Search

This paper investigates an intelligent reflecting surface (IRS) aided wireless powered communication network (WPCN) with a focus on multicast transmissions in a game-theoretic way, where a transmitter first harvests energy from a power station (PS) and then transmits the information to multiple Internet of Things (IoT) devices in a multicast form. An IRS is deployed to assist the wireless energy transfer (WET) and wireless information transfer (WIT) processes. Considering that the PS and the transmitter belong to different service providers, we propose two schemes based on Stackelberg game, namely IRS-aided transmitter dominant energy trading (IRS-TDET) with the transmitter as the leader and IRS-aided PS dominant energy trading (IRS-PDET) with the PS as the leader. To solve the non-convex optimization problem of the leader-level game in the IRS-TDET scheme, we first derive the closed-form optimal phase shifts in the WET stage, and then use the semidefinite relaxation approach to solve the max-min problem about the phase shifts of the WIT stage. Finally, a low-complexity alternating optimization algorithm is developed to solve the simplified non-convex optimization problem with regard to the energy price and WET time. Numerical results show that the performance of proposed IRS-TDET and IRS-PDET is superior to that of the corresponding non-IRS aided schemes, and the deployment of IRS can effectively improve the utilities of both the PS and the transmitter. In addition, it is preferred to deploy the IRS near the transmitter in WPCN, and the transmitter can be deployed near the PS or the IoT devices to obtain higher utility.

Knowledge graph completion(KGC) has attracted increasing attention in recent years, aiming at complementing missing relationships between entities in a Knowledge Graph(KG). While the existing KGC approaches utilizing the knowledge within KG could only complement a very limited number of missing relations, more and more approaches tend to study the completion of the multi-relationship knowledge graph. However, the existing completion methods of multi-relation knowledge graph regard knowledge graph as an undirected graph, which ignores the directionality of knowledge graph, so that the potential characteristics of multi-relation cannot be learned. Besides, most algorithms fail to explore the local information of knowledge because they ignore the different importance of entity adjacencies. In this paper, we propose to use local information fusion to join the entity and its adjacency relation, to acquiring the multi-relation representation. In addition, we try to specify distinct weights to model the direction of the relationship and apply the attention mechanism between entity nodes to obtain local information between entity nodes. Experiments conducted on three benchmark datasets and a medical domain knowledge graph dataset that we collect demonstrate the effectiveness of the proposed framework.

This article explores cooperative communications for drone swarms which become an essential part of space-air-ground integrated networks (SAGINs), since drones (also called unmanned aerial vehicles) have the advantage of fast and flexible deployment. However, the rapid spread of drone applications leads to a dramatic increase in the spectrum and energy demand for drone communications, resulting in a shortage of spectrum and energy resources. Moreover, the inherent broadcast nature of radio propagation makes the air interface of drone communications extremely vulnerable to jamming and eavesdropping attacks, especially considering the fact that an unknown drone may sneak into a swarm for the purpose of intentionally interfering with or tapping confidential transmissions of the legitimate drones. As a consequence, a cooperative networking architecture is proposed for drone-swarm-aided SAGINs. A heterogeneous spectrum sharing mechanism is presented for drone communications to alleviate the spectrum shortage by dynamically identifying and accessing idle spectrum resources with the aid of spectrum sensing and inference. Physical-layer security is also examined for cooperative drone communications with limited energy supply. Finally, a wide range of potential applications of drone communications are pointed out, including emergency wireless networks, beyond 5G and 6G systems, and smart grid.

Internet of Things (IoT) is the connections of embedded sensing technologies that can be integrated with intelligent transportation systems (ITS), which can take advantage of the advanced sensing fusion to implement the intelligent interaction. Moreover, the environmental wireless signals was harvested and utilized to power the charging sensors, which can be used to assist and enhance the ITS and have attracted extensive attentions in industries. The simultaneous wireless information and power transfer (SWIPT) enabled ITS was designed and proposed with the advanced sensing fusion, where the sensor nodes of the source and destination were all equipped with multiple antennas and the maximum ratio transmission (MRT) precoding and maximum ratio combination (MRC) were adopted respectively in the proposed system, where the sensing nodes can be embedded into the data addressing equipment to fusion the sensing information. For the established massive sensing node transmission system, the transmission power, the number of source antennas and destination antennas were considered to fuse and optimize to obtain the maximum energy efficiency (EE) of the ITS. Since the original fusion problem was non-convex and difficult to solve, which can be transformed to convex problem via the large SNR approximation method and thereby the optimal solution of the sensing fusion problem was derived through the Lagrange algorithm and the Lambert function W. The simulation results showed that the proposed fusion algorithm can achieve remarkable EE compared with other algorithms and be utilized to the massive sensors to improve the intelligent predication and transportation for vehicle network.

The progress on human pose estimation by deep neural networks has been significantly advanced in recent years. However, the problem of precision loss caused by the prediction of the coordinates back to the original image has been neglected. In this paper, we propose a simple but effective method using Heatmap and Offset for Pose Estimation (HOPE). In order to solve the human pose estimation problem, firstly a general top-down method is used in HOPE to generate the human detection box based on a detector, and then the keypoints in each cropped box image are located. To alleviate the precision loss of mapping process, HOPE embeds the coordinate offset into the structure of the neural network, allowing the network to self-learn the slight offset in the mapping process in an end-to-end manner, which improves the accuracy in the current field of pose estimation. Experimental results on the multi-person pose estimation dataset MSCOCO, the single-person pose estimation dataset MPII and CrowdPose Pose Estimation dataset indicate that our method achieves state-of-the-art performance in terms of accuracy and computational complexity.

