Search

In addition to spreading information among friends, information can also be pushed through marketing accounts to non-friends. Based on these two information dissemination channels, this paper establishes a Susceptible-Infection-Marketing-Removed (SIMR) rumor propagation model. First, we obtain the basic reproduction number $ R_0 $ through the next generation matrix. Second, we prove that the solutions of the model are uniformly bounded and discuss asymptotically stable of the rumor-free equilibrium point and the rumor-prevailing equilibrium point. Third, we propose an optimal control strategy for rumors to control the spread of rumors in the network. Finally, the above theories are verified by numerical simulation methods and the necessary conclusions are drawn.

Interconnect development for the new technology node requires coordinated efforts of multiple module teams working to co-optimize patterning and metallization solutions to meet performance and yield, and reliability targets. This paper presents the results of interconnect patterning optimization to improve overlay and process window control using a novel methodology, Design-for-Inspection™ (DFI). Using design-assisted voltage contrast measurement techniques, the method enables in-line test and monitoring of process induced overlay and CD variation of back-end-of line (BEOL) features built with litho-etch-litho-etch (LELE) patterning. While only some of the features of multi-color patterning scheme are chosen to be aligned directly, other combination of metal line and via colors may have uncontrolled misalignment risking open or short failures. The paper shows how the complete metrology coverage of multi-color combinations between dual patterned via and dual patterned metal lines helps driving the improvement of overlay and process margins in 14nm technology. The optimized process margin for via opens enables higher yields and better reliability.

Aiming at the problems existing in the assessment of existing vulnerability code slicing method, such as incomplete extraction of slicing information, high model complexity and poor generalization ability, and no feedback in the evaluation process, a dual-granularity lightweight vulnerability code slicing evaluation (VCSE) model was proposed.Aiming at the code snippet, a lightweight fusion model of TF-IDF and N-gram was constructed, which bypassed the OOV problem efficiently, and the semantic and statistical features of code slices were extracted based on the double granularity of words and characters.A heterogeneous integrated classifier with high accuracy and generalization performance was designed for vulnerability prediction and analysis.The experimental results show that the evaluation effect of lightweight VCSE is obviously better than that of the current widely used deep learning model.

The rapid development of Internet and information technology has intensified the spread of rumors. Making full use of the comment mechanism on the Internet can effectively prevent the spread of rumors. This article focuses on strategies to reduce rumors by controlling comments. Firstly, based on the comment mechanism, a rumor propagation model with mixed rumor and truth is established. Secondly, we measure the cost of the rumor and, on this basis, model the comment based rumor-truth problem as the optimal control problem. Thirdly, we prove the existence of optimal control and derive the optimal system. Finally, the superiority of the optimal control strategy is verified by numerical simulation, and the rumor propagation is effectively slowed down under the premise of cost control, and the influence of some parameter changes on rumor cost-effectiveness control is studied.

Deteriorated air quality is a significant hazard to human health and a crucial problem in numerous countries. Existing research has extensively investigated the air quality states and future trends, but neglects the analysis of the time-space relevance behavior and the pollutant propagation relation of air quality. Based on the theory of complex network, this paper characterizes the interaction process between different sites, constructs the air quality propagation network and mines the key stations of the whole region. Firstly, pollutant monitoring data from the Beijing-Tianjin-Hebei region are used for analysis, monitoring stations are abstracted into network nodes, and edge sets are constructed using pollutant paths obtained from spatial and temporal reachability calculation. By accumulating the effect of the paths of the propagation network at different timestamps, the weight of the network during the propagation process is assigned for reflecting the interactive degree of stations. Secondly, the PageRank algorithm evaluates the importance of nodes in the propagation network to find the influential pollution area. Key stations considering both the quantity and quality of propagation relations are then obtained. Experiment results demonstrate the relevance of the node rank distribution and the real pollution condition and verify the reliability of the network and the node ranking algorithm. This work provides a reliable theoretical guidance on the timely prevention and control of air pollution, and gives a basis of the location selection of monitoring stations.

