Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10065995
Shaojie He, Bihui Yu, Jingxuan Wei, Liping Bu
Cloud computing divides a huge program into countless subtasks through the network, which are calculated and analyzed by multiple servers, and then the results are returned to users. Therefore, the strategy of task scheduling is very important for computing performance. Aiming at the essence of cloud computing task scheduling and the optimization problem of seeking solutions, this paper proposes a hybrid algorithm called MMES algorithm (MA-MIX-ESDA). This algorithm not only guarantees the search space of electrostatic discharge algorithm (ESDA), but also accelerates its convergence speed, and solves the problem that mayfly algorithm (MA) is easy to fall into local optimization. Latin hypercube sampling is used for population initialization, exploration and development are balanced by the direction of the balance vector, and the step size control factor is added to jump out of local optimization. In order to evaluate the performance of the algorithm, 23 groups of test functions commonly used by CEC and 30 benchmark functions of CEC2014 are used to test the global search and local development functions of the algorithm, and the results are compared with the improved algorithm and classical algorithm. Experimental results show that the proposed MMES algorithm is more superior in search space and convergence speed.
{"title":"MMES: Improved Mayfly Algorithm Based on Electrostatic Optimization Algorithm","authors":"Shaojie He, Bihui Yu, Jingxuan Wei, Liping Bu","doi":"10.1109/ICCC56324.2022.10065995","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10065995","url":null,"abstract":"Cloud computing divides a huge program into countless subtasks through the network, which are calculated and analyzed by multiple servers, and then the results are returned to users. Therefore, the strategy of task scheduling is very important for computing performance. Aiming at the essence of cloud computing task scheduling and the optimization problem of seeking solutions, this paper proposes a hybrid algorithm called MMES algorithm (MA-MIX-ESDA). This algorithm not only guarantees the search space of electrostatic discharge algorithm (ESDA), but also accelerates its convergence speed, and solves the problem that mayfly algorithm (MA) is easy to fall into local optimization. Latin hypercube sampling is used for population initialization, exploration and development are balanced by the direction of the balance vector, and the step size control factor is added to jump out of local optimization. In order to evaluate the performance of the algorithm, 23 groups of test functions commonly used by CEC and 30 benchmark functions of CEC2014 are used to test the global search and local development functions of the algorithm, and the results are compared with the improved algorithm and classical algorithm. Experimental results show that the proposed MMES algorithm is more superior in search space and convergence speed.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129795394","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10066004
Xinjian Zhao, Fei Xia, Guoquan Yuan, Shi Chen, Hu Song
The group relationship (Community Relation) contained in the trajectory data can be used for hot spot exploration, community governance, and traffic diversion, which has broad application prospects. Trajectory group association privacy refers to the user relationship with a similar movement mode in the trajectory data. Publishing trajectory data to analysts without protection will cause the leakage of such privacy. Recently, trajectory correlation privacy has attracted the attention of researchers, proposing solutions based on differential privacy. Still, existing methods are limited to protecting the motion patterns of two users and cannot be used in multi-user scenarios. Moreover, existing methods use heuristic strategies to reconstruct trajectories, which have excessive noise increase and large loss of published trajectory availability. Because of the above problems, we design a probability differentiation tree (PDT) structure to describe the user's movement pattern, then define the probability differentiation tree similarity function. A noise probability differentiation tree generation algorithm (NPDT) is proposed to realize the trajectory of user-associated privacy protection by adding Laplace noise to the probability value of PDT. We also propose the trajectory reconstruction algorithm (TRA) to reconstruct each user trajectory through the noise probability differentiation tree, noise trajectory number distribution, and noise trajectory length distribution to form the final published trajectory data set. Theoretical analysis and experimental results show that the proposed privacy protection method effectively maintains the availability of trajectory data while improving the privacy protection intensity of group association.
{"title":"A Group-Correlated Privacy Protection Trajectory Publishing Method Based on Differential Privacy","authors":"Xinjian Zhao, Fei Xia, Guoquan Yuan, Shi Chen, Hu Song","doi":"10.1109/ICCC56324.2022.10066004","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10066004","url":null,"abstract":"The group relationship (Community Relation) contained in the trajectory data can be used for hot spot exploration, community governance, and traffic diversion, which has broad application prospects. Trajectory group association privacy refers to the user relationship with a similar movement mode in the trajectory data. Publishing trajectory data to analysts without protection will cause the leakage of such privacy. Recently, trajectory correlation privacy has attracted the attention of researchers, proposing solutions based on differential privacy. Still, existing methods are limited to protecting the motion patterns of two users and cannot be used in multi-user scenarios. Moreover, existing methods use heuristic strategies to reconstruct trajectories, which have excessive noise increase and large loss of published trajectory availability. Because of the above problems, we design a probability differentiation tree (PDT) structure to describe the user's movement pattern, then define the probability differentiation tree similarity function. A noise probability differentiation tree generation algorithm (NPDT) is proposed to realize the trajectory of user-associated privacy protection by adding Laplace noise to the probability value of PDT. We also propose the trajectory reconstruction algorithm (TRA) to reconstruct each user trajectory through the noise probability differentiation tree, noise trajectory number distribution, and noise trajectory length distribution to form the final published trajectory data set. Theoretical analysis and experimental results show that the proposed privacy protection method effectively maintains the availability of trajectory data while improving the privacy protection intensity of group association.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129633604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10065691
Emad Naji, Bin Dai
Reconfigurable Intelligent Surfaces (RISs) have the ability to make the concept of smart radio environments a reality, by utilizing the special characteristics of meta-surfaces. In this paper, we discuss how an IRS-assisted enhance link quality and coverage between an access point (AP) located on a wall and an antenna user in an indoor environment. specifically, we formulate and solve a non-convex constraint issue to minimize transmit power at the antenna of the transmitter and maximize the received power at user-end by optimizing both phase/amplitude shifts, as well as maximizing Energy Efficiency (EE) by proposing an Optimizing Alternating (OA) technique to solve that issue. The result of simulation show that IRS helps the indoor environment to gain a strong signal and make a virtual link between the AP and USER. Moreover, it is verified that the IRS by joint amplitude/phase shifts and OA are able to make a significant improvement of about 10 dBm by maximizing both discrete phase/amplitude shifts. Also, the IRS be able to create “signal hot-spots” in some points between the IRS and USER to deliver a strong signal and produce 30% improvement as well as maximizing the energy efficiency and keep it highest until 30 dB of (signal to noise ratio) SNR. In this paper, we assume that the user is in a bad situation and not be able to receive a good signal from the base station in which by helping RIS the user receives a higher SNR. