最新刊期
卷 21 ， 期 5 ， 2022
- 摘要：Knowledge graph is a research hotspot in the intersection of big data and artificial intelligence， which can extract and express the potential knowledge in data and support complex reasoning and the construction of diverse intelligent applications. In recent years， there have been initial applications of knowledge graphs in the fields of Internet， finance and medical insurance， however， less attention has been paid to the integration of knowledge graphs with cultural domain. With the integration of culture and technology as the background， help with the construction of the smart museum， discusses the background and necessity of the knowledge graph construction for smart cultural museums， summarizes and analyzes the achievements of relevant scholars in recent years， and gives framework of cultural knowledge graph system. Namely， the related methods and technical schemes of knowledge extraction and relational reasoning are expounded. Then， it summarizes the possible applications and the problems that may be faced in the operation and management of the cultural knowledge graphs. Finally， the challenges and possible works in the future are analyzed and prospected.关键词：knowledge graph;cultural big data;smart cultural museums;knowledge extraction
- 301
- |
- 373
- |
- 0
发布时间：2022-05-24 - 摘要：With the continuous development of anti-detection technology， a large number of diversified malicious code variants have been produced，traditional detection technology has been unable to accurately detect this unknown malicious code. Because data visualization methods can express the core of malicious code in image features， Therefore， the visualized malicious code detection method has received more and more attention. First， summarize the traditional malicious code detection technology， and then introduces the current mainstream malicious code visualization methods， and then analyzes the machine learning and deep learning detection methods based on malicious code images， which specifically covers the model structure， innovation points and evaluation method used in the method. Finally，summarize the problems faced by the current detection technology， and explain the possible future research directions， aiming to help the development of malicious code detection technology.关键词：anti-detection technology;malicious code;data visualization;machine learning;deep learning
- 114
- |
- 30
- |
- 0
发布时间：2022-05-24 - 摘要：With the popularization of the teaching application of virtual reality technology， its influence on learning effects has attracted researchers' attention. In recent years， meta-analysis methods have been widely used to analyze the validity of the application of technology in education and teaching， and the research conclusions have a certain reference. Meta-analysis refers to the application of specific design and statistical methods to conduct overall and systematic quantitative and qualitative analysis of past research results.Adopt meta-analysis method to analyze the results of 31 relevant empirical research documents， and explores the teaching effect of different types of virtual reality technology， as well as the impact of discipline， school period， use period， knowledge type， etc. on the effect. Then further discuss and analyze the research conclusions， and propose suggestions on the application of virtual reality technology to teaching based on this.关键词：virtual reality technology;adjustment variable;teaching effect;empirical research;meta-analysis
- 42
- |
- 12
- |
- 0
发布时间：2022-05-24 - 摘要：Using convolutional neural networks in object detection can greatly improve the accuracy. It is of great value to study how to effectively use convolutional neural networks for object detection. Knowledge distillation is a representative type of model compression and acceleration. It can transfer the knowledge learned by a large network to a small network， so that the small network can obtain accuracy close to that of the large network. First，discuss the research prospects of target detection and some difficulties encountered at this stage. Based on the possibility of solving this problem based on the knowledge distillation method， the introduction of knowledge distillation can not only simplify the network， but also save the corresponding computing power and limited resources. Secondly， introduce the basic structure， research process and progress of knowledge distillation； Finally， introduce and compare typical object detection algorithms and analyze the effects of different improved distillation algorithms.关键词：deep learning;convolutional neural network;knowledge distillation;object detection;model compression
- 73
- |
- 16
- |
- 0
发布时间：2022-05-24 - 摘要：A new classification model of Chinese short text based on external knowledge attention was proposed to solve the problems of short text classification caused by insufficient contextual information and fuzzy semantics. Firstly， the model captures deep semantic information and solves the problem of context information by multiplying word and character features with trainable matrix to generate feature matrix with two levels of alignment. Secondly， this model retrievals knowledge from external knowledge base to enhance semantic representation of short texts， and introduces two attention mechanisms： short text oriented conceptual attention and concept set oriented conceptual attention. The experimental results show that the accuracy of the model is better than that of the existing models by adding the aligned eigenmatrix and using the external knowledge base to retrieve the knowledge for data training. The classification model of Chinese short text based on external knowledge attention proves that the introduction of alignment feature matrix and external knowledge attention is effective.关键词：short text classification;external knowledge;conceptual attention;convolutional neural network
- 59
- |
- 14
- |
- 0
发布时间：2022-05-24 - 摘要：In order to solve the problem of large prediction error of time series and nonlinear gas concentration series， a gated cycle unit neural network integrating attention mechanism is proposed to study gas prediction. Firstly， the algorithm preprocesses the data set， then introduces the update gate and reset gate to design the algorithm structure of the gated circulating unit neural network， and adds the attention mechanism to adjust the parameters of the hidden layer of the network， so as to predict the gas concentration with the goal of minimizing the error loss. Taking the monitoring data of Jilin slate gas disaster risk management and control platform as an example， the minimum root mean square error between the prediction result and the actual value is 3.95% and the minimum average absolute error is 0.71%. Compared with convolution neural network， cyclic neural network and multilayer sensor， the experiment shows that the prediction accuracy of this algorithm is higher than that of traditional methods.关键词：GRU;CNN;attention mechanism;multilayer perceptron;gas concentration prediction