Rating predictions, as an application that is widely used in recommender systems, have gradually become a valuable way which can help user narrow down their choices quickly and make wise decisions from the vast amount of information. However, most existing collaborative recommendation models suffer from poor accuracy due to data sparsity and cold start problems that recommender systems contain only a few explicit data. To solve this problem, a new implicit trust recommendation approach (ITRA) is proposed to generate item rating prediction by mining and utilizing user implicit information in recommender systems. Specifically, user trust neighbor set that has similar preference and taste with a target user is first obtained by trust expansion strategy via user trust diffusion features in a trust network. Then, the trust ratings mined from user trust neighbors are used to compute trust similarity among users based on user collaborative filtering model. Finally, using the above filtered trust ratings and user trust similarity, the prediction results are generated by a trust weighting method. In addition, the empirical experiments are conducted on three real-world datasets, and the results demonstrate that our rating prediction model has obvious advantages over the state-of-the-art comparison methods in terms of the accuracy of recommendations.

With the rapidly increasing complexity and indispensable status of software systems, unprecedented challenges have been brought to software debugging and fault repair. Among the states of art automated fault localization techniques, spectrum-based fault localization (SBFL) is one of the most widely studied heuristic approaches. While, existing SBFL techniques are mostly focused on the analysis of the test spectrum, which loses sight of the effectiveness implied in interactions of software entities. In this paper, an optimized fault localization approach is proposed, which integrates spectrum-based fault localization and fault influence propagation analysis. The intuition is that the entity associated with more suspicious entities is more likely to be faulty. (1) A dynamic-instrumenting execution tracer is developed to record test coverage data and method call relations simultaneously. (2) The Fault Influence Network (FIN) based on complex network theory is constructed, of which the network topology is abstracted from method call relations and the weights of nodes are calculated applying a raw fault locator. Thus, the test spectrum and the method interactive relations are combined for further collaborative analysis. (3) A PageRank-based fault localization approach is proposed. By the computation of fault influence propagation, the suspiciousness score PR-Susp of the real faulty method can be enhanced significantly. Also, a candidate set pruning based on failing tests is implemented, which is used for the narrowing of investigation. Experiments are conducted over real-world fault dataset Defects4J with 33 raw spectrum-based fault locators, which proves that the proposed approach improves the baseline 14.9% averagely on the metric acc@5 with a growth rate of over 39% on all metrics of acc@1, acc@3, acc@5, MAP, and MWE.

Software vulnerabilities are the root causes of various security risks. Once a vulnerability is exploited by malicious attacks, it will greatly compromise the safety of the system, and may even cause catastrophic losses. Hence automatic classification methods are desirable to effectively manage the vulnerability in software, improve the security performance of the system, and reduce the risk of the system being attacked and damaged. In this paper, a new automatic vulnerability classification model (TFI-DNN) has been proposed. The model is built upon term frequency-inverse document frequency (TF-IDF), information gain (IG), and deep neural network (DNN): the TF-IDF is used to calculate the frequency and weight of each word from vulnerability description; the IG is used for feature selection to obtain an optimal set of feature word, and; the DNN neural network model is used to construct an automatic vulnerability classifier to achieve effective vulnerability classification. The National Vulnerability Database of the United States has been used to validate the effectiveness of the proposed model. Compared to SVM, Naive Bayes, and KNN, the TFI-DNN model has achieved better performance in multi-dimensional evaluation indexes including accuracy, recall rate, precision, and F1-score.

Diseases related to issues with blood pressure are becoming a major threat to human health. With the development of telemedicine monitoring applications, a growing number of corresponding devices are being marketed, such as the use of remote monitoring for the purposes of increasing the autonomy of the elderly and thus encouraging a healthier and longer health span. Using machine learning algorithms to measure blood pressure at a continuous rate is a feasible way to provide models and analysis for telemedicine monitoring data and predicting blood pressure. For this paper, we applied the gradient boosting decision tree (GBDT) while predicting blood pressure rates based on the human physiological data collected by the EIMO device. EIMO equipment-specific signal acquisition includes ECG and PPG. In order to avoid over-fitting, the optimal parameters are selected via the cross-validation method. Consequently, our method has displayed a higher accuracy rate and better performance in calculating the mean absolute error evaluation index than methods, such as the traditional least squares method, ridge regression, lasso regression, ElasticNet, SVR, and KNN algorithm. When predicting the blood pressure of a single individual, calculating the systolic pressure displays an accuracy rate of above 70% and above 64% for calculating the diastolic pressure with GBDT, with the prediction time being less than 0.1 s. In conclusion, applying the GBDT is the best method for predicting the blood pressure of multiple individuals: with the inclusion of data such as age, body fat, ratio, and height, algorithm accuracy improves, which in turn indicates that the inclusion of new features aids prediction performance.