Knowledge graph completion has become a well-studied problem and a non-trivial task with the broad application of the knowledge graphs. Previously, a lot of works have been proposed to solve the knowledge graph completion problem, for example, a series of Trans model, semantic matching models, convolutional neural networks based methods and so on. However, a series of Trans models and semantic matching models only focused on the shadow information of the knowledge graph, thus failed to capture the implicit fine-grained feature in the triple of knowledge graphs; convolutional neural networks based methods learned more expressive feature for knowledge graph completion, and it also ignored the directional relation characteristic and implicit fine-grained feature in the triple. In this paper, we propose a novel knowledge graph completion model named directional multi-dimensional attention convolution model that explores directional information and an inherent deep expressive characteristic of the triple. At last, we evaluate our directional multi-dimensional attention convolution model based on three standard evaluation criteria in two robust datasets, and the experiment shows that our model achieves state-of-the-art MeanRank.

Action recognition based on 3D skeleton data has attracted much attention due to its wide application, and it is one of the most popular research topics in computer vision. The 3D skeleton data is an effective representation of motion dynamics and is not easily affected by light, scene variation, etc. Previous research on action recognition has mainly focused on video or RGB data methods. In recent years, the advantages of combining skeleton data and deep learning have been gradually demonstrated, many impressive methods have been proposed, especially GCN‐based methods. In this survey, we first introduce the development process of 3D skeleton‐data action recognition and the classification of graph convolutional network, then introduce the commonly used NTU RGB + D and NTU RGB + D 120 datasets. Finally, a detailed review of existing variants of three mainstream technologies is provided based on deep learning and their performance was compared from three dimensions. To the best of our knowledge, this is the first research to integrate the research of GCN‐based method and its various evolutionary methods. Comparative investigation of existing variants of research in action‐recognition task from different perspectives is made, a generic framework is described, state‐of‐the‐art practices are summarized, and the emerging trends of this topic are explored.

This paper investigates the secrecy energy efficiency (SEE) for a two-tier heterogeneous cellular network having a macro cell and multiple small cells, where the macro cell consists of a macro base station (MBS), a macro user and an eavesdropper and meanwhile each small cell comprises of a small base station (SBS), a small user and an eavesdropper. We consider the use of artificial noise (AN) signals emitted at MBS and SBSs at the expense of transmit power to defend against eavesdroppers. For the sake of obtaining a better balance between the secrecy rate (SR) and energy efficiency of heterogeneous cellular networks, we propose an SEE maximization (SEEM) scheme in small cells through jointly optimizing the beamformers of confidential signals and AN at MBS and SBSs, called joint beamforming based SEEM (JBF-SEEM) while guaranteeing the SR requirements and the power constraints for MBS and each SBS. However, the proposed JBF-SEEM problem is a complicated optimization problem because of the non-convex fractional form. Thus, we firstly transform the original objective function in fractional form into a subtractive one through applying the Dinkelbach's method, and then the non-convex subtractive form is approximated into a convex optimization problem by combining the penalty function and the difference of two-convex functions approach. Furthermore, we propose a two-tier iterative JBF-SEEM algorithm to obtain the near-optimal beamforming solution. Numerical results show that the JBF-SEEM algorithm is convergent while achieving a better SEE performance than conventional approaches.

A protein complex is a group of associated polypeptide chains which plays essential roles in the biological process. Given a graph representing protein-protein interactions (PPI) network, it is critical but non-trivial to detect protein complexes, the subsets of proteins that are tightly coupled, from it. Network embedding is a technique to learn low-dimensional representations of vertices in networks. It has been proved quite useful for community detection in social networks in recent years. However, unlike social networks, PPI network does not contain rich metadata, so that existing network embedding methods cannot fully capture the network structure of PPI to improve the effect of protein complexes detection significantly. We propose a semi-supervised network embedding model by adopting graph convolutional networks to detect densely connected subgraphs effectively. We compare the performance of our model with state-of-the-art approaches on three popular PPI networks with various data sizes and densities. The experimental results show that our approach significantly outperforms other approaches on all three PPI networks.

Entropy has been widely measured the complexity of complex networks. At present, many measures about entropies were defined based on the directed connection of nodes. The similarity of nodes can better represent the relationship among all nodes in complex networks. In the definition of similarity of nodes, the importance of a node in the network depends not only on the degree of the node itself, but also on the extent of dependence of neighboring nodes on the node. In this paper, we proposed a new structure entropy based on nonextensive statistical mechanics and similarity of nodes. In the proposed method, the similarity of nodes and the betweenness of nodes are both quantified. The proposed method considers the extent of dependence between neighbouring nodes. For some complex networks, the proposed structure entropy can distinguish complexity of that while other entropies can not be. Meanwhile, we construct five ER random networks and small-world networks and some real-world complex networks such as the US Air Lines networks, the GD'01-GD Proceedings Self-Citing networks, the Science Theory networks, the Centrality Literature networks and the Yeast networks are measured by the proposed method. The results illustrated our method for quantifying the complexity of complex networks is effective.

This article investigates the physical layer security for a cooperative cognitive radio (CCR) system with energy-harvesting (EH) technique, which consists of a cognitive source (CS), multiple cognitive relays (CRs) and a cognitive destination (CD) as well as multiple eavesdroppers (Es) who are considered to tap the confidential transmission from CS via CRs to CD. Both CS and CRs are equipped with energy-harvesters to harvest energy from the surrounding radio frequency environment. Additionally, in order to guarantee the quality-of-service of primary communications, the transmit powers of CS and CRs are limited by the maximum tolerant power at primary destination. Depending on the availability of channel state information (CSI) of wireless energy links from primary transmitter (PT) to CRs, we respectively propose non-energy aware relay selection (NEARS) and energy aware relay selection (EARS) schemes for protecting the transmission from CS via CRs to CD from leaking to Es. We consider the use of the security-reliability tradeoff (SRT) to measure the performance of NEARS and EARS schemes. For the purpose of comparison, we also carry out the analysis of SRT performance for conventional direct communication (DC). Furthermore, the closed-form expressions of outage probability and intercept probability for NEARS and EARS schemes are derived and numerical results show that the SRT performance of EARS is better than that of DC and NEARS schemes.