- 21
- |
- 6
- |
- 0
发布时间：2022-05-24 - 摘要：With the rapid development of the we media industry， convenient information acquisition methods have created favorable conditions for the generation and dissemination of fake news. The widespread dissemination of fake news is extremely destructive to social stability. To study the text content of fake news， propose a fake news detection model based on the BERT model， while considering the semantic representation of sentences and the long-distance dependence between sentences. First， a comprehensive semantic representation of the we media news text is carried out by BERT model， and then the acquired semantic features of the words are input into the LSTM model for learning， finally detect fake news through the Softmax layer. In the experiments， the BERT model alone improved the performance over the FastText model by 4.58%， and the BtCNN model improved the detection performance over the Word2vec-based CNN model by 0.91%. Compared with the former models，BtLSTM model performed the best， with a detection performance of 92.15%.The experimental results on real data sets show that， compared with the baseline model， BtLSTM can better represent the semantic information of fake news and has the best fake news detection performance.关键词：BERT model;LSTM model;text semantic features;fake news detection
- 55
- |
- 10
- |
- 0
发布时间：2022-05-24 - 摘要：When the puncture robot performs the puncture action， the target lesion will be displaced due to the breathing motion， and the puncture system has inherent system delay， these factors will cause the puncture position to shift. In order to improve the accuracy of puncture， a relevant breathing prediction model can be established to compensate for breathing movement and system delays. Therefore propose a predictive model framework based on bidirectional long-term short-term memory neural network （Bi-LSTM）. A training set is made by selecting a piece of respiratory data， and then the trained predictive model is used to calculate the position information of the new respiratory data in the next 250ms. This article uses 29 sets of respiratory data （including self-test data and public data sets） for experiments， and does not deliberately intercept stable sections for experiments， which include complex conditions such as obvious changes in respiratory frequency amplitude and baseline drift. After counting the results of 29 sets of respiratory data with three algorithms of long short-term memory （LSTM）， gated recurrent unit （GRU）， and Bi-LSTM， the root mean square error （RMSE） is used， mean absolute value error （MAE） and the number of data points with an error greater than 0.5mm are used as reference standards. The experiment shows that the average RMSE and MAE of the Bi-LSTM prediction results are 0.091mm and 0.042mm， respectively， and its accuracy and robustness are better， and it can predict irregular breathing signals well.关键词：respiratory movement;prediction algorithm;Bi-LSTM;puncture robot
- 24
- |
- 9
- |
- 0
发布时间：2022-05-24 - 摘要：Due to the information overload caused by a large number of test questions， the personalized degree of test question recommendation is not high and the efficiency is low. According to the research of cognitive diagnosis， data mining and natural language processing， this paper proposes a test question recommendation method TCEGA based on genetic algorithm. This method determines the relationship between test questions and knowledge points according to the cognitive diagnosis model， and models the students' mastery level of test questions； Combined with the implicit semantic analysis， the data of the test database is processed， and the corresponding test questions are recommended by the difficulty of the test questions. TCEGA takes into account the individuality of the recommended students in learning and the generality of the group students in learning， so as to improve the rationality and accuracy of the recommendation. Finally， the comparative experiments show that the accuracy of the model is 90.17%， which is 11.79% higher than that of the traditional SOM algorithm， therefore， the method can be widely used in online learning application scenarios.关键词：recommendation of test questions;genetic algorithm;cognitive diagnosis;user interest;algorithm optimization
- 16
- |
- 5
- |
- 0
发布时间：2022-05-24 - 摘要：To address the problem that deep networks in the existing literature cannot adaptively select network layers based on traffic samples， a self-distillation-based adaptive malicious traffic classification algorithm is proposed. The method first pre-processes the original traffic as the input of the backbone network， constructs the weight distribution of the traffic by the self-attentive network layer， and then uses a one-dimensional convolutional neural network to extract the significant features in the traffic distribution as the input of the subsequent network. The branch network adaptively selects the network layer according to the entropy of the traffic samples， and returns early if it is less than the set threshold， otherwise the backbone network continues the inference. Experimentally verified， the method has an average detection rate of 99.9% for normal traffic and 99.96% for malicious traffic. The detection rate of malicious traffic is 2% higher than that of existing deep learning typical algorithms， and the detection rate of difficult samples is 5% higher， and the branching network has an adaptive function to avoid subsequent network inference.关键词：malicious traffic;self-distillation;self-attentive mechanism;self-adaptive
- 16
- |
- 3
- |
- 0