Complex functional integration leads to intricate logical control flows which in turn presents a great challenge to construct software behavior models. In this paper, we propose a probabilistic software behavior model by mining the execution traces using control flow analysis. To describe the interactions between software components, a semantic characterization method is developed. A tracing mechanism is designed to collect execution logs, based on which algorithms are developed to recognize detailed control relations. Moreover, dynamic behavioral frequencies are statistically estimated which provide quantitative data for behavior prediction. Finally, a multi-label enhanced software complex network model, which holds single or composite call relations and calling probabilities, is constructed with the purpose of profiling systematically and structurally. An illustrative example shows that the modeling approach can discover interactive patterns correctly and characterize software behavior scientifically. The method has been used in seven real-world projects, and results show that the proposed model is effective on discovering the complexities of software behavior and hence help to detect defects in software design and to improve performance.

Recommender systems can provide users with an ordered list of various items, which greatly assists users to purchase products that they are satisfied with. However, item recommendation has been confronted with some inherent problems, such as sparse ratings and long-tail distribution, resulting in low accuracy of recommendations and insignificant marketing. In this paper, we propose a novel learning model based on trust diffusion and global item (TDGIL) to improve the accuracy of item rating prediction for recommender systems. Specifically, first, the rating information on items is mined and aggregated to the greatest extent based on trust diffusion characteristics among users. The benchmark prediction of item recommendation is updated by a user trust neighbor set and its item ratings, which are obtained by a trust diffusion algorithm. Then, the difference weights and compensation coefficients for all items are defined to learn users' potential preferences in the proposed global item model. Finally, the TDGIL learning algorithm is presented to train and learn the target networks by random gradient descent. The extensive experiments and results on two real-world datasets demonstrated that our proposed model can achieve significant improvements in the accuracy of rating prediction compared with some state-of-the-art methods.

Network intrusion detection systems (NIDS) are essential tools in ensuring network information security, and neural networks have become an increasingly popular solution for NIDS. However, with the gradual complexity of the network environment, the existing solutions using the conventional neural network cannot make full use of the rich information in the network traffic data due to its single structure. More importantly, this will lead to the existing NIDS have incomplete knowledge of the intrusion detection domain, and making it unable to achieve a high detection rate and good stability in the new environment. In this paper, we take a step forward and extract the different level features from the network connection, rather than a long feature vector used in the traditional approach, which can process feature information separately more efficiently. And further, we propose multimodal-sequential intrusion detection approach with special structure of hierarchical progressive network, which is supported by multimodal deep auto encoder (MDAE) and LSTM technologies. By design the special structure of hierarchical progressive network, our approach can efficiently integrate the different level features information within a network connection and automatically learn temporal information between adjacent network connections at the same time. Based on the three benchmark datasets from 1999 to 2017, including NSL-KDD, UNSW-NB15, and CICIDS 2017, we investigated the performance of our proposed approach on the task of detecting attacks within modern network. The experimental results show that the average accuracy of this method is 94% in binary classification and 88% in multi-class classification, which is at least 2% and 4% super than other methods respectively, and demonstrated that our model has excellent stability. Moreover, we further explore the multimodality and complementarity in traffic data, the experimental results show that the performance of detection model can be further improved in the range 2% to 5% when using our MDAE model to process the features of traffic data.

As a widely known chronic disease, diabetes mellitus is called a silent killer. It makes the body produce less insulin and causes increased blood sugar, which leads to many complications and affects the normal functioning of various organs, such as eyes, kidneys, and nerves. Although diabetes has attracted high attention in research, due to the existence of missing values and class imbalance in the data, the overall performance of diabetes classification using machine learning is relatively low. In this paper, we propose an effective Prediction algorithm for Diabetes Mellitus classification on Imbalanced data with Missing values (DMP_MI). First, the missing values are compensated by the Naïve Bayes (NB) method for data normalization. Then, an adaptive synthetic sampling method (ADASYN) is adopted to reduce the influence of class imbalance on the prediction performance. Finally, a random forest (RF) classifier is used to generate predictions and evaluated using comprehensive set of evaluation indicators. Experiments performed on Pima Indians diabetes dataset from the University of California at Irvine, Irvine (UCI) Repository, have demonstrated the effectiveness and superiority of our proposed DMP_MI.

