adults in England who are identified as being at high risk of developing type 2 diabetes. The face-to-face service was launched in 2016, followed by a digital service (NHS Digital Diabetes Prevention Program [NHS-DDPP]) in 2019. A total of 4 service providers were commissioned to deliver the NHS-DDPP and were required to deliver the digital service in line with a program specification detailing the key intervention content. The fidelity of the behavior change content in the digital service (ie, the extent to which the program is delivered as intended) is currently unknown. Digital interventions may allow higher fidelity as staff do not have to be trained to deliver all intervention content. Assessing fidelity of the intervention design is particularly important to establish the planned behavior change content in the NHS-DDPP and the extent to which this adheres to the program specification. This is the first known independent assessment of design fidelity in a large-scale digital behavior change intervention.ObjectiveThis study aims to assess the fidelity of the behavior change content in each of the 4 NHS-DDPP providers’ intervention designs to the full program specification.MethodsWe conducted a document review of each provider’s NHS-DDPP intervention design, along with interviews with program developers employed by the 4 digital providers (n=6). Providers’ intervention design documents and interview transcripts were coded for behavior change techniques (BCTs; ie, the active ingredients of the intervention) using the Behavior Change Technique Taxonomy version 1 and underpinning theory using the Theory Coding Scheme framework. The BCTs identified in each digital provider’s intervention design were compared with the 19 BCTs included in the program specification.ResultsOf the 19 BCTs specified in the program specification, the 4 providers planned to deliver 16 (84%), 17 (89%), 16 (84%), and 16 (84%) BCTs, respectively. An additional 41 unspecified BCTs were included in at least one of the 4 digital providers’ intervention designs. By contrast, inconsistent use of the underpinning theory was apparent across providers, and none of the providers had produced a logic model to explain how their programs were expected to work. All providers linked some of their planned BCTs to theoretical constructs; however, justification for the inclusion of other BCTs was not described.ConclusionsThe fidelity of BCT content in the NHS-DDPP was higher than that previously documented for the face-to-face service. Thus, if service users engage with the NHS-DDPP, this should increase the effectiveness of the program. However, given that a clear theoretical underpinning supports the translation of BCTs in intervention designs to intervention delivery, the absence of a logic model describing the constructs to be targeted by specific BCTs is potentially problematic.

Enhancing bug localization with bug report decomposition and code hierarchical network

Bug localization, which aims to locate buggy source code files for given bug reports, is a crucial yet challenging software-mining task. Despite remarkable success, the state of the art falls short in handling (1) bug reports with diverse characteristics and (2) programs with wildly different behaviors. In response, this paper proposes a graph-based neural model BLoco for automated bug localization. To be specific, our proposed model decomposes bug reports into several bug clues to capture bug-related information from various perspectives for highly diverse bug reports. To understand the program in depth, we first design a code hierarchical network structure, Code-NoN, based on basic blocks to represent source code files. Correspondingly, a multilayer graph neural network is tailored to capture program behaviors from the Code-NoN structure of each source code file. Finally, BLoco further incorporates a bi-affine classifier to comprehensively predict the relationship between the bug reports and source files. Extensive experiments on five large-scale real-world projects demonstrate that the proposed model significantly outperforms existing techniques.

Predicting student performance in computing courses based on programming behavior

AbstractIn the context of programming language courses, offering early and personalized assistance to students with difficulties in solving code exercises is fundamental to helping them improve their performance. To this end, it is desirable to detect those students with behavioral patterns that could potentially result in poor performance in a timely manner. In this study, we analyze different behavioral metrics for predicting the performance of students in a 2nd‐year computer science course. Metrics are extracted from practical sessions, that is, code exercise labs that may be finished online during a period of 5–7 days. In addition to well‐known metrics (e.g., the relative time spent solving compilation/execution errors or skills for removing compilation errors), we also incorporate metrics that take into account the work habits of students (e.g., timestamps in which the student has undertaken 25%, 50%, and 75% of the exercise lab) and similarity between codes. With a sample of 224 students from seven different course groups, we construct robust predictors by training different machine learning models, reaching an R2 value of up to 0.40. We also perform a machine learning‐based analysis for detecting the most relevant response variables and analyze their relations to the final grades of students. The results show that the newly considered metrics have enough predictive power to be considered in the model. Finally, we graphically illustrate the nonlinear and seemingly interdependent relations that exist between predictor and response variables.