发布时间：2022-05-24 - 摘要：With the development of information technology， PDF documents， with good characteristics， have become a popular file format for data exchange. It has also become a file carrier that is often used in APT attacks. Existing malicious PDF document detection methods often use balanced sample data sets， but the number of malicious documents in the real environment is far less than that of benign documents. Therefore， in the case of unbalanced sample classification， a malicious PDF document detection method based on KM-TBSMOTE bi-directional sampling method is proposed. Based on the BSMOTE algorithm， the generated transition samples are used to synthesize new samples， and the TBSMOTE algorithm is given to increase the proportion of negative samples. The K-Means algorithm is used to down-sampling the samples of benign PDF documents， combined with the TBSMOTE algorithm， so that the sample classification reaches a balanced state. Finally， the random forest method is used for malicious detection. Experiments show that this method has a good detection effect on the unbalanced PDF sample set. The comprehensive evaluation index F1 reaches 98.98%， the recall rate is 98.91%， and the false positive rate is 0.026%. Compared with the traditional BSMOTE oversampling method， the evaluation index F1 is increased by 1.39%， the recall rate is increased by 1.96%， and the false detection rate is reduced by 0.048%. The malicious PDF documents detection method based on KM-TBSMOTE bi-directional sampling can effectively solve the impact of imbalanced sample classification on the classification model， improve the detection effect， and is suitable for the malicious detection of PDF documents in the real environment.关键词：malicious PDF;document detection;APT attack;unbalanced data;bi-directional sampling
- 16
- |
- 5
- |
- 0
发布时间：2022-05-24 - 摘要：Live video broadcasting is the hottest emerging industry in recent years.Due to the unique online real-time， simple language and Internet-based characteristics of the bullet screen， the existing methods are difficult to directly use the bullet screen's emotional analysis. In order to solve the problem of the accuracy of the analysis of the bullet screen text， a sentiment analysis model based on an improved SVM is proposed here for the characteristics of the language of the live broadcast bullet screen. On this basis， a method of fusing various features such as word vectors， sentiment words， negation words and punctuation is proposed， and the fusion results are mapped onto the vector space， and then the sentiment is classified by a classifier. The experimental results show that the improved SVM classifier model proposed is 3.8%， 2.3% and 1.1% higher than the evaluation indexes （accuracy， recall and F1 value） of the unimproved model， respectively.关键词：sentiment analysis;bullet screen;SVM algorithm;word vector;live online
- 18
- |
- 4
- |
- 0
发布时间：2022-05-24 - 摘要：The task of predicting drug-related legal provisions has difficulties such as high complexity of the case and high similarity between cases. Traditional methods mostly focus on the semantic learning of the case， while ignoring the role of legal knowledge， resulting in poor performance of legal predictions. high. Therefore， based on the KG-BERT algorithm， this article proposes an improved KG-Lawformer algorithm. The improved algorithm can learn case knowledge and legal knowledge at the same time， and better guide prediction through legal knowledge. Experiments have proved that the proposed method improves the macro F1 value by 10%~30% compared with the traditional method， reaching 79%， and has a certain improvement in the accuracy rate Acc， the macro precision rate MP， and the macro recall rate MR，which proved that the prediction performance can be improved by integrating law knowledge into law prediction.关键词：drug-related cases;law provision prediction;knowledge graph completion;KG-BERT;multi-label classification
- 18
- |
- 7
- |
- 0
发布时间：2022-05-24 - 摘要：In order to further strengthen road traffic safety management and improve the accuracy of road traffic safety early warning system， a road traffic accident severity prediction model based on XGBoost algorithm is proposed. First， SMOTE was used to process the unbalanced data set， and the number of positive and negative samples reached 1∶1. Then the random forest algorithm was used to rank the characteristics of urban road traffic accident severity in order of importance to find out the factors that had a greater impact on the prediction model. Finally， the prediction model is built based on XGBoost algorithm， and the grid search method is used to optimize the model parameters to improve the prediction accuracy. By comparing the results with KNN， Logistic and random forest models， the classification accuracy of XGBoost model was improved by 0.097 on average. The road traffic accident severity prediction model based on XGBoost algorithm has better prediction performance， which can provide a reliable reference for preventing and reducing the severity of traffic accidents.关键词：traffic safety;severity of traffic accidents;SMOTE;random forest;XGBoost
- 18
- |
- 6
- |
- 0
发布时间：2022-05-24 - 摘要：The cloud computing resource pool is composed of a large number of homogeneous physical machines. However， due to the different demands generated by users and applications， the computing energy consumption generated by each node is different， which makes certain nodes have long term high energy consumption and hot spots. Long-term hot spots on the node will damage the life of the physical machine. To solve the above problems， this paper proposes an energy prediction scheduling algorithm based on random forest（ECPRF）， which performs virtual machine migration according to energy consumption prediction， and balances the energy consumption of each physical machine. Experiments show that the scheduling algorithm can effectively balance server energy consumption and avoid hot spots.关键词：cloud computing;random forest;energy consumption prediction;ECPRF
- 32
- |
- 9
- |
- 0
发布时间：2022-05-24 - 摘要：The path planning aims to find a feasible path in the environment with obstacles， that is， the collision free path with the minimum length from the starting position to the target position. In order to solve the problem of obstacle avoidance path planning， proposes an improved method of grey wolf algorithm （GWO） for the optimization of group intelligence， which is used to solve the problem of inefficient path planning. The algorithm combines particle swarm optimization （PSO） and wolf algorithm， combines the two algorithms to improve the position updating formula of grey wolf algorithm. The algorithm is easy to converge early and fall into the problem of local optimal and low convergence speed. The experimental results show that compared with the basic grey wolf algorithm， the improved algorithm has different degrees of improvement in global optimal solution and convergence speed. The average solution time is only 89.8% of the basic algorithm， and the number of iterations is only 83%. The improved algorithm has better performance in path planning.关键词：robot;path optimization;particle swarm optimization;grey wolf algorithm