Identifying defective software entities is essential to ensure software quality during software development. However, the high dimensionality and class distribution imbalance of software defect data seriously affect software defect prediction performance. In order to solve this problem, this paper proposes an Ensemble MultiBoost based on RIPPER classifier for prediction of imbalanced Software Defect data, called EMR_SD. Firstly, the algorithm uses principal component analysis (PCA) method to find out the most effective features from the original features of the data set, so as to achieve the purpose of dimensionality reduction and redundancy removal. Furthermore, the combined sampling method of adaptive synthetic sampling (ADASYN) and random sampling without replacement is performed to solve the problem of data class imbalance. This classifier establishes association rules based on attributes and classes, using MultiBoost to reduce deviation and variance, so as to achieve the purpose of reducing classification error. The proposed prediction model is evaluated experimentally on the NASA MDP public datasets and compared with existing similar algorithms. The results show that EMR_SD algorithm is superior to DNC, CEL and other defect prediction techniques in most evaluation indicators, which proves the effectiveness of the algorithm.

Numerous countries in the world pay close attention to the change of air quality and the control of air pollution. Air quality prediction has become a challenging issue owing to the complex characteristics with time-space nonlinearity and multi-dimensional feature interaction. This paper inventively proposes a three-dimensional presentation of the air quality data and establishes a deep model using spatiotemporal collaborative strategy to predict regional air quality, which can properly deal with multiple characteristics of air quality. Firstly, a theory of regional air quality prediction is given with reasonable analyses and scientific definitions, for systematically analyzing the closely connected areas and simultaneously predicting multiple locations. Secondly, a three-dimensional data structure called Relevance Data Cube is constructed utilizing a clustering algorithm, time sliding windows and correlation analysis of factors. This structure depicts the whole relationships among multiple dimensions of air quality data clearly and provides a basic analysis foundation for subsequent processing. Thirdly, based on the above theory and data structure, a prediction model is proposed to demonstrate the applicability of our work. Spatial relevance and factor influence are processed using dimension reduction technique provided by CNN, which performs automatic mining of spatial correlation and factors reaction of air quality. Besides, LSTM is combined with CNN to deal with the temporal dependency among the data dynamically. Finally, the proposed model is compared with other neural networks by a large number of experiments. The model is proved to better adapt to the characteristics and predict the air quality more accurately, which is more suitable and reliable for the field of air quality.

This paper discusses how word-of-mouth marketing affects the profits of product sales in social network-based shopping under good after-sales service. First, a new word-of-mouth communication model based on silent evaluation, positive evaluation, and negative evaluation is proposed. Second, we use the way of increasing after-sales service to achieve high praise and thereby maximize the expected profits. Thus, the proportion control problem of after-sales service investment is modeled as an optimal control problem. Third, the existence of optimal control is proved, and an optimal control strategy for dynamic proportion of after-sales service investment is proposed. Fourth, through data simulation of different real-world networks, it is verified that the expected profits under the dynamic after-sales service strategy is higher than that under any uniform control strategy. Finally, sensitivity analysis is performed to explore how different parameters affect the expected profits.

Analysis on the modularity of software network plays a critical role in the detection of software vulnerabilities and in the improvement of software stability, reliability, and robustness. This paper intends to propose a method based on the most vital nodes to analyze the modularity of software of different versions. To this end, it first tracked software dynamic execution traces to build a dynamic software network model, and then it mapped these traces to a complex network of a dynamic invoke software network. Second, it identified the most vital nodes in two steps, namely the calInitialWeightForNode and calFinalWeightForNode in order to compute the initial weights and the final weights of nodes iteratively. Third, it mined top-k nodes as the original communities to create a framework for detecting new community, and expanded these nodes to the community by the algorithm expandTheCommunity and evaluated the results with criterion Q. Finally, it calculated the modularity of software of different versions. Experimental results show that the most vital nodes are reasonable to be taken in comparison with other ranking measures, and that the analysis approach to the modularity of software network is effective in evaluating software community structure, and can help assist the developer to refactor the software and improve the software quality.

Identifying the most influential spreaders in a weighted complex network is vital for optimizing utilization of the network structure and promoting the information propagation. Most existing algorithms focus on node centrality, which consider more connectivity than clustering. In this paper, a novel algorithm based on clustering degree algorithm (CDA) is proposed to identify the most influential spreaders in a weighted network. First, the weighted degree of a node is defined according to the node degree and strength. Then, based on the node weighted degree, the clustering degree of a node is calculated in respect to the network topological structure. Finally, the propagation capability of a node is achieved by accounting the clustering degree of the node and the contribution from its neighbors. In order to evaluate the performance of the proposed CDA algorithm, the susceptible-infected-recovered model is adopted to simulate the propagation process in real-world networks. The experiment results have showed that CDA is the most effective algorithm in terms of Kendall's tau coefficient and with the highest accuracy in influential spreader identification compared with other algorithms such as weighted degree centrality, weighted closeness centrality, evidential centrality, and evidential semilocal centrality.