In this paper, we consider an orthogonal frequency division multiplexing (OFDM) based heterogeneous cellular network consisting of a macro cell and a small cell, where a macro-cell base station (MBS) and a small-cell base station (SBS) communicate with respective macro-cell user (MU) and small-cell user (SU) with the existence of a common eavesdropper across different OFDM subcarriers. In order to make full use of the mutual interference between MBS and SBS against eavesdropping, an artificially designed signal is emitted at MBS to cancel out the interference received at MU as well as to interfere with the common eavesdropper at each subcarriers. For the purpose of further improving the secrecy performance, a sum secrecy rate (SSR) of OFDM-based heterogeneous cellular networks is maximized by optimizing the power allocation between MBS and SBS across different OFDM subcarriers when the eavesdropper owns the global instantaneous channel state information (ICSI), thus called ICSI based SSR maximization (ICSI-SSRM). As for the case when the ICSI of wiretap channels is unknown, we propose a statistical CSI based SSR maximization (SCSI-SSRM) scheme, where the statistical characteristics of channels from MBS and SBS to eavesdropper are employed to optimally allocate powers of MBS and SBS across different subcarriers under the constraint of MBS's total transmit power. The formulated ICSI-SSRM and SCSI-SSRM problems are all non-convex due to form of the difference between two-convex (D.C.) functions. Thus, we utilize the D.C. approximation approach to respectively convert original optimization problems into convex problems. Moreover, iterative optimal power allocation algorithms for ICSI-SSRM and SCSI-SSRM schemes are also presented to obtain their respective SSR values. Simulation results illustrate that the ICSI-SSRM and SCSI-SSRM algorithms can converge to their optimal values, which confirms the correctness and validation of the proposed algorithms. In addition, numerical results are also given to show that the ICSI-SSRM and SCSI-SSRM schemes outperform conventional power allocation methods in terms of their SSR performance.

In this paper, we consider an energy-harvesting (EH) relay system consisting of a source, a destination, and multiple EH decode-and-forward (DF) relays that can harvest the energy from their received radio signals. A power-splitting ratio is employed at an EH DF relay to characterize a trade-off between an energy used for decoding the source signal received at the relay and the remaining energy harvested for retransmitting the decode outcome. We propose an optimal power splitting and relay selection (OPS-RS) framework and also consider the conventional equal power splitting and relay selection (EPS-RS) for comparison purposes. We derive closed-form expressions of outage probability for both the OPS-RS and EPS-RS schemes and characterize their diversity gains through an asymptotic outage analysis in the high signal-to-noise ratio region. We further examine an extension of our OPS-RS framework to an energy-harvesting battery (EHB) enabled cooperative relay scenario, where the EH relays are equipped with batteries used to store their harvested energies. We propose an amplify-and-forward (AF) based EHB-OPS-RS and a DF based EHB-OPS-RS schemes for AF and DF relay networks, respectively. Numerical results show that the proposed OPS-RS always outperforms the EPS-RS in terms of outage probability. Moreover, the outage probabilities of AF and DF based EHB-OPS-RS schemes are much smaller than that of OPS-RS and EPS-RS methods, demonstrating the benefit of exploiting the batteries in EH relays. Additionally, the DF based EHB-OPS-RS substantially outperforms the AF based EHB-OPS-RS and such an outage advantage becomes more significant, as the number of EH relays increases., Comment: 14 pages

This paper proposes a method of measuring the powertrain noise, which enables for the indoor prediction of the sound pressure level (SPL) of pass-by noise in an early stage of vehicle development phase. The proposed method is implemented in two steps. Firstly, the indoor tyre/roller noise is obtained by the indoor coast-by experiment based on the proposed single coast-by approach. Secondly, the powertrain noise is computed by subtracting the tyre/roller noise from the total vehicle noise. Then the powertrain noise is synthesized with the outdoor tyre/road noise energetically to predict the level of pass-by noise, with the purpose of reducing the influence caused by the difference between indoor and outdoor tyre/road noise. The proposed method is validated by experimental comparisons with the existing methods, shielding method and slick-tyre experiment method, respectively. The results show that the proposed method not only produces better prediction accuracy than the existing ones, but also has the ability to predict the pass-by noise SPL for a realistic case in an evaluation time of less than one hour.

With the development of society, under the influence of high technology such as big data, cloud computing, network technology, and mobile Internet, education has gradually got rid of the traditional teaching methods and the limitations of traditional teaching methods. This article takes college students as the research theme. In order to obtain more complete survey data, the survey is based on an online questionnaire. In this online questionnaire survey, a total of 11 computer classroom teachers from colleges and universities at the city, county, and town levels were invited to supplement the paper questionnaire survey. After checking and verifying the retrieved questionnaire, the excel software technology and statistical data will be used to effectively analyze the questionnaire. Experimental research shows that 84% of teachers believe that big data education is related to computer teaching. They believe that the use of information technology to collect computer learning behaviors in colleges and universities can help improve the efficiency or quality of computer teaching, while only 16% of teachers are concerned about big data. The impact on computer teaching is unclear. Research confirms that high-data technology is related to computer classroom teaching, and the use of information technology to collect computer-based learning behavior can help improve the effectiveness or quality of computer classroom teaching. Big data technology has great potential in computer classroom teaching and will provide practical theoretical support and help for computer classroom teaching. Although the application of big data related technologies in the field of education in my country is still in the initial stage, it is still of great significance for deepening the reform of education and teaching.

