Loading...

Table of Content

    10 November 2022, Volume 42 Issue 11
    CCF Bigdata 2021
    Survey on imbalanced multi‑class classification algorithms
    Mengmeng LI, Yi LIU, Gengsong LI, Qibin ZHENG, Wei QIN, Xiaoguang REN
    2022, 42(11):  3307-3321.  DOI: 10.11772/j.issn.1001-9081.2021122060
    Asbtract ( )   HTML ( )   PDF (1861KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Imbalanced data classification is an important research content in machine learning, but most of the existing imbalanced data classification algorithms foucus on binary classification, and there are relatively few studies on imbalanced multi?class classification. However, datasets in practical applications usually have multiple classes and imbalanced data distribution, and the diversity of classes further increases the difficulty of imbalanced data classification, so the multi?class classification problem has become a research topic to be solved urgently. The imbalanced multi?class classification algorithms proposed in recent years were reviewed. According to whether the decomposition strategy was adopted, imbalanced multi?class classification algorithms were divided into decomposition methods and ad?hoc methods. Furthermore, according to the different adopted decomposition strategies, the decomposition methods were divided into two frameworks: One Vs. One (OVO) and One Vs. All (OVA). And according to different used technologies, the ad?hoc methods were divided into data?level methods, algorithm?level methods, cost?sensitive methods, ensemble methods and deep network?based methods. The advantages and disadvantages of these methods and their representative algorithms were systematically described, the evaluation indicators of imbalanced multi?class classification methods were summarized, the performance of the representative methods were deeply analyzed through experiments, and the future development directions of imbalanced multi?class classification were discussed.

    K‑nearest neighbor imputation subspace clustering algorithm for high‑dimensional data with feature missing
    Yongjian QIAO, Xiaolin LIU, Liang BAI
    2022, 42(11):  3322-3329.  DOI: 10.11772/j.issn.1001-9081.2021111964
    Asbtract ( )   HTML ( )   PDF (1207KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    During the clustering process of high?dimensional data with feature missing, there are problems of the curse of dimensionality caused by data high dimension and the invalidity of effective distance calculation between samples caused by data feature missing. To resolve above issues, a K?Nearest Neighbor (KNN) imputation subspace clustering algorithm for high?dimensional data with feature missing was proposed, namely KISC. Firstly, the nearest neighbor relationship in the subspace of the high?dimensional data with feature missing was used to perform KNN imputation on the feature missing data in the original space. Then, multiple iterations of matrix decomposition and KNN imputation were used to obtain the final reliable subspace structure of the data, and the clustering analysis was performed in that obtained subspace structure. The clustering results in the original space of six image datasets show that the KISC algorithm has better performance than the comparison algorithm which clusters directly after interpolation, indicating that the subspace structure can identify the potential clustering structure of the data more easily and effectively; the clustering results in the subspace of six high?dimensional datasets shows that the KISC algorithm outperforms the comparison algorithm in all datasets, and has the optimal clustering Accuracy and Normalized Mutual Information (NMI) on most of the datasets. The KISC algorithm can deal with high?dimensional data with feature missing more effectively and improve the clustering performance of these data.

    Neural tangent kernel K‑Means clustering
    Mei WANG, Xiaohui SONG, Yong LIU, Chuanhai XU
    2022, 42(11):  3330-3336.  DOI: 10.11772/j.issn.1001-9081.2021111961
    Asbtract ( )   HTML ( )   PDF (2237KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Aiming at the problem that the clustering results of K-Means clustering algorithm are affected by the sample distribution because of using the mean to update the cluster centers, a Neural Tangent Kernel K-Means (NTKKM) clustering algorithm was proposed. Firstly, the data of the input space were mapped to the high-dimensional feature space through the Neural Tangent Kernel (NTK), then the K-Means clustering was performed in the high-dimensional feature space, and the cluster centers were updated by taking into account the distance between clusters and within clusters at the same time. Finally, the clustering results were obtained. On the car and breast-tissue datasets, three evaluation indexes including accuracy, Adjusted Rand Index (ARI) and FM index of NTKKM clustering algorithm and comparison algorithms were counted. Experimental results show that the effect of clustering and the stability of NTKKM clustering algorithm are better than those of K-Means clustering algorithm and Gaussian kernel K?Means clustering algorithm. Compared with the traditional K?Means clustering algorithm, NTKKM clustering algorithm has the accuracy increased by 14.9% and 9.4% respectively, the ARI increased by 9.7% and 18.0% respectively, and the FM index increased by 12.0% and 12.0% respectively, indicating the excellent clustering performance of NTKKM clustering algorithm.

    Efficient failure recovery method for stream data processing system
    Yang LIU, Yangyang ZHANG, Haoyi ZHOU
    2022, 42(11):  3337-3345.  DOI: 10.11772/j.issn.1001-9081.2021122108
    Asbtract ( )   HTML ( )   PDF (2031KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Focusing on the issue that the single point of failure cannot be efficiently handled by streaming data processing system Flink, a new fault?tolerant system based on incremental state and backup, Flink+, was proposed. Firstly, backup operators and data paths were established in advance. Secondly, the output data in the data flow diagram was cached, and disks were used if necessary. Thirdly, task state synchronization was performed during system snapshots. Finally, backup tasks and cached data were used to recover calculation in case of system failure. In the system experiment and test, Flink+ dose not significantly increase the additional fault tolerance overhead during fault?free operation; when dealing with the single point of failure in both single?machine and distributed environments, compared with Flink system, the proposed system has the failure recovery time reduced by 96.98% in single?machine 8?task parallelism and by 88.75% in distributed 16?task parallelism. Experimental results show that using incremental state and backup method together can effectively reduce the recovery time of the single point of failure of the stream system and enhance the robustness of the system.

    Multi‑agent reinforcement learning based on attentional message sharing
    Rong ZANG, Li WANG, Tengfei SHI
    2022, 42(11):  3346-3353.  DOI: 10.11772/j.issn.1001-9081.2021122169
    Asbtract ( )   HTML ( )   PDF (1668KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Communication is an important way to achieve effective cooperation among multiple agents in a non? omniscient environment. When there are a large number of agents, redundant messages may be generated in the communication process. To handle the communication messages effectively, a multi?agent reinforcement learning algorithm based on attentional message sharing was proposed, called AMSAC (Attentional Message Sharing multi?agent Actor?Critic). Firstly, a message sharing network was built for effective communication among agents, and information sharing was achieved through message reading and writing by the agents, thus solving the problem of lack of communication among agents in non?omniscient environment with complex tasks. Then, in the message sharing network, the communication messages were processed adaptively by the attentional message sharing mechanism, and the messages from different agents were processed with importance order to solve the problem that large?scale multi?agent system cannot effectively identify and utilize the messages during the communication process. Moreover, in the centralized Critic network, the Native Critic was used to update the Actor network parameters according to Temporal Difference (TD) advantage policy gradient, so that the action values of agents were evaluated effectively. Finally, during the execution period, the decision was made by the agent distributed Actor network based on its own observations and messages from message sharing network. Experimental results in the StarCraft Multi?Agent Challenge (SMAC) environment show that compared with Native Actor?Critic (Native AC), Game Abstraction Communication (GA?Comm) and other multi?agent reinforcement learning methods, AMSAC has an average win rate improvement of 4 - 32 percentage points in four different scenarios. AMSAC’s attentional message sharing mechanism provides a reasonable solution for processing communication messages among agents in a multi?agent system, and has broad application prospects in both transportation hub control and unmanned aerial vehicle collaboration.