- 20
- |
- 5
- |
- 0
发布时间：2022-05-24 -
Heuristic Optimization Algorithm for Multi-constrained Three-dimensional Container Loading Problem AI导读
摘要：Aiming at the multi-constrained three-dimensional container loading problem， traditional heuristic algorithms are often difficult to meet all constraints to improve the container loading rate. Therefore， a heuristic optimization algorithm based on the block loading algorithm is proposed， which obtains simple blocks according to the block loading algorithm under the condition of satisfying multiple constraints， and determines the target space of block loading by dividing and merging the remaining space. The loading sequence is used to optimize the block selection in each loading stage to obtain the optimal loading scheme. The algorithm is tested by multi-constrained BRw cases generated by the BR classical cases combined with the normal distribution method. The experimental results show that the average loading rate of the heuristic optimization algorithm based on the block loading algorithm reaches 83.7%， which is 4% higher than the average loading rate of the traditional heuristic algorithm. The heuristic optimization algorithm based on the block loading algorithm can not only effectively deal with the various constraints in the loading process， but also increase the volume of each container loading， reduce the loading time， and improve the container loading rate.关键词：three-dimensional container loading problem;block loading algorithm;heuristic optimization algorithm;BRw cases- 13
- |
- 3
- |
- 0
发布时间：2022-05-24 - 摘要：In the field of earthquake prediction， topographic deformation observation is an important research direction. Ground tilt is a kind of topographical change， so the tiltmeter used to measure the amount of ground tilt has a certain guiding significance for the prediction of earthquake precursors. In order to better analyze the characteristics of the vertical pendulum tiltmeter circuit and the linearity of the test circuit output， this article uses Multisim to simulate the vertical pendulum tiltmeter circuit， and obtain the corresponding voltage data by setting the capacitance corresponding to different pendulum tilt angles. The simulation results show that the amplitude of the output voltage of the circuit changes with the tilt angle of the pendulum， and the positive and negative output voltage can reflect the forward and reverse bias of the pendulum， which is consistent with the actual circuit of the instrument； the least squares method is used to fit The linear equation of the output voltage is calculated， and the linearity of the output voltage is less than 0.4% through calculation， which verifies that the output and input of the vertical pendulum tiltmeter have a good linear relationship within its range.关键词：vertical pendulum tiltmeter;differential capacitance displacement sensor;circuit simulation;Multisim;linearity
- 13
- |
- 3
- |
- 0
发布时间：2022-05-24 - 摘要：SMOTE algorithm has a good classification for imbalanced data sets but a poor effect on the intra-class imbalance datasets， therefore the improved SMOTE algorithm based on Gaussian mixture clustering for imbalanced data sets is designed. Firstly， the GMM algorithm is used to cluster a small number of sample sets. Then the redundant samples overlapping with the cluster center point are deleted. Finally， the SMOTE algorithm is used to make the final balance of the data according to different clusters. The experiment compares the classification effect of RF，SMOTE+RF，GMM-SMOTE+RF algorithms on six sets of UCI standard open data sets. The results showed that the AUC value of the model proposed increased by 6.09% on average and it can effectively balance the imbalance datasets.关键词：GMM clustering;SMOTE algorithm;random forest;elbow method;imbalanced data sets
- 16
- |
- 5
- |
- 0
发布时间：2022-05-24 -
Application of Modified Hierarchical Base Scale Entropy in the Fault Diagnosis of Hydraulic Pump AI导读
摘要：The multi-scale entropy theory based on the definition of sample entropy has a large deviation when quantifying complex signals， and the coarse-grained method it adopts cannot effectively analyze the high-frequency components of vibration signals， and the utilization of fault information is low. Aiming at the above problems， a new method for quantifying vibration signal irregularities is proposed—modified hierarchical base-scale entropy （MHBSE）. MHBSE can not only overcome the shortcomings of sample entropy in the analysis of complex signals by performing hierarchical symbolization on time series， but also make full use of the faults information in the high frequency components of vibration signals， thereby improving the quality of features. In view of the excellent performance of MHBSE， a new method for detecting the health of hydraulic pumps is proposed. The effectiveness of the proposed method is tested using the collected hydraulic pump vibration experimental data. The experimental results show that the proposed method can fully extract the fault information in the hydraulic pump vibration signal， the extracted features can well characterize the different states of hydraulic pump， and the final fault recognition rate reached 100%.关键词：Modified hierarchical analysis;base-scale entropy;t-SNE;random forest;hydraulic pump;fault diagnosis- 10
- |
- 3
- |
- 0
发布时间：2022-05-24 - 摘要：In order to improve the utilization of the bus voltage of the common-bus open-winding permanent magnet synchronous motor （OW-PMSM） drive system and reduce the zero-sequence current in the system， an improved voltage space vector pulse width modulation （SVPWM） algorithm was proposed. By selecting a new basic voltage space vector， the linear modulation area of the dual inverter system is maximized， and at the same time， the zero-voltage vector is redistributed during the modulation process to suppress the zero-sequence current. Finally， a common bus OW-PMSM drive control system simulation is built based on MATLAB/Simulink environment. The simulation results show that the improved SVPWM algorithm has a higher utilization rate of the bus voltage， and the zero sequence current in the OW-PMSM phase current is reduced by 3A， and the third harmonic content is reduced by 51.47%.关键词：OW-PMSM;voltage utilization;zero sequence current;SVPWM;Simulink
- 15
- |
- 3
- |
- 0
发布时间：2022-05-24 -
Design and Implementation of Quality Evaluation System for Reservoir Safety Management Documents AI导读