Predictive Analysis on Sensor Data Using Distributed Machine Learning

The science of getting computers to act without being explicitly programmed is known as machine learning. Machine learning is comparable to data mining in terms of how it works. Both systems sift through data in order to find patterns. Machine learning, on the other hand, instead of extracting data for human comprehension as in data mining applications, uses that data to increase the program's own understanding. Data patterns are detected by machine learning programmes, which then alter Programme behaviours accordingly. This research is based on real-world data obtained from sensors in an oil and Gas Corporation that monitor drilling procedures and equipment. The sensor data is streamed in at a one-second period, resulting in 86400 rows of data per day. Ox data's H2O has chosen for this problem after researching state-of-the-art Big Data analytics tools such as Mahout, RHadoop, and Spark because of its rapid in-memory processing, robust machine learning engine, and ease of use. Missed values can be estimated using accurate predictive analytics of massive sensor data, or wrong readings can be replaced owing to malfunctioning sensors or a broken communication line. It can also be used to predict circumstances to aid in various decision-making processes, such as maintenance planning and operation. In this project, sensor data has been evaluated and anticipate output using the H2O tool. The machine learning techniques has been employed in distributed systems, such as connecting five nodes to accomplish parallel processing.

Childcare, Parental Behavior and Well-Being during Childhood

Abstract In this paper, we study the long-term impacts on parental health and behaviors of a low-fee universal childcare program for children aged zero to five years old. We follow families for more than 10 years after the reform. For families with preschool children, we show that the adverse effects documented in previous studies on maternal depression and parental behaviors persist over time. However, once children enter school, most negative effects of the program on parental mental health and behaviors fade away. Only the negative effect on positive interactions may have persisted over time.

Human-affective Dimensions, Entrepreneurial Intentions, and Entrepreneurial Behaviors in Entrepreneurship Training Programs for Vulnerable Populations

This research investigated the entrepreneurial behaviors of vulnerable migrants who participated in an entrepreneurial education program and its management. By applying a structured questionnaire to the program participants, the variables that stood out most were: entrepreneurial intention, attitude towards the entrepreneur, social integration, and the usefulness of the entrepreneurial education program.

Online Power Management for Multi-Cores: A Reinforcement Learning Based Approach

Power and energy is the first-class design constraint for multi-core processors and is a limiting factor for future-generation supercomputers. While modern processor design provides a wide range of mechanisms for power and energy optimization, it remains unclear how software can make the best use of them. This article presents a novel approach for runtime power optimization on modern multi-core systems. Our policy combines power capping and uncore frequency scaling to match the hardware power profile to the dynamically changing program behavior at runtime. We achieve this by employing reinforcement learning (RL) to automatically explore the energy-performance optimization space from training programs, learning the subtle relationships between the hardware power profile, the program characteristics, power consumption and program running times. Our RL framework then uses the learned knowledge to adapt the chip's power budget and uncore frequency to match the changing program phases for any new, previously unseen program. We evaluate our approach on two computing clusters by applying our techniques to 11 parallel programs that were not seen by our RL framework at the training stage. Experimental results show that our approach can reduce the system-level energy consumption by 12 percent, on average, with less than 3 percent of slowdown on the application performance. By lowering the uncore frequency to leave more energy budget to allow the processor cores to run at a higher frequency, our approach can reduce the energy consumption by up to 17 percent while improving the application performance by 5 percent for specific workloads.