    Graph convolutional network method based on hybrid feature modeling
    Zhuoran LI, Zhonglin YE, Haixing ZHAO, Jingjing LIN
    2022, 42(11):  3354-3363.  DOI: 10.11772/j.issn.1001-9081.2021111981
    Asbtract ( )   HTML ( )   PDF (3410KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    For the complex information contained in the network, more ways are needed to extract useful information from it, but the relevant characteristics in the network cannot be completely described by the existing single?feature Graph Neural Network (GNN). To resolve the above problems, a Hybrid feature?based Dual Graph Convolutional Network (HDGCN) was proposed. Firstly, the structure feature vectors and semantic feature vectors of nodes were obtained by Graph Convolutional Network (GCN). Secondly, the features of nodes were aggregated selectively so that the feature expression ability of nodes was enhanced by the aggregation function based on attention mechanism or gating mechanism. Finally, the hybrid feature vectors of nodes were gained by the fusion mechanism based on a feasible dual?channel GCN, and the structure features and semantic features of nodes were modeled jointly to make the features be supplement for each other and promote the method's performance on subsequent machine learning tasks. Verification was performed on the datasets CiteSeer, DBLP (DataBase systems and Logic Programming) and SDBLP (Simplified DataBase systems and Logic Programming). Experimental results show that compared with the graph convolutional network model based on structure feature training, the dual channel graph convolutional network model based on hybrid feature training has the average value of Micro?F1 increased by 2.43, 2.14, 1.86 and 2.13 percentage points respectively, and the average value of Macro?F1 increased by 1.38, 0.33, 1.06 and 0.86 percentage points respectively when the training set proportion is 20%, 40%, 60% and 80%. The difference in accuracy is no more than 0.5 percentage points when using concat or mean as the fusion strategy, which shows that both concat and mean can be used as the fusion strategy. HDGCN has higher accuracy on node classification and clustering tasks than models trained by structure or semantic network alone, and has the best results when the output dimension is 64, the learning rate is 0.001, the graph convolutional layer number is 2 and the attention vector dimension is 128.

    Popularity prediction method of Twitter topics based on evolution patterns
    Weifan XIE, Yan GUO, Guangsheng KUANG, Zhihua YU, Yuanhai XUE, Huawei SHEN
    2022, 42(11):  3364-3370.  DOI: 10.11772/j.issn.1001-9081.2022010045
    Asbtract ( )   HTML ( )   PDF (934KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    A popularity prediction method of Twitter topics based on evolution patterns was proposed to address the problem that the differences between evolution patterns and the time?effectiveness of prediction were not taken into account in previous popularity prediction methods. Firstly, the K?SC (K?Spectral Centroid) algorithm was used to cluster the popularity sequences of a large number of historical topics, and 6 evolution patterns were obtained. Then, a Fully Connected Network (FCN) was trained as the prediction model by using historical topic data of each evolution pattern. Finally, in order to select the prediction model for the topic to be predicted, Amplitude?Alignment Dynamic Time Warping (AADTW) algorithm was proposed to calculate the similarity between the known popularity sequence of the topic to be predicted and each evolution pattern, and the prediction model of the evolution pattern with the highest similarity was selected to predict the popularity. In the task of predicting the popularity of the next 5 hours based on the known popularity of the first 20 hours, the Mean Absolute Percentage Error (MAPE) of the prediction results of the proposed method was reduced by 58.2% and 31.0% respectively, compared with those of the Auto?Regressive Integrated Moving Average (ARIMA) method and method using a single fully connected network. Experimental results show that the model group based on the evolution patterns can predict the popularity of Twitter topic more accurately than single model.

    Process tracking multi‑task rumor verification model combined with stance
    Bin ZHANG, Li WANG, Yanjie YANG
    2022, 42(11):  3371-3378.  DOI: 10.11772/j.issn.1001-9081.2021122148
    Asbtract ( )   HTML ( )   PDF (1420KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    At present, social media platforms have become the main ways for people to publish and obtain information, but the convenience of information publish may lead to the rapid spread of rumors, so verifying whether information is a rumor and stoping the spread of rumors has become an urgent problem to be solved. Previous studies have shown that people's stance on information can help determining whether the information is a rumor or not. Aiming at the problem of rumor spread, a Joint Stance Process Multi?Task Rumor Verification Model (JSP?MRVM) was proposed on the basis of the above result. Firstly, three propagation processes of information were represented by using topology map, feature map and common Graph Convolutional Network (GCN) respectively. Then, the attention mechanism was used to obtain the stance features of the information and fuse the stance features with the tweet features. Finally, a multi?task objective function was designed to make the stance classification task better assist in verifying rumors. Experimental results prove that the accuracy and Macro?F1 of the proposed model on RumorEval dataset are improved by 10.7 percentage points and 11.2 percentage points respectively compared to those of the baseline model RV?ML (Rumor Verification scheme based on Multitask Learning model), verifying that the proposed model is effective and can reduce the spread of rumors.

    Detection of unsupervised offensive speech based on multilingual BERT
    Xiayang SHI, Fengyuan ZHANG, Jiaqi YUAN, Min HUANG
    2022, 42(11):  3379-3385.  DOI: 10.11772/j.issn.1001-9081.2021112005
    Asbtract ( )   HTML ( )   PDF (1536KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Offensive speech has a serious negative impact on social stability. Currently, automatic detection of offensive speech focuses on a few high?resource languages, and the lack of sufficient offensive speech tagged corpus for low?resource languages makes it difficult to detect offensive speech in low?resource languages. In order to solve the above problem, a cross?language unsupervised offensiveness transfer detection method was proposed. Firstly, an original model was obtained by using the multilingual BERT (multilingual Bidirectional Encoder Representation from Transformers, mBERT) model to learn the offensive features on the high?resource English dataset. Then, by analyzing the language similarity between English and Danish, Arabic, Turkish, Greek, the obtained original model was transferred to the above four low?resource languages to achieve automatic detection of offensive speech on low?resource languages. Experimental results show that compared with the four methods of BERT, Linear Regression (LR), Support Vector Machine (SVM) and Multi?Layer Perceptron (MLP), the proposed method increases both the accuracy and F1 score of detecting offensive speech of languages such as Danish, Arabic, Turkish, and Greek by nearly 2 percentage points, which are close to those of the current supervised detection, showing that the combination of cross?language model transfer learning and transfer detection can achieve unsupervised offensiveness detection of low?resource languages.