Due to the gradual improvement of computing power and the development of big data, the error rate of voice recognition systems and voice assistance functions gradually decreases with the gradual application of deep learning. In terms of commercial contracts, it is a legally binding text. This article first analyzes the theory of functional equivalence translation and then analyzes the theory of functional equivalence translation from the perspective of professionalism and accuracy and determines the principles of business contract English vocabulary translation. At the same time, it studies English in terms of vocabulary equivalence. In the process of teaching trade English, some common vocabulary in trade English is easy to be misunderstood. This article analyzes these common translation errors and their main causes and proposes some appropriate solutions. Business English vocabulary translation must follow the principle of pragmatic equivalence, and the translation must be flexible to deal with communication obstacles caused by cultural background, language habits, and other factors, because they are subject to two different language and cultural characteristic rules (English and Chinese). The seasonal autoregressive summed moving average (ARIMA) model is used to simulate the time series of the surface water temperature in the coastal waters, and the structure of the model is determined through the principle of irrelevance and simplicity. On this basis, the optimal ARIMA prediction model is established and the surface water temperature of the coastal waters is monthly. The prediction results are consistent with the monthly change trend of the sea surface temperature, indicating that the model is suitable for predicting the offshore surface temperature.

To improve detection performance, this paper proposes a non-cooperative unmanned aerial vehicle (UAV) detection strategy based on multichannel detection of remote signal with energy detector (ED). More specifically, the sampling point of remote signal on each subchannel adaptively varies with environmental signal-to-noise (SNR) within constrained scope. The decision on the presence or absence of remote signal is made at the fusion center (FC) based on the majority voting rule. Both theoretical derivation and simulation experiments validate the effectiveness of the proposed scheme.

We introduce a unified system for SCHOLAT to import information into Encyclopedia, called Semi-automatic Scholar Encyclopedia Generating System, which consists of four modules: website display page, backstage administration module, version comparison and website security module. Scholar Encyclopedia is implemented as a subsystem of Semi-automatic Scholar Encyclopedia Generating System. Scholar Encyclopedia is a scholar information website that has been independently compiled by the scholars and filled in by personal information. Scholar Encyclopedia Import Model consists of three modules: user evaluation, information import and manual review. At present, Semi-automatic Scholar Encyclopedia Generating System has been used, and the beta version of Scholar Encyclopedia has been deployed in our school network center. The online application shows that Scholar Encyclopedia Import Model has good accuracy and practicability.

In this paper, we consider a multi-relay cooperative communication network that consists of a source node transmitting to its destination with the help of multiple decode-and-forward (DF) relays. Specifically, the DF relays that succeed in decoding the source signal are allowed to re-transmit their decoded results simultaneously to the destination in a cooperative beamforming manner. In order to carry out the cooperative beamforming, the destination needs to send the quantized channel state information (CSI) to the relays through a limited feedback channel in the face of channel quantization errors (CQE). We propose a CQE oriented multi-relay beamforming (MRB) scheme, denoted CQE-MRB for short, for the sake of improving the throughput of relay-destination transmissions. An effective throughput defined as the difference between the transmission rate and the feedback rate is used to measure an outage probability of the source-destination transmission. Simulation results demonstrate that the outage performance of proposed CQE-MRB scheme is improved substantially with an increasing number of relays. Moreover, it is shown that the number of channel quantization bits can be further optimized to minimize the outage probability of proposed CQE-MRB scheme.

Knowledge graphs have shown increasing importance in broad applications such as question answering, web search, and recommendation systems. The objective of knowledge graph embedding is to encode both entities and relations of knowledge graphs into continuous low-dimensional vector spaces to perform various machine learning tasks. Most of the existing works only focused on the local structure of knowledge graphs when utilizing structural information of entities, which may not sincerely preserve the global structure of knowledge graphs.In this paper, we propose a semi-supervised model by adopting graph convolutional networks to utilize both local and global structural information of entities. Specifically, our model takes textual information of each entity into consideration as entity attributes in the process of learning. We show the effectiveness of our model by applying it to two traditional tasks for knowledge graph: entity classification and link prediction. Experimental results on two well-known corpora reveal the advantages of this model compared to state-of-the-art methods on both tasks. Moreover, the results show that even with only 1% labeled data to train, our model can still achieve good performance.

In this paper, we investigate physical-layer security for a heterogeneous cellular network consisting of a macro cell and a small cell in the presence of a passive eavesdropper that intends to tap both the macro-cell and small-cell transmissions. Both the orthogonal spectrum sharing (OSS) and non-orthogonal spectrum sharing (NOSS) are considered for the heterogeneous cellular network. The OSS allows the macro cell and small cell to access a given spectrum band in an orthogonal manner, whereas the NOSS enables them to access the same spectrum simultaneously and mutual interference exits. As a consequence, we present two NOSS schemes, namely the interference-limited NOSS (IL-NOSS) and interference-canceled NOSS (IC-NOSS), where the mutual interference is constrained below a tolerable level in the IL-NOSS through power control and the IC-NOSS scheme exploits a specially-designed signal for canceling out the interference received a legitimate cellular user while confusing the passive eavesdropper. We derive closed-form expressions for an overall secrecy outage probability of the OSS, IL-NOSS, and IC-NOSS schemes, which take into account the transmission secrecy of both the macro cell and small cell. We further characterize the secrecy diversity of OSS, IL-NOSS and IC-NOSS schemes through an asymptotic secrecy outage analysis in the high signal-to-noise ratio region. It is shown that the OSS and IL-NOSS methods obtain the same secrecy diversity gain of zero only, and a higher secrecy diversity gain of one is achieved by the IC-NOSS scheme. Additionally, numerical results demonstrate that the IC-NOSS scheme significantly performs better than the OSS and IL-NOSS methods in terms of the overall secrecy outage probability., 32 pages

Modeling the interests of authors over time from documents has important applications in broad applications such as recommendation systems, authorship identification and opinion extraction. In this paper, we propose an Ordering-sensitive and Semantic-aware Dynamic Author Topic Model (OSDATM), which monitors the evolution of author interest in time-stamped documents. The model further uses the discovered author interest information to discover better topics. Unlike traditional topic models, OSDATM is sensitive to the ordering of words, thus it extracts more information from the semantic meaning of the context. The experimental results show that OSDATM learns better topics than state-of-the-art topic models. In addition, the dynamic interests of authors that the OSDATM model discovers are interpretable and consistent with the truth.