Counterfactual Models for Fair and Adequate Explanations

Recent efforts have uncovered various methods for providing explanations that can help interpret the behavior of machine learning programs. Exact explanations with a rigorous logical foundation provide valid and complete explanations, but they have an epistemological problem: they are often too complex for humans to understand and too expensive to compute even with automated reasoning methods. Interpretability requires good explanations that humans can grasp and can compute. We take an important step toward specifying what good explanations are by analyzing the epistemically accessible and pragmatic aspects of explanations. We characterize sufficiently good, or fair and adequate, explanations in terms of counterfactuals and what we call the conundra of the explainee, the agent that requested the explanation. We provide a correspondence between logical and mathematical formulations for counterfactuals to examine the partiality of counterfactual explanations that can hide biases; we define fair and adequate explanations in such a setting. We provide formal results about the algorithmic complexity of fair and adequate explanations. We then detail two sophisticated counterfactual models, one based on causal graphs, and one based on transport theories. We show transport based models have several theoretical advantages over the competition as explanation frameworks for machine learning algorithms.

A polynomial reduction of forks into logic programs

In this research note we present additional results for an earlier published paper [1]. There, we studied the problem of projective strong equivalence (PSE) of logic programs, that is, checking whether two logic programs (or propositional formulas) have the same behaviour (under the stable model semantics) regardless of a common context and ignoring the effect of local auxiliary atoms. PSE is related to another problem called strongly persistent forgetting that consists in keeping a program's behaviour after removing its auxiliary atoms, something that is known to be not always possible in Answer Set Programming. In [1], we introduced a new connective ‘|’ called fork and proved that, in this extended language, it is always possible to forget auxiliary atoms, but at the price of obtaining a result containing forks. We also proved that forks can be translated back to logic programs introducing new hidden auxiliary atoms, but this translation was exponential in the worst case. In this note we provide a new polynomial translation of arbitrary forks into regular programs that allows us to prove that brave and cautious reasoning with forks has the same complexity as that of ordinary (disjunctive) logic programs and paves the way for an efficient implementation of forks. To this aim, we rely on a pair of new PSE invariance properties.

The \u2018new normal\u2019 includes online prenatal exercise: exploring pregnant women\u2019s experiences during the pandemic and the role of virtual group fitness on maternal mental health

BackgroundPrenatal anxiety and depressive symptoms have significantly increased since the onset of the coronavirus (COVID-19) pandemic In addition, home confinement regulations have caused a drastic increase in time spent sedentary. Online group fitness classes may be an effective strategy that can increase maternal physical activity levels and improve mental health outcomes by providing an opportunity for social connectedness. The present study explores the experiences of pregnant women who participated in an online group exercise program during the pandemic and identifies relationships with maternal mental health and well-being. In addition, we present person-informed recommendations on how to improve the delivery of future online prenatal exercise programs.MethodsSemi-structured interviews were conducted with pregnant women (8-39 weeks of pregnancy) who participated in an online group exercise program, from March to October 2020 in Spain. A phenomenological approach was taken, and open-ended questions were asked to understand women’s experiences throughout the pandemic and the role the online exercise classes may have had on their physical activity levels, mental health, and other health behaviours such as diet. A thematic analysis was performed to evaluate data. In addition, women completed the State-Trait Anxiety Inventory and these data supplemented qualitative findings.ResultsTwenty-four women were interviewed, and the anxiety scores were on average 32.23 ± 9.31, ranging from low to moderate levels. Thematic analysis revealed that women felt safe exercising from home, an increased availability of time to schedule a structured exercise class, and consequently an improvement in their adherence to the program and other behaviours (i.e., healthier diet). Women emphasized feeling connected to other pregnant women when they exercised online together, and overall, this had a positive effect on their mental well-being. Women suggested that future online exercise programs should include flexible options, detailed instructions and facilitation by a qualified exercise professional.ConclusionPregnant women are receptive to online group exercise classes and expressed that they are an accessible option to accommodating physical activity during the pandemic. In addition, the online group environment provides an important sense of connectivity among pregnant women exercising together and this may mitigate the detrimental effect of COVID-19 on maternal mental health.