    Neural machine translation method based on source language syntax enhanced decoding
    Longchao GONG, Junjun GUO, Zhengtao YU
    2022, 42(11):  3386-3394.  DOI: 10.11772/j.issn.1001-9081.2021111963
    Asbtract ( )   HTML ( )   PDF (1267KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Transformer, one of the best existing machine translation models, is based on the standard end?to?end structure and only relies on pairs of parallel sentences, which is believed to be able to learn knowledge in the corpus automatically. However, this modeling method lacks explicit guidance and cannot effectively mine deep language knowledge, especially in the low?resource environment with limited corpus size and quality, where the sentence encoding has no prior knowledge constraints, leading to the decline of translation quality. In order to alleviate the issues above, a neural machine translation model based on source language syntax enhanced decoding was proposed to explicitly use the source language syntax to guide the encoding, namely SSED (Source language Syntax Enhanced Decoding). A syntax?aware mask mechanism based on the syntactic information of the source sentence was constructed at first, and an additional syntax?dependent representation was generated by guiding the encoding self?attention. Then the syntax?dependent representation was used as a supplement to the representation of the original sentence and the decoding process was integrated by attention mechanism, which jointly guided the generation of the target language, realizing the enhancement of the prior syntax. Experimental results on several standard IWSLT (International Conference on Spoken Language Translation) and WMT (Conference on Machine Translation) machine translation evaluation task test sets show that compared with the baseline model Transformer, the proposed method obtains a BLEU score improvement of 0.84 to 3.41 respectively, achieving the state?of?the?art results of the syntactic related research. The fusion of syntactic information and self?attention mechanism is effective, the use of source language syntax can guide the decoding process of the neural machine translation system and significantly improve the quality of translation.

    User incentive based bike‑sharing dispatching strategy
    Bing SHI, Xizi HUANG, Zhaoxiang SONG, Jianqiao XU
    2022, 42(11):  3395-3403.  DOI: 10.11772/j.issn.1001-9081.2021122109
    Asbtract ( )   HTML ( )   PDF (2192KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    To address the dispatching problem of bike?sharing, considering the budget constraints, user maximum walking distance restrictions, user temporal and spatial demands and dynamic changes in the distribution of shared bicycles, a bike?sharing dispatching strategy with user incentives was proposed to improve the long?term user service rate of the bike?sharing platform. The dispatching strategy consists of a task generation algorithm, a budget allocation algorithm and a task allocation algorithm. In the task generation algorithm, the Long Short?Term Memory (LSTM) network was used to predict the future bike demand of users; in the budget allocation algorithm, the Deep Deterministic Policy Gradient (DDPG) algorithm was used to design a budget allocation strategy; after the budget was allocated to the tasks, the tasks needed to be allocated to the user for execution, so a greedy matching strategy was used for task allocation. Experiments were carried out on the Mobike dataset to compare the proposed strategy with the dispatching strategy with unlimited budget (that is, the platform is not limited by budget and can use any money to encourage users to ride to the target area), the greedy dispatching strategy, the dispatching strategy with truck hauling, and the situation without dispatching. Experimental results show that the proposed dispatching strategy with user incentive can effectively improve the service rate in the bike?sharing system compared to the greedy dispatching strategy and dispatching strategy with truck hauling.

    Deep fusion model for predicting differential gene expression by histone modification data
    Xin LI, Tao JIA
    2022, 42(11):  3404-3412.  DOI: 10.11772/j.issn.1001-9081.2021111956
    Asbtract ( )   HTML ( )   PDF (1734KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Concering the problem that the Cell type?Specificity (CS) and similarity and difference information between different cell types are not properly used when predicting Differential Gene Expression (DGE) with large?scale Histone Modification (HM) data, as well as large volume of input and high computational cost, a deep learning?based method named dcsDiff was proposed. Firstly, multiple AutoEncoders (AEs) and Bi?directional Long Short?Term Memory (Bi?LSTM) networks were introduced to reduce the dimensionality of HM signals and model them to obtain the embedded representation. Then, multiple Convolutional Neural Networks (CNNs) were used to mine the HM combined effects in each single cell type, and the similarity and difference information of each HM and joint effects of all HMs between two cell types. Finally, the two kinds of information were fused to predict DGE between two cell types. In the comparison experiments with DeepDiff on 10 pairs of cell types in the REMC (Roadmap Epigenomics Mapping Consortium) database, the Pearson Correlation Coefficient (PCC) of dcsDiff in DGE prediction was increased by 7.2% at the highest and 3.9% on average, the number of differentially expressed genes accurately detected by dcsDiff was increased by 36 at most and 17.6 on average, and the running time of dcsDiff was saved by 78.7%. The validity of reasonable integration of the above two kinds of information was proved in the component analysis experiment. The parameters of dcsDiff were also determined by experiments. Experimental results show that the proposed dcsDiff can effectively improve the efficiency of DGE prediction.

    2021 CCF China Blockchain Conference (CCF CBCC 2021)
    Research progress of blockchain‑based federated learning
    Rui SUN, Chao LI, Wei WANG, Endong TONG, Jian WANG, Jiqiang LIU
    2022, 42(11):  3413-3420.  DOI: 10.11772/j.issn.1001-9081.2021111934
    Asbtract ( )   HTML ( )   PDF (1086KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Federated Learning (FL) is a novel privacy?preserving learning paradigm that can keep users' data locally. With the progress of the research on FL, the shortcomings of FL, such as single point of failure and lack of credibility, are gradually gaining attention. In recent years, the blockchain technology originated from Bitcoin has achieved rapid development, which pioneers the construction of decentralized trust and provides a new possibility for the development of FL. The existing research works on blockchain?based FL were reviewed, the frameworks for blockchain?based FL were compared and analyzed. Then, key points of FL solved by the combination of blockchain and FL were discussed. Finally, the application prospects of blockchain?based FL were presented in various fields, such as Internet of Things (IoT), Industrial Internet of Things (IIoT), Internet of Vehicles (IoV) and medical services.

    Consensus transaction trajectory visualization tracking method for Fabric based on custom logs
    Shanshan LI, Yanze WANG, Yinglong ZOU, Huanlei CHEN, He ZHANG, Ou WU
    2022, 42(11):  3421-3428.  DOI: 10.11772/j.issn.1001-9081.2021111935
    Asbtract ( )   HTML ( )   PDF (3301KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Concerning that the federated chain lacks visualization methods to show the resource usage, health status, mutual relationship and consensus transaction process of each node, a Fabric consensus transaction Tracking method based on custom Log (FTL) was proposed. Firstly, Hyperledger Fabric, a typical federation framework, was used as the infrastructure to build the bottom layer of FTL. Then, the custom consensus transaction logs of the Fabric were collected and parsed by using the ELK (Elasticsearch, Logstash, Kibana) tool chain, and Spring Boot was used as the business logic processing framework. Finally, Graphin which focuses on graph analysis was utilized to realize the visualization of consensus trade trajectory. Experimental results show that compared with native Fabric applications, FTL Fabric?based application framework only experienced an 8.8% average performance decline after the implementation of visual tracking basis without significant latency, which can provide a more intelligent blockchain supervision solution for regulators.

    Blockchain construction and query method for spatio‑temporal data
    Yazhou HUA, Linlin DING, Ze CHEN, Junlu WANG, Zhu ZHU
    2022, 42(11):  3429-3437.  DOI: 10.11772/j.issn.1001-9081.2021111933
    Asbtract ( )   HTML ( )   PDF (2236KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    As a type of data with both temporal and spatial dimensions, spatio?temporal data is widely used in supply chain management, e?commerce and other fields, which integrity and security are of great importance in practical applications. Aiming at the problems of lack of transparency and easily being tampered of data in the current centralized storage of spatial?temporal datasets, a blockchain construction and query method for spatio?temporal data was proposed by combining the decentralized, tamper?proof and traceable characteristics of blockchain technology with spatio?temporal data management. Firstly, an improved Directed Asycline Graph Blockchain (Block?DAG) based blockchain architecture for spatio?temporal data, namely ST_Block?DAG (Spatio?Temporal Block?DAG), was proposed. Secondly, to improve the efficiency of spatio?temporal data storage and query, a storage structure based on quadtree and single linked list was adopted to store spatio?temporal data in the ST_Block?DAG blockchain. Finally, a variety of spatio?temporal data query algorithms were implemented on the basis of the storage structure of ST_Block?DAG, such as single?value query and range query. Experimental results show that compared with STBitcoin (Spatio?Temporal Bitcoin), Block?DAG and STEth (Spatio?Temporal Ethereum), ST_Block?DAG has the spatio?temporal data processing efficiency improved by more than 70% and the comprehensive query performance of spatio?temporal data improved by more than 60%. The proposed method can realize fast storage and query of spatio?temporal data, and can effectively support the management of spatio?temporal data.