摘要：In order to strengthen the supervision of reservoir safety management and improve the efficiency of the evaluation of reservoir safety management documents， by studying Chinese word segmentation technology， keyword learning method and document quality evaluation method， and proposes a document quality evaluation method based on term frequency for reservoir safety management. Choose jieba term segmentation tool and uses Django framework to develop quality evaluation system for reservoir safety management documents. Experimental results show that the system is accurate in evaluating the quality of reservoir safety management documents， which is basically consistent with the actual situation. The system can obtain the evaluation standard by learning all the similar documents and calculate the document quality index. It can be widely used in quality evaluation of reservoir safety management documents and has certain practical value.关键词：reservoir safety management;document quality evaluation;term frequency;jieba term segmentation;Django- 31
- |
- 12
- |
- 0
发布时间：2022-05-24 - 摘要：In order to solve the problem that there is no testing equipment for lower extremity training function in wheelchair-type lower extremity training equipment， a control system of testing platform for lower extremity training equipment was designed. The system adopted STM32F429IGT single-chip microcomputer as the master control chip， combined with the power supply module， communication module， sensor signal transmitting module， stepper motor driver module and the brake module， a test platform of hardware system， which can realize sensor collected data processing and transmission of power load and motor drive， restraint of tilt motion control. In terms of stepping motor control， SPWM wave was used on the basis of PWM wave to drive control， which can make the movement of the mechanism more stable. The motion simulation results show that the displacement of the centroid projection point changes steadily， which proves the feasibility of the control system.关键词：lower limb training test platform;STM32;SPWM;control system
- 24
- |
- 7
- |
- 0
发布时间：2022-05-24 - 摘要：Cloud transformation has brought an impact on traditional applications. Applications based on cloud native can give full play to the advantages of cloud platform and make the application adapt best to cloud platform. Introduce a video management system based on cloud native micro service architecture and container deployment. The system has the advantages of high expansion of cloud platform， elastic storage and load balancing. It can restart the micro service node at the second level and expand the node at the minute level. At the same time， micro services are divided by weighing the four elements of business， performance， reliability and stability， reducing the coupling between services， and adapting to business changes while improving R&D efficiency， so as to support the efficient management of video media files by enterprises.关键词：cloud native;micro-service architecture;video management system;Dubbo
- 37
- |
- 8
- |
- 0
发布时间：2022-05-24 - 摘要：At present， some scholars have tried to build the Internet of Things system based on blockchain technology， but due to the high demand of blockchain technology in computing resources， these Internet of Things systems often can not fit the actual application environment. In order to reduce the complexity of the Internet of Things system based on blockchain technology and meet the needs of practical application environment， a system architecture of the Internet of Things based on blockchain technology is proposed by analyzing the existing blockchain consensus mechanism. Firstly， the architecture divides all terminal devices into several regional networks with small coverage， which improves the flexibility of the system； Secondly， all the regional networks are organized into a blockchain network using lightweight consensus mechanism， which improves the security of the system； Finally， data users must apply to the blockchain network to obtain data， which increases the ability to control data access for the system and further improves the security of the system.关键词：blockchain;Internet of Things;consensus mechanism
- 28
- |
- 7
- |
- 0
发布时间：2022-05-24 - 摘要：In order to prevent single-column billboards from tilting or even collapsing under the influence of natural wind and surface subsidence， a single-column billboard monitoring system based on Internet of Things is designed. Through the sensor-based data acquisition unit， the system can monitor the inclination， amplitude and other state information of a billboard in real time. After the data preprocessing and the preliminary analysis of the original data， the recent health status of a billboard is calculated， and the future trend of the billboard status information is predicted through the Prophet time series prediction model. If any abnormality is found， the relevant staff will be reminded by an early warning. In order to reduce the cost of regional management of single-column billboards， GIS（Geographic Information System） and BIM（Building Information Model） are combined. The combined system connects the billboard and the basic geographic data of its surroundings， making the remote monitoring more intelligent and efficient. The system is developed using SpringBoot + Vue framework and is able to conduct real-time monitoring， visual management， early warning and alarm. When put into use， the system can help enterprises remotely monitor the status of billboards， reduce the probability of accidents， and provide a reference for relevant research.关键词：single-column billboard;sensor;BIM;GIS;Prophet time series forecasting
- 46
- |
- 7
- |
- 0
发布时间：2022-05-24 - 摘要：The data of Chinese patent medicine has the characteristics of large quantity and complex relationship. How to effectively store， manage， track and use the clinical， circulation and standard data of Chinese patent medicine has become the focus of the drug regulatory authorities. In order to help realize the knowledge integration of Chinese patent medicine， improve the data association， and mine the potential value of data. Methods the project used the knowledge map storage structure combined with visualization technology to organize the clinical technology， commercial circulation， standards and other information of Chinese patent medicine. Results the database system of Chinese patent medicine knowledge map was constructed， and the visualization platform of Chinese patent medicine knowledge map was built. The visualization platform based on knowledge mapping technology can better mine the potential value of the data. At the same time， the database of Chinese patent medicine knowledge mapping can provide data basis for intelligent question answering research， which has a great prospect of knowledge service.关键词：Chinese patent medicine;knowledge map;visualization;platform construction