Blood pressure (BP) diseases have become one of the major threats to human health. Continuous measurement of BP has proven to be a prerequisite for effective incident prevention. In contrast with the traditional prediction models with low measurement accuracy or long training time, noninvasive BP measurement is a promising use for continuous measurement. Thus in this paper, classification and regression trees (CARTs) are proposed and applied to tackle the problem. First, according to the characteristics of different information, different CART models are constructed. Second, in order to avoid the over-fitting problem of these models, the cross-validation method is used for selecting the optimum parameters so as to achieve the best generalization of these models. Based on the biological data collected from CM400 monitor, this approach has achieved better performance than the common existing models such as linear regression, ridge regression, the support vector machine and neural network in terms of accuracy rate, root mean square error, deviation rate, and Theil inequality coefficient, and the required training time is also comparatively less. With increasing data, the accuracy rate of predicting systolic BP and diastolic BP by CART exceeds 90%, and the training time is less than 0.5 s.

The user interaction in online social networks can not only reveal the social relationships among users in e-commerce systems, but also imply the social preferences of a target user for recommendation services. However, the current research has rarely explored the impact of social interaction on recommendation performance, especially now that recommender systems face increasing challenges and suffer from poor efficiency due to social data overload. Therefore, applied research on user interaction has become increasingly necessary in the field of social recommendation. In this paper, we develop a novel social recommendation method based on the user interaction in complex social networks, called the SRUI model, to present a basis for improving the efficiency of the recommender systems. Specifically, a weighted social interaction network is first mapped to represent the interactions among social users according to the gathered information about historical user behavior. Thereafter, the complete path set is mined by the complete path mining (CPM) algorithm to find social similar neighbors with tastes similar to those of the target user. Finally, the social similar tendencies of the users on the complete paths are obtained to predict the final ratings of items through the SRUI model. A series of experimental results based on two real public datasets show that our approach performs better than other state-of-the-art methods in terms of recommendation performance.

A significant approach for the discovery of biological regulatory rules of genes, protein and their inheritance relationships is the extraction of meaningful patterns from biological sequence data. The existing algorithms of sequence pattern discovery, like MSPM and FBSB, suffice their low efficiency and accuracy. In order to deal with this issue, this paper presents a new algorithm for biological sequence pattern mining abbreviated MpBsmi based on the data index structure. The MpBsmi algorithm employs a sequence position table abbreviated ST and a sequence database index structure named DB-Index for data storing, mining and pattern expansion. The ST and DB-Index of single items are firstly obtained through scanning sequence database once. Then a new algorithm for fast support counting is developed to mine the table ST to identify the frequent single items. Based on a connection strategy, the frequent patterns are expanded and the expanded table ST is updated by scanning the DB-Index. The fast support counting algorithm is used for obtaining the frequent expansion patterns. Finally, a new pruning technique is developed for extended pattern to avoid the generation of unnecessarily large number of candidate patterns. The experiments results on multiple classical protein sequences from the Pfam database validate the performance of the proposed algorithm including the accuracy, stability and scalability. It is showed that the proposed algorithm has achieved the better space efficiency, stability and scalability comparing with MSPM, FBSB which are the two main algorithms for biological sequence mining.

As the quality of crucial entities can directly affect that of software, their identification and protection become an important premise for effective software development, management, maintenance and testing, which thus contribute to improving the software quality and its attack-defending ability. Most analysis and evaluation on important entities like codes-based static structure analysis are on the destruction of the actual software running. In this paper, from the perspective of software execution process, we proposed an approach to mine dynamic noteworthy functions (DNFM)in software execution sequences. First, according to software decompiling and tracking stack changes, the execution traces composed of a series of function addresses were acquired. Then these traces were modeled as execution sequences and then simplified so as to get simplified sequences (SFS), followed by the extraction of patterns through pattern extraction (PE) algorithm from SFS. After that, evaluating indicators inner-importance and inter-importance were designed to measure the noteworthiness of functions in DNFM algorithm. Finally, these functions were sorted by their noteworthiness. Comparison and contrast were conducted on the experiment results from two traditional complex network-based node mining methods, namely PageRank and DegreeRank. The results show that the DNFM method can mine noteworthy functions in software effectively and precisely.