    Cross-chain interaction safety model based on notary groups
    Chuyu JIANG, Lixi FANG, Ning ZHANG, Jianming ZHU
    2022, 42(11):  3438-3443.  DOI: 10.11772/j.issn.1001-9081.2021111915
    Asbtract ( )   HTML ( )   PDF (1065KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Concerning the problems of centralized functions of notary nodes and low cross?chain transaction efficiency in notary mechanism, a cross?chain interaction safety model based on notary groups was proposed. Firstly, notary nodes were divided into three kinds of roles, i.e. transaction verifiers, connectors and supervisors, and multiple transactions with consensus were packaged to a big deal by transaction verification group, and the threshold signature technique was used to sign it. Secondly, the confirmed transactions were placed in a cross?chain wait?to?be?transferred pool, some transactions were selected randomly by the connectors, and the technologies such as secure multiparty computation and fully homomorphic encryption were used to judge the authenticity of these transactions. Finally, if the hash values of all eligible transactions were true and reliable as well as verified by the transaction verification group, a batch task of multiple cross?chain transactions was able to be continued by the connector and be interacted with the blockchain in information. Security analysis shows that the proposed cross?chain mechanism is helpful to protect the confidentiality of information and the integrity of data, realizes the collaborative computing of data without leaving the database, and guarantees the stability of the cross?chain system of blockchain. Compared with the traditional cross?chain interaction security model, the complexity of the number of signatures and the number of notary groups that need to be assigned decreases from O(n) to O(1).

    Federated‑autonomy‑based cross‑chain scheme for blockchain
    Jianhui ZHENG, Feilong LIN, Zhongyu CHEN, Zhaolong HU, Changbing TANG
    2022, 42(11):  3444-3457.  DOI: 10.11772/j.issn.1001-9081.2021111922
    Asbtract ( )   HTML ( )   PDF (2880KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    To deal with the phenomenon of "information and value islands" caused by the lack of interoperation among the increasingly emerging blockchain systems, a federated?autonomy?based cross?chain scheme was proposed. The elemental idea of this scheme is to form a relay alliance chain maintained by participated blockchain systems using blockchain philosophy, which is supposed to solve the data sharing, value circulation and business collaboration problems among different blockchain systems. Firstly, a relay mode based cross?chain structure was proposed to provide interoperation services for heterogeneous blockchain systems. Secondly, the detailed design of the relay alliance chain was presented as well as the rules for the participated blockchain systems and their users. Then, the basic types of cross?chain interactions were summarized, and a process for implementing cross?chain interoperability based on smart contracts were designed. Finally, through multiple experiments, the feasibility of the cross?chain scheme was validated, the performance of the cross?chain system was evaluated, and the security of the whole cross?chain network was analyzed. Simulation results and security analysis prove that the proposed channel allocation strategy and block?out right allocation scheme of the proposed scheme are practically feasible, the throughput of the proposed shceme can reach up to 758 TPS (Transactions Per Second) when asset transactions are involved, and up to 960 TPS when asset transactions are not involved; the proposed scheme has high?level security and coarse? and fine?grained privacy protection mechanism. The proposed federated?autonomy?based cross?chain scheme for blockchain can provide secure and efficient cross?chain services, which is suitable for most of the current cross?chain scenarios.

    Fake news detection method based on blockchain technology
    Shengjia GONG, Linlin ZHANG, Kai ZHAO, Juntao LIU, Han YANG
    2022, 42(11):  3458-3464.  DOI: 10.11772/j.issn.1001-9081.2021111885
    Asbtract ( )   HTML ( )   PDF (1557KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Fake news not only leads to misconceptions and damages people's right to know the truth, but also reduces the credibility of news websites. In view of the occurrence of fake news in news websites, a fake news detection method based on blockchain technology was proposed. Firstly, the smart contract was invoked to randomly assign reviewers for the news for determining the authenticity of the news. Then, the credibility of the review results was improved by adjusting the number of reviewers and ensuring the number of effective reviewers. At the same time, the incentive mechanism was designed with rewards distributed according to the reviewers' behaviors, and the reviewers' behaviors and rewards were analyzed by game theory. In order to gain the maximum benefit, the reviewers' behaviors should be honest. An auditing mechanism was designed to detect malicious reviewers to improve system security. Finally, a simple blockchain fake news detection system was implemented by using Ethereum smart contract and simulated for fake news detection, and the results show that the accuracy of news authenticity detection of the proposed method reaches 95%, indicating that the proposed method can effectively prevent the release of fake news.

    Blockchain‑based electronic medical record secure sharing
    Chao LIN, Debiao HE, Xinyi HUANG
    2022, 42(11):  3465-3472.  DOI: 10.11772/j.issn.1001-9081.2021111895
    Asbtract ( )   HTML ( )   PDF (1451KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    To solve various issues faced by Electronic Medical Record (EMR) sharing, such as centralized data provider, passive patient data management, low interoperability efficiency and malicious dissemination, a blockchain-based EMR secure sharing method was proposed. Firstly, a more secure and efficient Universal Designated Verifier Signature Proof (UDVSP) scheme based on the commercial cryptography SM2 digital signature algorithm was proposed. Then, a smart contract with functionalities of uploading, verification, retrieval and revocation was designed, and a blockchain-based EMR secure sharing system was constructed. Finally, the feasibilities of UDVSP scheme and sharing system were demonstrated through security analysis and performance analysis. The security analysis shows that the proposed UDVSP is probably secure. The performance analysis shows that compared with existing UDVSP/UDVS schemes, the proposed UDVSP scheme saves the computation cost at least 87.42% and communication overhead at least 93.75%. The prototype of blockchain smart contract further demonstrates the security and efficiency of the sharing system.

    ChinaService 2021
    Data field classification algorithm for edge intelligent computing
    Zhiyu SUN, Qi WANG, Bin GAO, Zhongjun LIANG, Xiaobin XU, Shangguang WANG
    2022, 42(11):  3473-3478.  DOI: 10.11772/j.issn.1001-9081.2021091692
    Asbtract ( )   HTML ( )   PDF (2398KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    In view of the general problems of not fully utilizing historical information and slow parameter optimization process in the research of clustering algorithms, an adaptive classification algorithm based on data field was proposed in combination with edge intelligent computing, which can be deployed on Edge Computing (EC) nodes to provide local intelligent classification service. By introducing supervision information to modify the structure of the traditional data field clustering model, the proposed algorithm enabled the traditional data field to be applied to classification problems, extending the applicable fields of data field theory. Based on the idea of the data field, the proposed algorithm transformed the domain value space of the data into the data potential field space, and divided the data into several unlabeled cluster results according to the spatial potential value. After comparing the cluster results with the historical supervision information for cloud similarity, the cluster results were attributed to the most similar category. Besides, a parameter search strategy based on sliding step length was proposed to speeded up the parameter optimization of the proposed algorithm. Based on this algorithm, a distributed data processing scheme was proposed. Through the cooperation of cloud center and edge devices, classification tasks were cut and distributed to different levels of nodes to achieve modularity and low coupling. Simulation results show that the precision and recall of the proposed algorithm maintained above 96%, and the Hamming loss was less than 0.022. Experimental results show that the proposed algorithm can accurately classify and accelerate the speed of parameter optimization, and outperforms than Logistic Regression (LR) algorithm and Random Forest (RF) algorithm in overall performance.