Exploring the Barriers and Challenges Saudi Arabian Nursing Students Encounter during their Internship Program: A Qualitative Study

Negative experiences on an internship can adversely affect the attitude and confidence of nursing students. This highlights the need to explore the experiences of nursing students as they progress through their training. The present study examines the challenges and problems faced by nursing students during their internship program at a hospital in Medina, Saudi Arabia. The study employed a qualitative exploratory study design by recruiting eight nursing students. Face-to-face semi-structured interviews were conducted to gather data. Moreover, the data were thematically analyzed. The major challenges and problems experienced by the nursing students were lack of support, distrust, improper behavior, and absence of a preceptorship program. There is a negative impact of these issues on nursing practice in clinical settings. The study concludes that positive outcomes are likely to be achieved by improving the internship program.

Feature-FL: Feature-Based Fault Localization

Fault localization aims at developing an effective methodology identifying suspicious statements potentially responsible for program failures. The spectrum-based fault localization is the widely used methodology by analyzing the statistical coincidences viewed from the spectrum to evaluate the suspiciousness of each statement of being faulty. However, just analyzing statistical coincidences in the coverage information perspective and without combining diverse amount of information may restrict fault localization effectiveness. Thus, this article proposes feature-based fault localization ( <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Feature-FL</monospace> ): A family fault localization methodology of feature-based metrics by combining the feature diversity from the view of program features into suspiciousness evaluation. Specifically, <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Feature-FL</monospace> defines a concept of branching execution probability to abstract program behaviors as the values of features. Then, <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Feature-FL</monospace> uses feature selection (i.e., a family of feature-based metrics) to evaluate the relevance of each feature with program failures. Finally, <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Feature-FL</monospace> associates each feature with its corresponding statement, and uses the relevance as the suspiciousness to locate suspicious statements. We present six feature-based metrics for <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Feature-FL</monospace> , and conduct an extensive study to evaluate the effectiveness of <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Feature-FL</monospace> and its potential over the state-of-the-art spectrum-based formulas. Our results provide insight into the potential among different feature-based metrics and also show <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Feature-FL</monospace> significantly outperforms the state-of-the-art spectrum-based formulas, e.g., an average <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">saving</i> of at least 30% over spectrum-based formulas in case of real faults.

Characterization of Program Behavior under Faulty Instruction Encoding

As process technology scales, electronic devices become more susceptible to soft errors. Soft errors can lead to silent data corruptions (SDCs), seriously compromising the reliability of a system. Researchers have explored error resilient encodings, which leverage crash patterns to detect SDCs. Despite its importance, much still remains to be determined regarding how errors propagate to cause SDCs or crashes. Understanding error propagation patterns could lead to more efficient implementation of error detection. An experimental study of program behavior in the presence of faulty instruction encoding under the IA-32 architecture is described in this study. Extensive fault injection experiments including over 70,000 faults were conducted, targeting all fields of instruction encoding. The analysis of the obtained data shows the following: (1) If the alignment of an instruction sequence is not preserved after injection, it causes crashes in a high probability (93.2%). (2) The SDC rate of an alignment-preserved category is close to that of a typical data injection. The SDC-prone fields include the opcode field, reg field, and immediate field. (3) Several crash patterns, such as violation of calling conventions, are revealed to extend the detection methods. These findings help us identify the vulnerable parts of instruction encoding, which need to be protected against soft errors. By applying the implications provided by the findings, we discuss feasible modifications, including swapping reg encodings, to reduce SDC rate, thus increasing the resilience of instruction set to soft errors.

From Requirements to Source Code: Evolution of Behavioral Programs

Automatically generating executable code has a long history of arguably modest success, mostly limited to the generation of small programs of up to 200 lines of code, and genetic improvement of existing code. We present the use of genetic programming (GP) in conjunction with context-oriented behavioral programming (COBP), the latter being a programming paradigm with unique characteristics that facilitate automatic coding. COBP models a program as a set of behavioral threads (b-threads), each aligned to a single behavior or requirement of the system. To evolve behavioral programs we design viable and effective genetic operators, a genetic representation, and evaluation methods. The simplicity of the COBP paradigm, its straightforward syntax, the ability to use verification and formal-method techniques to validate program correctness, and a program comprising small independent chunks all allow us to effectively generate behavioral programs using GP. To demonstrate our approach we evolve complete programs from scratch of a highly competent O player for the game of tic-tac-toe. The evolved programs are well structured, consisting of multiple, explainable modules that specify the different behavioral aspects of the program and are similar to our handcrafted program. To validate the correctness of our individuals, we utilize the mathematical characteristics of COBP to analyze program behavior under all possible execution paths. Our analysis of an evolved program proved that it plays as expected more than 99% of the times.