In the last decades, many similarity measures are proposed, such as Jaccard coefficient, cosine similarity, BM25, language model, etc. Despite the effectiveness of the existing similarity measures, we observe that none of them can consistently outperform the others in most typical situations. Choosing which similarity predicate to use is usually treated as an empirical question by evaluating a particular task with a number of different similarity predicates, which is not computationally efficient and the obtained results are not portable. In this paper, we propose a novel approach to combine different similarity predicates together to form a committee so that we do not need to worry about choosing which of them to use. Empirically, we can obtain a better result than any individual similarity predicate, which is quite meaningful in practice. Specifically, our method models the problem of committee generation as a 0–1 integer programming problem based on the confidence of similarity predicates and the reliability of attributes. We demonstrate the effectiveness of our model by applying it on three datasets with controlled errors. Experimental results demonstrate that our similarity predicate committee is more robust and superior over existing individual similarity predicates.

Community question answering (CQA) is an important but challenging task. Meantime, as the theory of deep learning develops, remarkable progress has been made by deep neural networks. This paper studies an advanced deep neural network that not only uses external knowledge to learn better representations of questions and answers but also improves representation learning by considering question categorization as an auxiliary task. Specifically, we propose a novel Multi-task and Knowledge enhanced Multi-head Interactive Attention network for Community Question Answering (MKMIA-CQA). It contains a document modeling module responsible for utilizing external commonsense knowledge to help identify background information (entity mentions and their relations) and filter out noise information from the long text which has complicated semantic and syntactic structures. Moreover, the model is trained in a multi-task manner. It regards community question answering as the primary task and question categorization as the auxiliary task, which aims to learn a category-aware encoder and improve the quality of locating the salient information of a long question. The experimental results on three widely used CQA datasets demonstrate that our model achieves impressive results compared to other strong competitors.

Unlike English and other Western languages, many Asian languages such as Chinese and Japanese do not delimit words by space. Word segmentation and new word detection are therefore key steps in processing these languages. Chinese word segmentation can be considered as a part-of-speech (POS)-tagging problem. We can segment corpus by assigning a label for each character which indicates the position of the character in a word (e.g., “B” for word beginning, and “E” for the end of the word, etc.). Chinese word segmentation seems to be well studied. Machine learning models such as conditional random field (CRF) and bi-directional long short-term memory (LSTM) have shown outstanding performances on this task. However, the segmentation accuracies drop significantly when applying the same approaches to out-domain cases, in which high-quality in-domain training data are not available. An example of out-domain applications is the new word detection in Chinese microblogs for which the availability of high-quality corpus is limited. In this paper, we focus on out-domain Chinese new word detection. We first design a new method Edge Likelihood (EL) for Chinese word boundary detection. Then we propose a domain-independent Chinese new word detector (DICND); each Chinese character is represented as a low-dimensional vector in the proposed framework, and segmentation-related features of the character are used as the values in the vector.

Despite that neural networks have demonstrated their good potential to be used in constructing learners which exhibit strong predictive performance, there are still some uncertainty issues that can greatly affect the effectiveness of the employed supervised learning algorithms, such as class imbalance and labeling errors (or class noise). Technically, imbalanced data resource can cause more difficulties or limitations for learning algorithms to distinguish different classes, while data with labeling errors can lead to an unreasonable problem formulation due to incorrect hypotheses. Indeed, noise and class imbalance are pervasive problems in the domain of educational data analytics. This study aims at developing improved randomized learning algorithms by investigating a novel type of cost function that focuses on the combined effects of class imbalance and class noise. Instead of concerning these uncertainty issues isolation, we present a convex combination of robust and imbalanced modelling objectives, contributing to a generalized formulation of weighted least squares problems by which the improved randomized learner models can be built. Our experimental study on several educational data classification tasks have verified the advantages of our proposed algorithms, in comparison with some existing methods that either takes no account of class imbalance and labeling errors, or merely consider one specific aspect in problem-solving.

With the broad adoption of the global positioning system-enabled devices, the massive amount of trajectory data is generated every day, which results in meaningful traffic patterns. This study aims to predict the future state of road traffic instead of merely showing the current traffic condition. First, we use a clustering method to group similar trajectories of a particular period together into a cluster for each road. Second, for each cluster, we average the lengths and angles of the entire trajectories in the group as the representative trajectory, which is regarded as the road pattern. Third, we create a feature vector for each road based on its historical traffic conditions and neighbor road patterns. Finally, we design a recurrent convolutional neural network for modeling the complex nonlinear relationship among features to predict road traffic conditions. Experimental results show that our approach performs more favorably compared with several traditional machine learning and state-of-the-art algorithms.