- 20
- |
- 7
- |
- 0
发布时间：2022-05-24 - 摘要：The evaluation results are inconsistent due to the different standards used to evaluate the anti epidemic level in various regions. Therefore， the anti epidemic level of five provinces and regions in China was compared and analyzed by WSM， TOPSIS， VIKOR and ELECTRE （MCDM method for short）. Firstly， the Shannon entropy weight method is used to evaluate the importance of each standard to the anti epidemic level of five provinces， cities and regions in China， and then the quantitative evaluation is carried out based on MCDM method. Considering that the ranking results depend on the standard weight， the sensitivity analysis of the weight is carried out at last. The practice results show that transmission control is the most important standard to evaluate the level of epidemic resistance， followed by emergency rescue and vaccine research and development， and the economic and social impact is the least. According to the obtained standards， in order to provide reference and reference for combating the epidemic situation in various regions.关键词：multi-criteria decision making;COVID-19 epidemic;sensitivity analysis;standard weight
- 13
- |
- 4
- |
- 0
发布时间：2022-05-24 -
Construction of Vehicle Driving Cycle Based on Principal Component Analysis and K-means Clustering AI导读
摘要：The vehicle driving condition diagram is the most basic basis for vehicle development and evaluation. Based on the GPS data of vehicle driving， use a variety of data processing and data analysis methods comprehensively and construct the vehicle driving cycle diagram with small errors according to the kinematics segment. First， the data were cleaned and the kinematics fragments were extracted. Secondly， 15 characteristic parameters were used to extract features from the kinematics fragments. Then， principal component analysis and K-means clustering were carried out on the characteristic parameter matrix to construct a driving cycle curve. Finally， error analysis was conducted on the constructed driving cycle curve and the sampling population， and the relative error was 0.84. The results verified the validity and rationality of the vehicle driving cycle diagram constructed.关键词：clustering;driving cycle;kinematic segment;principal component analysis- 18
- |
- 3
- |
- 0
发布时间：2022-05-24 - 摘要：The artificial bee colony algorithm has the shortcomings of insufficient convergence accuracy and easy to fall into the local optimal solution. An artificial bee colony algorithm based on adaptive logarithmic convergence double search strategy is proposed. The adaptive coefficient and logarithmic convergence coefficient are introduced to improve the search strategy in the hiring bee stage and the observation bee stage to prevent the algorithm from falling into local extremum； at the same time， a double search strategy is introduced to speed up the convergence speed and improve the accuracy. This algorithm is applied and an improved artificial bee colony algorithm is proposed for the registration of heterologous images of apples in orchards in natural scenes. The algorithm uses the non-rotation factor of the SURF algorithm and the response threshold of the Hessian matrix as the initial two-dimensional vector of the improved artificial bee colony algorithm， and uses the root mean square error function as the fitness function for optimization search. In the experimental stage， the salps swarm algorithm， the bat algorithm and an improved artificial bee swarm algorithm are introduced to compare the performance. The experimental results show that the proposed algorithm improves the optimization performance of artificial bee colony algorithm by 50%~90% under the condition of fixed global evolution times； The results of heterologous image registration can get more correct feature point pairing than the traditional SURF algorithm.关键词：artificial bee colony algorithm;adaptive coefficient;logarithmic convergence coefficient;double search strategy;heterogeneous image registration
- 12
- |
- 4
- |
- 0
发布时间：2022-05-24 - 摘要：For the characteristics of rice blast microscopic images， a method of applying SVM to the rapid classification of rice blast in rice is proposed. Firstly， the rice blast microscopic images are preprocessed， then shape and texture features are extracted， and finally the extracted features are identified as whether they are rice blast using SVM. The results show that SVM is suitable for the classification problem of small sample rice plague spore microscopic images. Four kernel functions were used in the study for the classification accuracy comparison test， and the highest SVM classification accuracy of 98.7% was achieved using the radial basis kernel function.关键词：SVM;image classification;rice blast microscopic image;image processing;feature extraction
- 15
- |
- 4
- |
- 0
发布时间：2022-05-24 - 摘要：In order to solve the problem of misrecognition of similar characters in license plate image，propose a template matching method integrating the improved local HOG feature to recognize license plate. Compared with the template matching method， the template matching method combined with jump feature and the template matching method combined with local HOG feature， the recognition rate of the method proposed increases by 52%， 12% and 4% respectively when testing the first class license plate images. When the second type of license plate image is tested， the recognition rate is improved by 36%， 28% and 4% respectively. The improved local HOG feature determines the optimal parameters of the local feature block of the character image， obtains the local optimal feature block， reduces the dimension of feature descriptor， and improves the recognition rate while ensuring the recognition rate.关键词：improved local HOG feature;template matching;character recognition
- 17
- |
- 5
- |
- 0
发布时间：2022-05-24 - 摘要：Course ideology and politics is an important method to actively implement the Party’s educational policy in the new era. Expound the concept and connotation of course ideology and politics， and discuss three methods of integrating course ideology and politics into professional courses. Furthermore， explore the specific methods of course ideology and politics for the computer vision course. Through identifying the entry point in the course design and using a variety of teaching methods in the course teaching， it successfully helps students establish a scientific and rigorous way of thinking， and stimulates students' patriotic feelings and cultural self-confidence.关键词：course ideology and politics;computer vision;professional course