    Cache cooperation strategy for maximizing revenue in mobile edge computing
    Yali WANG, Jiachao CHEN, Junna ZHANG
    2022, 42(11):  3479-3485.  DOI: 10.11772/j.issn.1001-9081.2022020194
    Asbtract ( )   HTML ( )   PDF (1553KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Mobile Edge Computing (MEC) can reduce the energy consumption of mobile devices and the delay of users’ acquisition to services by deploying resources in users’ neighborhood; however, most relevant caching studies ignore the regional differences of the services requested by users. A cache cooperation strategy for maximizing revenue was proposed by considering the features of requested content in different regions and the dynamic characteristic of content. Firstly, considering the regional features of user preferences, the base stations were partitioned into several collaborative domains, and the base stations in each collaboration domain was able to serve users with the same preferences. Then, the content popularity in each region was predicted by the Auto?Regressive Integrated Moving Average (ARIMA) model and the similarity of the content. Finally, the cache cooperation problem was transformed into a revenue maximization problem, and the greedy algorithm was used to solve the content placement and replacement problems according to the revenue obtained by content storage. Simulation results showed that compared with the Grouping?based and Hierarchical Collaborative Caching (GHCC) algorithm based on MEC, the proposed algorithm improved the cache hit rate by 28% with lower average transmission delay. It can be seen that the proposed algorithm can effectively improve the cache hit rate and reduce the average transmission delay at the same time.

    Requirement acquisition approach for intelligent computing services
    Ye WANG, Aohui ZHOU, Siyuan ZHOU, Bo JIANG, Junwu CHEN, Shizhe SONG
    2022, 42(11):  3486-3492.  DOI: 10.11772/j.issn.1001-9081.2022010059
    Asbtract ( )   HTML ( )   PDF (728KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    In intelligent computing services, data analysis and processing are provided for the service consumer by the service provider through Internet, and a learning model is established to complete intelligent computing function. Due to the lack of effective communication channels between service providers and service consumers, as well as the fuzzy and messy requirement descriptions of the service consumer feedback, there is a lack of a unified service requirement acquisition method to effectively analyze, organize and regulate the continuously changing requirement of users, which leads to the failure of intelligent computing services to make a rapid improvement according to the user’s requirements. Aiming at the problems of continuity and uncertainty of requirement changes in service development, a requirement acquisition method for intelligent computing services was proposed. The application feedback and questions of intelligent computing services were firstly obtained from Stack Overflow question and answer forum. Then, the knowledge classification and prioritization were performed on them by using different learning models (including Support Vector Machine (SVM), naive Bayes and TextCNN) according to the types of requirements concerned by the service consumer. Finally, a customized service requirement template was used to describe the requirements of intelligent computing services.

    Event‑driven dynamic collection method for microservice invocation link data
    Peng LI, Zhuofeng ZHAO, Han LI
    2022, 42(11):  3493-3499.  DOI: 10.11772/j.issn.1001-9081.2021101735
    Asbtract ( )   HTML ( )   PDF (2643KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Microservice invocation link data is a type of important data generated in the daily operation of the microservice application system, which records a series of service invocation information corresponding to a user request in the microservice application in the form of link. Microservice invocation link data are generated at different microservice deployment nodes due to the distribution characteristic of the system, and the current collection methods for these distributed data include full collection and sampling collection. Full collection may bring large data transmission and data storage costs, while sampling collection may miss critical invocation data. Therefore, an event?driven and pipeline sampling based dynamic collection method for microservice invocation link data was proposed, and a microservice invocation link system that supports dynamic collection of invocation link data was designed and implemented based on the open?source software Zipkin. Firstly, the pipeline sampling was performed on the link data of different nodes that met the predefined event features, that is the same link data of all nodes were collected by the data collection server only when the event defined data was generated by a node; meanwhile, to address the problem of inconsistent data generation rates of different nodes, multi?threaded streaming data processing technology based on time window and data synchronization technology were used to realize the data collection and transmission of different nodes. Finally, considering the problem that the link data of each node arrives at the server in different sequential order, the synchronization and summary of the full link data were realized through the timing alignment method. Experimental results on the public microservice lrevocation ink dataset prove that compared to the full collection and sampling collection methods, the proposed method has higher accuracy and more efficient collection on link data containing specific events such as anomalies and slow responces.

    Service integration method based on adaptive multi‑objective reinforcement learning
    Xiao GUO, Chunshan LI, Yuyue ZHANG, Dianhui CHU
    2022, 42(11):  3500-3505.  DOI: 10.11772/j.issn.1001-9081.2021122041
    Asbtract ( )   HTML ( )   PDF (1403KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    The current service resources in Internet of Services (IoS) show a trend of refinement and specialization. Services with single function cannot meet the complex and changeable requirements of users. Service integrating and scheduling methods have become hot spots in the field of service computing. However, most existing service integrating and scheduling methods only consider the satisfaction of user requirements and do not consider the sustainability of the IoS ecosystem. In response to the above problems, a service integration method based on adaptive multi?objective reinforcement learning was proposed. In this method, a multi?objective optimization strategy was introduced into the framework of Asynchronous Advantage Actor?Critic (A3C) algorithm, so as to ensure the healthy development of the IoS ecosystem while satisfying user needs. The integrated weight of the multi?objective value was able to adjusted dynamically according to the regret value, which improved the imbalance of sub?objective values in multi?objective reinforcement learning. The service integration verification was carried out in a real large?scale service environment. Experimental results show that the proposed method is faster than traditional machine learning methods in large?scale service environment, and has a more balanced solution quality of each objective compared with Reinforcement Learning (RL) with fixed weights.

    Personalized recommendation service system based on cloud-client-convergence
    Jialiang HAN, Yudong HAN, Xuanzhe LIU, Yaoshuai ZHAO, Di FENG
    2022, 42(11):  3506-3512.  DOI: 10.11772/j.issn.1001-9081.2021111992
    Asbtract ( )   HTML ( )   PDF (1160KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Mainstream personalized recommendation systems usually use models deployed in the cloud to perform recommendation, so the private data such as user interaction behaviors need to be uploaded to the cloud, which may cause potential risks of user privacy leakage. In order to protect user privacy, user-sensitive data can be processed on the client, however, there are communication bottleneck and computation resource bottleneck in clients. Aiming at the above challenges, a personalized recommendation service system based on cloud-client-convergence was proposed. In this system, the cloud-based recommendation model was divided into a user representation model and a sorting model. After being pre-trained on the cloud, the user representation model was deployed to the client, while the sorting model was deployed to the cloud. A small-scale Recurrent Neural Network (RNN) was used to model the user behavior characteristics by extracting temporal information from user interaction logs, and the Lasso (Least absolute shrinkage and selection operator) algorithm was used to compress user representations, thereby preventing a drop in recommendation accuracy while reducing the communication overhead between the cloud and the client as well as the computation overhead of the client. Experiments were conducted on RecSys Challenge 2015 dataset, and the results show that the recommendation accuracy of the proposed system is comparable to that of the GRU4REC model, while the volume of the compressed user representations is only 34.8% of that before compression, with a lower computational overhead.