Autonomous underwater vehicle(AUV) plays an important role in modern aquaculture. For example, it can be applied to monitor water quality as well as the fish biomass. In these activities, path planning is significant for AUV. This paper proposed a path planning algorithm for underwater vehicles based on improved rapidly expanding random tree(RRT). The artificial potential field(APF) algorithm is introduced into the heuristic function of bidirectional RRT to improve the expansion efficiency as a global path. To solve the problem of dynamic obstacles, global path planning and local path planning are combined. The dynamic window method is used in local path planning to deal with dynamic obstacles. This approach takes the robot’s speed as the reference point. It sets the window and then selects a better path by the evaluation function. In this way, the path planning of underwater vehicle in a dynamic environment is realized. Finally, the algorithm was simulated in both static and dynamic environments. The results show that the proposed algorithm can provide effective paths for underwater vehicles quickly.

The smart contract is a self-executing code that is managed by blockchain nodes, providing coordination and enforcement framework for agreements between network participants. However, smart contracts are not particularly “smart” since the virtual machine (VM) does not support the running environment of machine learning. To make the smart contract smarter, we propose a decentralized blockchain oracle framework to support smart contract training the machine learning model. In view of malicious nodes which may attack the process of training, we propose a consensus algorithm to prevent the malicious attack from malicious nodes. At the end of this paper, we do an experiment on two datasets: MNIST and CIFAR10. The result shows that our framework can prevent malicious attack efficiently and keep high accuracy. With our proposed framework, smart contracts have an ability to train or call machine learning model, making a smart contracts “smarter”.

This work-in-progress (WIP) paper in the research track explored the muddiest point and self-efficacy for students in introductory computer science courses, starting with one course at a southern public institution. With the increasing demand in computing-related careers, the increased enrollments in undergraduate computer science courses are facing challenges to ensure the average passing rate. In this work, we applied the muddiest point to capture student’s perceptions of concepts of difficulty and used self-efficacy to better contextualize students’ perfections of self-efficacy in relation to concepts of difficulty. In this paper we present the results of one section of one introductory computing course, that revealed the perceived difficult topics as well as changes in self-efficacy throughout the semester. Furthermore, a positive correlation was found between self-efficacy and performance. Keywords: self-efficacy, muddiest points, student performance, computer science education

This research full paper describes how web scraping and natural language processing can be utilized to answer complex questions in computer science education. In this work, we apply connectivism as the theoretical framework, and demonstrate how web scraping can be useful for extrapolating large amounts of data from publicly available web pages to pool data from a wider array of sources and to further knowledge in the field. In addition, we discuss how natural language processing can be used to reliably obtain salient information from textual data, and how it can complement qualitative analysis. To illustrate these techniques in practice, we provide a specific application in which we examine the current trends in the job market for computer science students. The information gathered in this example provides additional areas for educational consideration, such as offering students Python programming language and machine learning. Also, the job postings delineate a clear need for applicants to exhibit programming and testing skills. Although programming may be taught already, testing is widely considered a knowledge deficiency, which suggests that educators should consider placing an increased emphasis on this area to ensure their students are adequately prepared for their career endeavors, and able to transfer the knowledge taught to critically assess and debug their own programs.

The operating status of power equipment is closely related to the stability and safety of the power system. Comprehensively and accurately grasp the internal and external multi-source data of power equipment, and use scientific means to summarize and integrate the information, so as to make accurate assessment and prediction of the operating status and change trend of the equipment, and arrange a reasonable operation and maintenance plan, Is the key prerequisite and important foundation for the reliable and economic operation of the entire power system. The development and breakthrough of artificial intelligence technology (AI) in the aspects of efficiency, precision, and self-learning ability in identification, prediction, optimization, and decision-making tasks provide a new technical means and research ideas for the operation and maintenance of power equipment. The paper reviewed the key technologies of artificial intelligence in the operation and maintenance of power transmission and transformation equipment, sorted out the relevant data situation and existing problems of equipment operation and maintenance, pointed out the key technology framework of artificial intelligence and the application characteristics and scope of various algorithms. Finally, the problems faced by the application of artificial intelligence technology are analysed, and their development trends are prospected and suggested.

Software testing is an important part of the software development life cycle, and it is also the guarantee of software quality and reliability for a long time now and in the future. The main method of sex. In recent years, people have begun to try to apply artificial intelligence technology to the problem of test case generation, resulting in an evolutionary test method. The article introduces the current research status of the automatic generation method of test cases based on structured, focusing on the evolutionary test, and compares and summarizes its application based on genetic algorithm. The application results show that the evolutionary testing method is currently the most efficient method of automatically generating software test cases. Finally, some opinions on the further research direction of evolutionary testing are put forward.

With the widespread application of traditional electric power customer service in recent years, the scale of users continues to increase, and the traffic pressure on the communication service hotlines of power users such as 95598, 1000 service hotlines, and power dispatching command systems will increase sharply. However, manual agents and traditional self-service voice response systems are limited by factors such as customer service manpower, working time, knowledge level, and volume of traffic consultation due to the use of button interaction methods, which seriously affect the customer experience. In this paper, core technologies such as intelligent speech recognition, semantic understanding and speech synthesis are studied, and the application of current intelligent speech analysis technology in electric customer service is introduced.

A dynamic Smagorinsky&ndash, Lilly model (DSLM) subgrid model on the basis of the Smagorinsky&ndash, Lilly subgrid model (SLM) was introduced in the OpenFOAM software. The flow field of the vehicle was simulated, and the pressure coefficient and sound pressure curve of the monitoring points were compared with the wind tunnel test results. The results show that the DSLM subgrid model with a wall function can achieve high simulation accuracy. The investigation of the flow field structure revealed an intermittent detachment of the turbulent vortex after the airflow passed through the rearview mirror, thereby resulting in a violent pressure pulsation on the side window around the rearview mirror. Airflow passed through the A-pillar, separated, and reattached on the upper side window, thereby producing aerodynamic noise. The research results can serve as a good reference for the simulation and test of aerodynamic noises outside the vehicle, and for the reduction of the aerodynamic noises of vehicles.