- 40
- |
- 14
- |
- 0
发布时间：2022-05-24 - 摘要：Pattern recognition is a comprehensive and cross-disciplinary course involving computers， statistics， psychology and many other disciplines. It mainly introduces the basic concepts， basic theories， typical methods and applications of pattern recognition technology. It is a series of professional courses for undergraduates majoring in computer science and technology， and related majors. Therefore，first analyzes and summarizes the current research status and existing problems of of pattern recognition， then designs a kind of teaching model for pattern recognition based on the hybrid teaching strategy， so as to provide a reference for teachers engaged in the teaching work related to pattern recognition， thus futher improve the teaching quality of pattern recognition and further improve students' professional ability and literacy.关键词：pattern recognition;online and offline;mixed teaching;teaching design
- 16
- |
- 8
- |
- 0
发布时间：2022-05-24 - 摘要：To cultivate artificial intelligence talents suitable for the needs of society， teaching methods must be reformed. Discrete mathematics course is both a theoretical foundation and a practical tool. The teaching of discrete mathematics is reformed by taking the content according to the professional needs， optimizing the teaching methods， changing the teaching mode， selecting interesting cases and fully using mathematical software for teaching； adopting diversified evaluation methods to stimulate students' innovation consciousness， strengthen the scientific spirit and apply what they have learned for practical innovation. The experimental class was compared with the control class， and the results showed that the experimental class had 14% higher excellence rate， 22% higher pass rate and 14% higher student satisfaction rate for the course teaching than the control group， and the course reform achieved more obvious effect.关键词：information technology;artificial intelligence;discrete mathematics
- 15
- |
- 3
- |
- 0
发布时间：2022-05-24 - 摘要：According to the implementation opinions on deepening the reform of innovation and entrepreneurship education in Colleges and universities put forward by the State Council， and implement the goal of cultivating high-quality applied "three innovation" talents with "employment ability， the foundation for further study， the potential for development and entrepreneurship". This paper explores from the aspects of formulating the overall goal and orientation of the "three innovations" talent training， planning the overall hierarchical structure of the "three innovations" talent implementation scheme， the reform of the educational curriculum system of the integration of specialty and innovation， the whole process of the "three innovations" curriculum construction of computer majors， the reform of teaching contents and teaching methods， and the reform of assessment methods and evaluation methods. Taking the computing major of our university as an example， the students' computing thinking ability， programming practice ability， project development ability， innovative application ability， and comprehensive ability have been significantly improved.关键词：three innovations;computer speciality;integration of expertise and innovation;application type;talents cultivation
- 21
- |
- 12
- |
- 0
发布时间：2022-05-24 - 摘要：Under the background of the new engineering disciplines， the cultivation of talents should not only continuously reform and improve teaching methods， but also pay attention to ideological and political education. In order to cultivate computer professionals with both ability and political integrity in a better way， the first course-- "Introduction to Computer" encountered by freshmen in computer science is reformed from the two dimensions of teaching methods and ideological and political education，thus propose a "student-centered" BOPPPS teaching model， and explores ideological and political elements in teaching and integrates them into teaching design. Practice shows that the new teaching model can stimulate students' interest in learning， improve students' knowledge level of the introduction to computer course， and cultivate students' correct understanding of their outlook on life and values. This kind of teaching mode has a good application effect， and has certain guiding significance for the teaching reform of the introduction to computer course.关键词：introduction to computer;freshmen;teaching methods;ideological and political education;BOPPPS
- 24
- |
- 5
- |
- 0
发布时间：2022-05-24 - 摘要：In order to adapt to the 1+X vocational skill level examination of vocational colleges， improves the teaching content and teaching means of Java courses，and propose a "graded project" teaching scheme based on OBE. The scheme was guided by the skill requirements of skill level examination， and based on the linking knowledge points of textual research， divides the project-based teaching process into four levels of projects， and assists in classroom screen recording and online courses at the same time. The teaching plan was implemented among 130 students and compared with project-based teaching. The implementation results showed that the students who use the teaching scheme in this paper have improved their task scores even when the knowledge point is more difficult， and they also approve the teaching scheme very much. OBE-based "graded project" teaching not only improves students' skill level， improves the pass rate of the certificate， but more importantly， inspires students' programming ideas， develops their learning potential and promotes students' independent learning.关键词：OBE;graded project;Java;1 + X
- 28
- |
- 3
- |
- 0