    Personal event detection method based on text mining in social media
    Rui XIAO, Mingyi LIU, Zhiying TU, Zhongjie WANG
    2022, 42(11):  3513-3519.  DOI: 10.11772/j.issn.1001-9081.2022010106
    Asbtract ( )   HTML ( )   PDF (2013KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Users’ social media contains their past personal experiences and potential life patterns, and the study of their patterns is of great value for predicting users’ future behaviors and performing personalized recommendations for users. By collecting Weibo data, 11 types of events were defined, and a three?stage Pipeline system was proposed to detect personal events by using BERT (Bidirectional Encoder Representations from Transformers) pre?trained models in three stages respectively, including BERT+BiLSTM+Attention, BERT+FullConnect and BERT+BiLSTM+CRF. The information of whether the text contained defined events, the event types of events contained, and the elements contained in each event were extracted from the Weibo, and the specific elements are Subject (subject of the event), Object (event element), Time (event occurrence time), Place (place where the event occurred) and Tense (tense of the event), thereby exploring the change law of user’s personal event timeline to predict personal events. Comparative experiments and analysis were conducted with classification algorithms such as logistic regression, naive Bayes, random forest and decision tree on a collected real user Weibo dataset. Experimental results show that the BERT+BiLSTM+Attention, BERT+FullConnect, BERT+BiLSTM+CRF methods used in three stages achieve the highest F1?score, verifying the effectiveness of the proposed methods. Finally, the personal event timeline was visually built according to the extracted events with time information.

    Recommendation service for API use cases based on open source community analysis
    Jiaqi ZHANG, Yanchun SUN, Gang HUANG
    2022, 42(11):  3520-3526.  DOI: 10.11772/j.issn.1001-9081.2021122070
    Asbtract ( )   HTML ( )   PDF (1339KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Current research on Application Program Interface (API) learning and code reuse focuses on mining frequent API usage patterns, extracting component information, and recommending personalized API services based on user requirements and target functions. However, as beginners in software development who lack professional knowledge, experience and skills to implement specific use cases, they often need real code use cases as a reference except reading official documents. Most of the existing research about code recommendation is in single fragment mode. The lack of cross function case in case selection is not conducive for beginners to learn to build a complete use scenario or a functional module. At the same time, the semantic description extracted from a single function annotation is not enough for learners to understand the complete function implementation method of the project. To solve the above problems, an API use case recommendation service based on open source community analysis was proposed. Taking the software development back?end framework Spring Boot as an example, a cross function case recommendation assistant learning service was constructed. Then, the feasibility and effectiveness of the proposed API use case recommendation service was verified through questionnaires and expert verification.

    Product and service quality analysis based on customer service dialogues
    Jiaju ZHANG, Huiping LIN
    2022, 42(11):  3527-3533.  DOI: 10.11772/j.issn.1001-9081.2022010073
    Asbtract ( )   HTML ( )   PDF (2371KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Existing product and service quality analysis is often based on questionnaire survey or product reviews, but there are problems such as difficulty in questionnaire collection and invalid data in product reviews. As a bridge between customers and businesses, the customer service dialogue contains rich customer opinions from product to service perspective, however, there are still few studies using customer service dialogues to analyze product and service quality. A product and service quality analysis method based on customer service dialogues was proposed, which firstly combined the product features and service blueprint to determine product and service quality evaluation factors, and used the Important?Performance Analysis (IPA) method to define the importance and performance index of evaluation factors. Then, quantitative analysis of the importance and satisfaction of products and services was performed by using the dialogue topic extraction and sentiment analysis. The method was applied on the real customer service dialogues of a Taobao flagship store which sells disinfection and sterilization products, and 18 evaluation factors were established, whose importance and performance were quantified based on more than 900 thousand real historical customer service dialogues, thereby analyzing the quality of products and services of the flagship store. Finally, a questionnaire on the professional customer service employees was carried out to verify the effectiveness of the proposed method.

    ChinaVR 2021
    Review of eye movement‑based interaction techniques for virtual reality systems
    Shouming HOU, Chaolan JIA, Mingmin ZHANG
    2022, 42(11):  3534-3543.  DOI: 10.11772/j.issn.1001-9081.2021122134
    Asbtract ( )   HTML ( )   PDF (1617KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Eye movement?based human?computer interaction can enhance the immersion and improve comfort of users by using eye?movement characteristics, and the incorporating of eye movement?based interaction techniques in Virtual Reality (VR) system plays a vital role in the popularity of VR systems, which has become a research hotspot in recent years. Firstly, the principles and categories of VR eye movement?based interaction techniques were described, the advantages of combining VR systems with eye movement?based interaction techniques were analyzed, and the current mainstream VR head?mounted display devices and typical application scenarios were summarized. Then, based on the analysis of experiments related to VR eye tracking, the research hotspots of VR eye movement were summarized, including miniaturized equipment, diopter correction, lack of high?quality content, blurring and distortion of eyeball images, positioning accuracy and near?eye display system, and the corresponding solutions were prospected for those related hot issues.

    Passive haptic interaction method for multiple virtual targets in vast virtual reality space
    Jieke WANG, Lin LI, Hailong ZHANG, Liping ZHENG
    2022, 42(11):  3544-3550.  DOI: 10.11772/j.issn.1001-9081.2021122123
    Asbtract ( )   HTML ( )   PDF (2818KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Focused on the issue that real interaction targets cannot be matched with the virtual interaction targets one by one when providing passive haptics for redirected walking users in a vast Virtual Reality (VR) space, a method with two physical proxies acting as haptic proxies to provide haptic feedback for multiple virtual targets was proposed, in order to meet the user’s passive haptic needs alternately during the redirected walking process based on Artificial Potential Field (APF). Aiming at the misalignment of virtual and real targets caused by the redirected walking algorithm itself and inaccurate calibration, the position and orientation of the virtual target were designed and haptic retargeting was introduced in the interaction stage. Simulation experimental results show that the design of the virtual target position and orientation can reduce the alignment error greatly. User experiments prove that haptic retargeting further improves the interaction accuracy and can bring users a richer and more immersive experience.

    Visual‑saliency‑driven reuse algorithm of indirect lighting in 3D scene rendering
    Shujie QI, Chunyi CHEN, Xiaojuan HU, Haiyang YU
    2022, 42(11):  3551-3557.  DOI: 10.11772/j.issn.1001-9081.2021122181
    Asbtract ( )   HTML ( )   PDF (2946KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    In order to accelerate rendering of 3D scenes by path tracing, a visual?saliency?driven reuse algorithm of indirect lighting in 3D scene rendering was proposed. Firstly, according to the characteristics of visual perception that the regions of interest have high saliency, while other regions have low saliency, a 2D saliency map of the scene image was obtained, which consists of color information, edge information, depth information and motion information of the image. Then, the indirect lighting in the high?saliency area was re?rendered, while the indirect lighting of the previous frame was reused in the low?saliency area under certain conditions, thereby accelerating the rendering. Experimental results show that the global lighting effect of the image generated by this method is real, and the rendering speed of the method is improved in several experimental scenes, and the speed can reach up to 5.89 times of that of the high?quality rendering.