Depressive Symptoms Moderate the Association Between the Recent History of Alcohol Use Severity and Suicide Attempt History Among Adults in a Pretrial Jail Diversion Program.

Background: Justice-involved individuals are at elevated risk for suicidal behavior; however, research examining risk for suicidal behavior in pretrial jail diversion programs (pretrial diversion) is limited. Aims: We aimed to test (1) associations between depressive symptoms and alcohol and drug use, and suicide attempt history (SAH), and (2) interactions between depressive symptoms and alcohol and drug use in relation to SAH among adults in an urban pretrial jail diversion program. Method: The design was cross-sectional, including self-report assessments and file reviews of historical information. Adults (N = 274; Mage = 33.72; 73.7% men; 52.6% non-Hispanic Black) completed assessments within two weeks of beginning the pretrial program. Results: Depressive symptoms were positively associated with SAH after adjusting for other covariates. The significant depressive symptoms by alcohol use interaction indicated that adults with low to average depressive symptoms evidenced a similar likelihood of SAH to those with high depressive symptoms when they also had elevated alcohol use scores. Limitations: We used cross-sectional self-report data. Conclusion: It may be important to conduct suicide risk assessments for justice-involved people who use alcohol even when depressive symptoms are low.

The Influence of E-learning Behavior on Students’ Learning Performance of Disaster Emergency Knowledge

E-learning has advantages of huge learning resources, enthusiastic learners, friendly interaction and flexible space-time, especially, the large-scale application of many new educational technologies makes e-learning possible. Based on the theory of deconstruction program behavior and task technology fit model, four assumptions of e-learning behavior’s influence on learning performance of students’ disaster emergency knowledge were put forward from two aspects, namely the individual factors (self-efficacy), information literacy and environmental factors (technical characteristics, and promote conditions), the PLS (partial least squares) method was used to carry out regression analysis of affecting factors. Results show that the Cronbach’s ɑ values of the questionnaire are all greater than 0.8, indicating that the measurement indexes have high reliability. AVE (average variance extracted) values of all latent variables are greater than the threshold level of 0.5, showing good aggregation validity. The square root of AVE value is greater than correlation coefficient of each latent variable, and the discrimination validity is good. Information literacy, technical characteristics and facilitation conditions have a positive and significant impact on learning performance of students adopting e-learning of disaster emergency knowledge. The conclusions provide practical suggestions and opinions for enriching the analysis of learning behavior under the e-learning environment, depicting the characteristics of students’ e-learning behavior.

Computing within bacteria: Programming of bacterial behavior by means of a plasmid encoding a perceptron neural network

In nature, bacteria exhibit a limited repertoire of behaviors in response to environmental changes. Synthetic biology has now opened up the possibility of programming cells or unicellular organisms in order to enable them to perform certain tasks, which would allow the programming of 'intelligent' bacteria. Many of the theoretical ideas that Liberman proposed last century, for example his seminal idea that a cell is a computer, are now being put into practice with bacterial colonies in both wet and in silico experiments.These bacteria may one day be used to solve a wide range of problems whose solution requires their adaptation to external changes either within a bioreactor, organ or tissue of a patient or through the design of microbial-synthetic consortia oriented to their use in bioprocesses to produce medicines, biofuels or biomaterials. In this work, we show the possibility of programming synthetic bacteria with a previously trained perceptron neural network. First, we illustrate how a colony of bacteria endowed with a perceptron is able to solve an optimization problem in silico. Secondly, we study by means of in silico simulations how a perceptron can be applied to program behaviors in bacteria leading to social interactions and to the formation of complex communities that in the future would be useful in biotechnology. Finally, we go a step further, and study how the above perceptron designed to program bacterial behavior is implemented in a genetic circuit designed for this purpose. Once the genetic circuit was obtained, it was engineered into a plasmid.