Background It is a crucial task of brain science researches to explore functional connective maps of Biological Neural Networks (BNN). The maps help to deeply study the dominant relationship between the structures of the BNNs and their network functions. Results In this study, the ideas of linear Granger causality modeling and causality identification are extended to those of nonlinear Granger causality modeling and network structure identification. We employed Radial Basis Functions to fit the nonlinear multivariate dynamical responses of BNNs with neuronal pulse firing. By introducing the contributions from presynaptic neurons and detecting whether the predictions for postsynaptic neurons’ pulse firing signals are improved or not, we can reveal the information flows distribution of BNNs. Thus, the functional connections from presynaptic neurons can be identified from the obtained network information flows. To verify the effectiveness of the proposed method, the Nonlinear Granger Causality Identification Method (NGCIM) is applied to the network structure discovery processes of Spiking Neural Networks (SNN). SNN is a simulation model based on an Integrate-and-Fire mechanism. By network simulations, the multi-channel neuronal pulse sequence data of the SNNs can be used to reversely identify the synaptic connections and strengths of the SNNs. Conclusions The identification results show: for 2–6 nodes small-scale neural networks, 20 nodes medium-scale neural networks, and 100 nodes large-scale neural networks, the identification accuracy of NGCIM with the Gaussian kernel function was 100%, 99.64%, 98.64%, 98.37%, 98.31%, 84.87% and 80.56%, respectively. The identification accuracies were significantly higher than those of a traditional Linear Granger Causality Identification Method with the same network sizes. Thus, with an accumulation of the data obtained by the existing measurement methods, such as Electroencephalography, functional Magnetic Resonance Imaging, and Multi-Electrode Array, the NGCIM can be a promising network modeling method to infer the functional connective maps of BNNs.

Knowledge tracing is a well-established problem and non-trivial task in personalized education. In recent years, many existing works have been proposed to handle the knowledge tracing task, particularly recurrent neural networks based methods, e.g., Deep Knowledge Tracing (DKT). However, DKT has the problem of vibration in prediction outputs. In this paper, to better understand the problem of DKT, we utilize a mathematical computation model named Finite State Automaton(FSA), which can change from one state to another in response to the external input, to interpret the hidden state transition of the DKT when receiving inputs. And we discover the root cause of the two problems is that the DKT can not handle the long sequence input with the help of FSA. Accordingly, we propose an effective attention-based model, which can solve the above problem by directly capturing the relationships among each item of the input regardless of the length of the input sequence. The experimental results show that our proposed model can significantly outperform state-of-the-art approaches on several well-known corpora.

This paper describes our system submitted to task 4 of SemEval 2020: Commonsense Validation and Explanation (ComVE) which consists of three sub-tasks. The task is to directly validate the given sentence whether or not to make sense and require the model to explain it. Based on BERT architecture with the multi-task setting, we propose an effective and interpretable “Explain, Reason and Predict” (ERP) system to solve the three sub-tasks about commonsense: (a) Validation, (b) Reasoning, and (c) Explanation. Inspired by cognitive studies of common sense, our system first generates a reason or understanding of the sentences and then choose which one statement makes sense, which is achieved by multi-task learning. During the post-evaluation, our system has reached 92.9% accuracy in subtask A (rank 11), 89.7% accuracy in subtask B (rank 9), and BLEU score of 12.9 in subtask C (rank 8).

Review text has been widely studied in traditional tasks such as sentiment analysis and aspect extraction. However, to date, no work is toward the end-to-end abstractive review summarization that is essential for business organizations and individual consumers to make informed decisions. This study takes the lead to study the aspect/sentiment-aware abstractive review summarization in domain adaptation scenario. Our novel model Abstractive review Summarization with Topic modeling and Reinforcement deep learning (ASTR) leverages the benefits of the supervised deep neural networks, reinforcement learning, and unsupervised probabilistic generative model to strengthen the aspect/sentiment-aware review representation learning. ASTR is a multi-task learning system, which simultaneously optimizes two coupled objectives: domain classification (auxiliary task) and abstractive review summarization (primary task), in which a document modeling module is shared across tasks. The main purpose of our multi-task model is to strengthen the representation learning of documents and safeguard the performance of cross-domain abstractive review summarization. Specifically, ASTR consists of two key components: (1) a domain classifier, working on datasets of both source and target domains to recognize the domain information of texts and transfer knowledge from the source domain to the target domain. In particular, we propose a weakly supervised LDA model to learn the domain-specific aspect and sentiment lexicon representations, which are then fed into the neural hidden states of given reviews to form aspect/sentiment-aware review representations; (2) an abstractive review summarizer, sharing the document modeling module with the domain classifier. The learned aspect/lexicon-aware review representations are fed into a pointer-generator network to generate aspect/sentiment-aware abstractive summaries of given reviews by employing a reinforcement learning algorithm. We conduct extensive experiments on real-life Amazon reviews to evaluate the effectiveness of our model. Quantitatively, ASTR achieves better performance than the state-of-the-art summarization methods in terms of ROUGE score and human evaluation in both out-of-domain and in-domain setups. Qualitatively, our model can generate better sentiment-aware summarization for reviews with different categories and aspects.

The particle swarm optimization (PSO) algorithm is prone to local optimal position in the late iteration when solving the reactive power optimization problem of active distribution network. Aiming at the above problems, a reactive power optimization algorithm, improved quantum particle swarm optimization (IQPSO), is proposed in this paper. Based on quantum particle swarm optimization (QPSO) algorithm, the inverse point of worst particle position in the iteration process is introduced into the update formula of particle position, which makes the particle jump out of the local optimal position constraint in the iterative process. In this paper, an optimization model is established with the aim of minimum voltage deviation and system loss in a typical 34 node system, and the effectiveness of proposed method is verified by comparing with QPSO.