    Object detection algorithm combined with optimized feature extraction structure
    Nan XIANG, Chuanzhong PAN, Gaoxiang YU
    2022, 42(11):  3558-3563.  DOI: 10.11772/j.issn.1001-9081.2021122122
    Asbtract ( )   HTML ( )   PDF (1607KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Concerning the problem of low object detection precision of DEtection TRansformer (DETR) for small targets, an object detection algorithm with optimized feature extraction structure, called CF?DETR (DETR combined CSP?Darknet53 and Feature pyramid network), was proposed on the basis of DETR. Firstly, CSP?Darknet53 combined with the optimized Cross Stage Partial (CSP) network was used to extract the features of the original image, and feature maps of 4 scales were output. Secondly, the Feature Pyramid Network (FPN) was used to splice and fuse the 4 scale feature maps after down?sampling and up?sampling, and output a 52×52 size feature map. Finally, the obtained feature map and the location coding information were combined and input into the Transformer to obtain the feature sequence. Through the Forward Feedback Networks (FFNs) as the prediction head, the category and location information of the prediction object was output. On COCO2017 dataset, compared with DETR, CF?DETR has the number of model hyperparameters reduced by 2×106, the average detection precision of small objects improved by 2.1 percentage points, and the average detection precision of medium? and large?sized objects improved by 2.3 percentage points. Experimental results show that the optimized feature extraction structure can effectively improve the DETR detection precision while reducing the number of model hyperparameters.

    Violence detection in video based on temporal attention mechanism and EfficientNet
    Xingquan CAI, Dingwei FENG, Tong WANG, Chen SUN, Haiyan SUN
    2022, 42(11):  3564-3572.  DOI: 10.11772/j.issn.1001-9081.2021122153
    Asbtract ( )   HTML ( )   PDF (2885KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Aiming at the problems of large model parameters, high computational complexity and low accuracy of traditional violence detection methods, a method of violence detection in video based on temporal attention mechanism and EfficientNet was proposed. Firstly, the foreground image obtained by preprocessing the dataset was input to the network model to extract the video features, meanwhile, the frame-level spatial features of violence were extracted by using the lightweight EfficientNet, and the global spatial-temporal features of the video sequence were further extracted by using the Convolutional Long Short-Term Memory (ConvLSTM) network. Then, combined with temporal attention mechanism, the video-level feature representations were obtained. Finally, the video-level feature representations were mapped to the classification space, and the Softmax classifier was used to classify the video violence and output the detection results, realizing the violence detection of video. Experimental results show that the proposed method can decrease the number of model parameters, reduce the computational complexity, increase the accuracy of violence detection and improve the comprehensive performance of the model with limited resources.

    Cross‑resolution person re‑identification by generative adversarial network based on multi‑granularity features
    Yanbing GENG, Yongjian LIAN
    2022, 42(11):  3573-3579.  DOI: 10.11772/j.issn.1001-9081.2021122124
    Asbtract ( )   HTML ( )   PDF (2598KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Existing Super Resolution (SR) reconstruction methods based on Generative Adversarial Network (GAN) for cross?resolution person Re?IDentification (ReID) suffer from deficiencies in both texture structure content recovery and feature consistency maintenance of the reconstructed images. To solve these problems, a cross?resolution pedestrian re?identification method based on multi?granularity information generation network was proposed. Firstly, a self?attention mechanism was introduced into multiple layers of generator to focus on multi?granularity stable regions with structural correlation, focusing on recovering the texture and structure information of the Low Resolution (LR) person image. At the same time, an identifier was added at the end of the generator to minimize the loss in different granularity features between the generated image and the real image during the training process, improving the feature consistency between the generated image and the real image in terms of features. Secondly, the self?attention generator and identifier were jointed, then they were optimized alternately with the discriminator to improve the generated image on content and features. Finally, the improved GAN and person re?identification network were combined to train the model parameters of the optimized network alternately until the model converged. Comparison Experimental results on several cross?resolution person re?identification datasets show that the proposed algorithm improves rank?1 accuracy on Cumulative Match Characteristic(CMC) by 10 percentage points on average, and has better performance in enhancing both content consistency and feature expression consistency of SR images.

    Forest pest detection method based on attention model and lightweight YOLOv4
    Haiyan SUN, Yunbo CHEN, Dingwei FENG, Tong WANG, Xingquan CAI
    2022, 42(11):  3580-3587.  DOI: 10.11772/j.issn.1001-9081.2021122164
    Asbtract ( )   HTML ( )   PDF (4972KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Aiming at the problems of slow detection speed, low precision, missed detection and false detection of current forest pest detection methods, a forest pest detection method based on attention model and lightweight YOLOv4 was proposed. Firstly, a dataset was constructed and preprocessed by using geometric transformation, random color dithering and mosaic data augmentation techniques. Secondly, the backbone network of YOLOv4 was replaced with a lightweight network MobileNetV3, and the Convolutional Block Attention Module (CBAM) was added to the improved Path Aggregation Network (PANet) to build the improved lightweight YOLOv4 network. Thirdly, Focal Loss was introduced to optimize the loss function of the YOLOv4 network model. Finally, the preprocessed dataset was input into the improved network model, and the detection results containing pest species and location information were output. Experimental results show that all the improvements of the network contribute to the performance improvement of the model; compared with the original YOLOv4 model, the proposed model has faster detection speed and higher detection mean Average Precision (mAP), and effectively solves the problem of missed detection and false detection. The proposed new model is superior to the existing mainstream network models and can meet the precision and speed requirements of real?time detection of forest pests.

    Artificial intelligence
    Review on interpretability of deep learning
    Xia LEI, Xionglin LUO
    2022, 42(11):  3588-3602.  DOI: 10.11772/j.issn.1001-9081.2021122118
    Asbtract ( )   HTML ( )   PDF (1703KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    With the widespread application of deep learning, human beings are increasingly relying on a large number of complex systems that adopt deep learning techniques. However, the black?box property of deep learning models offers challenges to the use of these models in mission?critical applications and raises ethical and legal concerns. Therefore, making deep learning models interpretable is the first problem to be solved to make them trustworthy. As a result, researches in the field of interpretable artificial intelligence have emerged. These researches mainly focus on explaining model decisions or behaviors explicitly to human observers. A review of interpretability for deep learning was performed to build a good foundation for further in?depth research and establishment of more efficient and interpretable deep learning models. Firstly, the interpretability of deep learning was outlined, the requirements and definitions of interpretability research were clarified. Then, several typical models and algorithms of interpretability research were introduced from the three aspects of explaining the logic rules, decision attribution and internal structure representation of deep learning models. In addition, three common methods for constructing intrinsically interpretable models were pointed out. Finally, the four evaluation indicators of fidelity, accuracy, robustness and comprehensibility were introduced briefly, and the possible future development directions of deep learning interpretability were discussed.

    Empathy prediction from texts based on transfer learning
    Chenguang LI, Bo ZHANG, Qian ZHAO, Xiaoping CHEN, Xingfu WANG
    2022, 42(11):  3603-3609.  DOI: 10.11772/j.issn.1001-9081.2021091632
    Asbtract ( )   HTML ( )   PDF (777KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Empathy prediction from texts achieves little progress due to the lack of sufficient labeled data, while the related task of text sentiment polarity classification has a large number of labeled samples. Since there is a strong correlation between empathy prediction and polarity classification, a transfer learning?based text empathy prediction method was proposed. Transferable public features were learned from the sentiment polarity classification task to assist text empathy prediction task. Firstly, a dynamic weighted fusion of public and private features between two tasks was performed through an attention mechanism. Secondly, in order to eliminate domain differences in datasets between two tasks, an adversarial learning strategy was used to distinguish the domain?unique features from the domain?public features between two tasks. Finally, a Hinge?loss constraint strategy was proposed to make common features be generic for different target labels and private features be unique to different target labels. Experimental results on two benchmark datasets show that compared to the comparison transfer learning methods, the proposed method has higher Pearson Correlation Coefficient (PCC) and coefficient of determination (R2), and has lower Mean?Square Error (MSE), which fully demonstrates the effectiveness of the proposed method.

    Session recommendation method based on graph model and attention model
    Weichao DANG, Zhiyu YAO, Shangwang BAI, Gaimei GAO, Chunxia LIU
    2022, 42(11):  3610-3616.  DOI: 10.11772/j.issn.1001-9081.2021091696
    Asbtract ( )   HTML ( )   PDF (1175KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    To solve the problem that representation of interest preferences based on the Recurrent Neural Network (RNN) is incomplete and inaccurate in session recommendation, a Session Recommendation method based on Graph Model and Attention Model (SR?GM?AM) was proposed. Firstly, the graph model used global graph and session graph to obtain neighborhood information and session information respectively, and used Graph Neural Network (GNN) to extract item graph features, which were passed through the global item representation layer and session item representation layer to obtain the global? level embedding and the session?level embedding, and the two levels of embedding were combined into graph embedding. Then, attention model used soft attention to fuse graph embedding and reverse position embedding, target attention activated the relevance of the target items, as well as attention model generated session embedding through linear transformation. Finally, SR?GM?AM outputted the recommended list of the N items for the next click through the prediction layer. Comparative experiments of SR?GM?AM and Lossless Edge?order preserving aggregation and Shortcut graph attention for Session?based Recommendation (LESSR) were conducted on two real public e?commerce datasets Yoochoose and Diginetica, and the results showed that SR?GM?AM had the highest P@20 of 72.41% and MRR@20 of 35.34%, verifying the effectiveness of it.

    Few‑shot target detection based on negative‑margin loss
    Yunyan DU, Hong LI, Jinhui YANG, Yu JIANG, Yao MAO
    2022, 42(11):  3617-3624.  DOI: 10.11772/j.issn.1001-9081.2021091683
    Asbtract ( )   HTML ( )   PDF (1509KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Most of the existing target detection algorithms rely on large?scale annotation datasets to ensure the accuracy of detection, however, it is difficult for some scenes to obtain a large number of annotation data and it consums a lot of human and material resources. In order to resolve this problem, a Few?Shot Target Detection method based on Negative Margin loss (NM?FSTD) was proposed. The negative margin loss method belonging to metric learning in Few?Shot Learning (FSL) was introduced into target detection, which could avoid mistakenly mapping the samples of the same novel classes to multiple peaks or clusters and helping to the classification of novel classes in few?shot target detection. Firstly, a large number of training samples and the target detection framework based on negative margin loss were used to train the model with good generalization performance. Then, the model was finetuned through a small number of labeled target category samples. Finally, the finetuned model was used to detect the new sample of target category. To verify the detection effect of NM?FSTD, MS COCO was used for training and evaluation. Experimental results show that the AP50 of NM?FSTD reaches 22.8%; compared with Meta R?CNN (Meta Regions with CNN features) and MPSR (Multi?Scale Positive Sample Refinement), the accuracies are improved by 3.7 and 4.9 percentage points, respectively. NM?FSTD can effectively improve the detection performance of target categories in the case of few?shot, and solve the problem of insufficient data in the field of target detection.

    Motor imagery electroencephalography classification based on data augmentation
    Yu PENG, Yaolian SONG, Jun YANG
    2022, 42(11):  3625-3632.  DOI: 10.11772/j.issn.1001-9081.2021091701
    Asbtract ( )   HTML ( )   PDF (3924KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Aiming at the multi?classification problem for Motor Imagery ElectroEncephaloGraphy (MI?EEG), Lightweight convolutional neural Network (L?Net) and Lightweight Hybrid Network (LH?Net) based on deep separable convolution were built on the basis of existing research. Experiments and analyses were carried out on the BCI competition IV-2a data set. It was shown that L?Net could fit the data faster than LH?Net, and the training time was shorter. However, LH?Net is more stable than L?Net and has better robustness in classification performance on the test set, the average accuracy and average Kappa coefficient of LH?Net were increased by 3.6% and 4.8%, respectively compared with L?Net. In order to further improve the classification performance of the model, a new method of adding Gaussian noise based on the time?frequency domain was adopted to apply Data Augmentation (DA) on the training samples, and simulation verification of the noise intensity was carried out, thus the optimal noise intensity ranges of the two models were inferred. With the DA method, the average accuracies of the two models were increased by at least 4% in the simulation results, the four classification effects were significantly improved.

    Cyber security
    High-capacity reversible data hiding in encrypted videos based on histogram shifting
    Pei CHEN, Shuaiwei ZHANG, Yangping LIN, Ke NIU, Xiaoyuan YANG
    2022, 42(11):  3633-3638.  DOI: 10.11772/j.issn.1001-9081.2021101722
    Asbtract ( )   HTML ( )   PDF (1692KB) ( )  
    Figures and Tables | References | Related Articles | Metrics

    Aiming at the low embedding capacity of Reversible Data Hiding (RDH) in encrypted videos, a high-capacity RDH scheme in encrypted videos based on histogram shifting was proposed. Firstly, 4×4 luminance intra-prediction mode and the sign bits of Motion Vector Difference (MVD) were encrypted by stream cipher, and then a two-dimensional histogram of MVD was constructed, and (0,0) symmetric histogram shifting algorithm was designed. Finally, (0,0) symmetric histogram shifting algorithm was carried out in the encrypted MVD domain to realize separable RDH in encrypted videos. Experimental results show that the embedding capacity of the proposed scheme is increased by 263.3% on average compared with the comparison schemes, the average Peak Signal-to-Noise Ratio (PSNR) of encrypted video is less than 15.956 dB, and the average PSNR of decrypted video with secret can reach more than 30 dB. The proposed scheme effectively improves the embedding capacity and is suitable for more types of video sequences.

2024 Vol.44 No.3

Current Issue
Archive
Honorary Editor-in-Chief: ZHANG Jingzhong
Editor-in-Chief: XU Zongben
Associate Editor: SHEN Hengtao XIA Zhaohui
Domestic Post Distribution Code: 62-110
Foreign Distribution Code: M4616
Address:
No. 9, 4th Section of South Renmin Road, Chengdu 610041, China
Tel: 028-85224283-803
  028-85222239-803
Website: www.joca.cn
E-mail: bjb@joca.cn
WeChat
Join CCF