发布时间：2022-05-24 - 摘要：Functional programming in Haskell basic is an introductory public course for all undergraduate majors， which uses functional language to teach programming. In response to the problems of small number of elective students， students worrying about the difficulty of the course and early withdrawal from the course with low final scores， more students persist in completing the course through the practice of flipped classroom teaching by combining offline and online， using electronic classroom and other technical means， and students gradually adapt to and enjoy the flipped classroom； designing typical examples to show the features and simple beauty of functional programming， more students like programming. Statistics showed that the number of students who completed the academic year in which the combination of online and offline teaching was implemented increased by 20% compared with the previous year， and the average grade， pass rate and merit rate of students also increased significantly compared with previous years.关键词：functional programming in Haskell basic;course completion rate;flipped classroom;MOOC;online evaluation
- 19
- |
- 7
- |
- 0
发布时间：2022-05-24 - 摘要：Under the background of new liberal arts construction， it is more in line with the needs of the industry to cultivate students' data thinking and ability.Taking the news construction of the Department of media of Central South University as the research object， reforms the curriculum system structure of relevant majors， completes the construction of information cutting-edge courses and plates， constructs the media big data case base， explores the professional direction of emerging disciplines of media intelligence， improves students' Industrial data analysis and practice ability， forms the data thinking of new liberal arts talents， and cultivates innovative media practitioners with interdisciplinary literacy.After eight years of reform and practice， students began to learn professional core knowledge independently， spontaneously explore cutting-edge hot issues in the discipline field， actively devote themselves to practical activities， form a comprehensive knowledge reserve and data thinking， and better meet the social demand for compound new cultural， scientific and innovative talents.关键词：new liberal arts;data thinking;cultivation of talents
- 15
- |
- 4
- |
- 0
发布时间：2022-05-24 - 摘要：In order to improve the academic challenge of college students， a "golden course" with depth， difficulty and challenge should be built. Firstly， taking the course of software project management as an example， put forward the construction and application of software engineering course with the integration of industry and education， the parallel of science and education and the combination of innovation and entrepreneurship. Then， a "golden course" with the combination of theory and practice， the coexistence of tradition and innovation， and challenging teaching assessment is explored and established. Finally， the construction and application of "golden course" have achieved good results， providing more experience for professional construction， integration of industry and education and other "golden course" construction.关键词：new engineering;integration of industry and education;collaborative education;Golden Course;software project management
- 15
- |
- 5
- |
- 0
发布时间：2022-05-24 - 摘要：In order to solve the problems of lack of self-learning resources， poor methods to improve practical ability， difficult comparison of curriculum assessment and no path of curriculum thought and politics in the teaching of C language programming course in local colleges and universities. Through the systematic construction of the curriculum objectives， teachers' team， curriculum resources， teaching content， teaching organization， assessment and evaluation of the course， a teaching mode of "three classrooms， four connections and five combinations" integrating "learning， training and breakthrough" has been formed， which has achieved good teaching results and significantly improved students' performance， programming ability， innovative practice ability and value guidance. Practice has proved that in the past five years， students have won 157 awards in program design competitions， applied for 139 undergraduate research-based learning and innovative experimental projects， applied for 196 patents and software copyrights， and participated in more than 20 local software development projects. The curriculum construction has achieved phased success.关键词：C language programming design;course construction;teaching mode;innovative and practical ability
- 28
- |
- 6
- |
- 0
发布时间：2022-05-24 -
Exploration on the Teaching Reform of "Fundamentals of Secrecy Technology" under the New Situations AI导读
摘要：With the rise of a new round of scientific and technological revolution and industrial reform， there are huge security risks in the use of foreign introduced technology. Information secrecy is facing various new situations， and the development of secrecy technology is facing unprecedented new challenges. Therefore， it is urgent to cultivate secrecy technical talents under the current new situation. Taking the course of "Fundamentals of secrecy technology" as the object， this paper puts forward a "four in one" teaching and education reform system： innovating the training objectives， strengthening the confidentiality awareness， optimizing the teaching content， enriching the teaching means， and cultivating compound security technology talents who adapt to the new situation and meet the new needs. Practice has proved that the curriculum construction has achieved good results， and the students' professional knowledge level and ability have been effectively improved.关键词：four in one;secrecy technology;teaching reform- 13
- |
- 5
- |
- 0
发布时间：2022-05-24
- 摘要：In actual production， different suppliers and transporters have different supply and transportation of raw materials at different times. In order to reduce resource consumption and meet production demand， time series and principal component analysis are used to study the material ordering and transportation planning of an enterprise. Taking the weekly order quantity of the enterprise in five years， the supply quantity of 402 suppliers and the transportation loss rate of 8 transporters as the research object， the comprehensive evaluation and quantitative analysis of its suppliers and transporters are carried out， and the specific ranking is given. According to the ranking， the enterprise's raw material ordering and transshipment planning strategy for the next 24 weeks is formulated through seasonal periodic time prediction and target planning. This strategy can not only improve the ordering and transportation efficiency of raw materials， but also improve the production efficiency.关键词：quantitative analysis;main component analysis;seasonal periodic time series;target planning
- 15
- |
- 4
- |
- 0
发布时间：2022-05-24
大数据与信息处理
- Technical support is provided by Beijing Founder electronics co., LTD 鄂ICP备12016542号-70
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0