Principal component analysis is an unsupervised machine learning algorithm for linear dimensionality reduction of data. The essence of principal component analysis is to solve the problem of maximum variance of projection point. Generally, the gradient rise method is used to solve the optimal parameters of the objective function, but it is easy to fall into the local maximum trap. In this paper, the principal component analysis method with Gaussian quantum particle swarm is proposed to overcome the shortcomings of gradient ascent method, which makes the algorithm independent of gradient and has stronger global optimization performance.

Open-domain conversation is one of the most challenging artificial intelligence problems, which involves language understanding, reasoning, and the utilization of common sense knowledge. The goal of this paper is to further improve the response generation, using personalization criteria. We propose a novel method called PRGDDA (Personalized Response Generation by Dual-learning based Domain Adaptation) which is a personalized response generation model based on theories of domain adaptation and dual learning. During the training procedure, PRGDDA first learns the human responding style from large general data (without user-specific information), and then fine-tunes the model on a small size of personalized data to generate personalized conversations with a dual learning mechanism. We conduct experiments to verify the effectiveness of the proposed model on two real-world datasets in both English and Chinese. Experimental results show that our model can generate better personalized responses for different users.

Automatic real-time summarization of massive document streams on the Web has become an important tool for quickly transforming theoverwhelming documents into a novel, comprehensive and concise overview of an event for users. Significant progresses have been made in static text summarization. However, most previous work does not consider the temporal features of the document streams which are valuable in real-time event summarization. In this paper, we propose a novel M ultitask learning A lgorithm for Web-scale R eal-time E vent S ummarization (MARES), which leverages the benefits of supervised deep neural networks as well as a reinforcement learning algorithm to strengthen the representation learning of documents. Specifically, MARES consists two key components: (i) A relevance prediction classifier, in which a hierarchical LSTM model is used to learn the representations of queries and documents; (ii) A document filtering model learns to maximize the long-term rewards with reinforcement learning algorithm, working on a shared document encoding layer with the relevance prediction component. To verify the effectiveness of the proposed model, extensive experiments are conducted on two real-life document stream datasets: TREC Real-Time Summarization Track data and TREC Temporal Summarization Track data. The experimental results demonstrate that our model can achieve significantly better results than the state-of-the-art baseline methods.

Online prediction is a process that repeatedly predicts the next element in the coming period from a sequence of given previous elements. This process has a broad range of applications in various areas, such as medical, streaming media, and finance. The greatest challenge for online prediction is that the sequence data may not have explicit features because the data is frequently updated, which means good predictions are difficult to maintain. One of the popular solutions is to make the prediction with expert advice, and the challenge is to pick the right experts with minimum cumulative loss. In this research, we use the forex trading prediction, which is a good example for online prediction, as a case study. We also propose an improved expert selection model to select a good set of forex experts by learning previously observed sequences. Our model considers not only the average mistakes made by experts, but also the average profit earned by experts, to achieve a better performance, particularly in terms of financial profit. We demonstrate the merits of our model on two real major currency pairs corpora with extensive experiments.

Discovering useful hidden patterns from learner data for online learning systems is valuable in education technology. Studies on personalized learning full-path recommendation are particularly important for the development of advanced E-learning systems. In this paper, we present a novel model of full-path learning recommendation. This model relies on clustering and machine learning techniques . Based on a feature similarity metric on learners, we first cluster a collection of learners and train a long short-term memory (LSTM) model in order to predict their learning paths and performance. Personalized learning full-paths are then selected from the results of path prediction. Finally, a suitable learning full-path is recommended specifically to a test learner. In this study, a series of experiments have been carried out against learning resource datasets. By comparisons, experimental results indicate that our proposed methods are able to make sound recommendations on appropriate learning paths with significantly improved learning results in terms of accuracy and efficiency.

Semantic web image retrieval is useful to end-users for semantic image searches over the Internet. This paper aims to develop image retrieval techniques for large-scale web image databases. An advanced retrieval system, termed Multi-concept Retrieval using Bimodal Deep Learning (MRBDL), is proposed and implemented using Convolutional Neural Networks (CNNs) which can effectively capture semantic correlations between a visual image and its free contextual tags. Different from existing approaches using multiple and independent concepts in a query, MRBDL considers multiple concepts as a holistic scene for retrieval model learning. In particular, we first use a bimodal CNN to train a holistic scene classifier in two modalities, and then semantic correlations of the sub-concepts included in the images are leveraged to boost holistic scene recognition. The predicted semantic scores obtained from holistic scene classifier are combined with complementary information on web images to improve the retrieval performance. Experiments have been carried out over two publicly available web image databases. The results show that our proposed approach performs favorably compared with several other state-of-the-art methods .

Entity resolution refers to disambiguating and resolving entities in structured and unstructured data. Developments of effective resolution algorithms are significant for processing scientific documents, particularly for biomedical literature. Specifically, name ambiguity among biomedical entities is a primary task that needs to be solved in the knowledge extraction process. In this paper, we present a novel approach to disambiguating gene/protein names by using context graphs. A set of abstracts of documents is used to build the context graphs through disclosing the indirect co-occurrence relationships among words. Feature vectors of the graphs can be constructed according to information gain (IG) on the word set. To evaluate the IG values, we propose a new metrics that integrates the word frequency (WF), dispersion degree (DD) and concentration degree (CD). Finally, entity resolution is performed by applying a support vector machine (SVM). Compared to existing approaches, the proposed method is capable of discovering latent information from the context of entity names, rather than using some statistical information such as the number of occurrences of words. Based on the results from comprehensive experiments over two benchmark datasets, we conclude that our proposed method, compared to several existing solutions, for resolving ambiguity entities is promising.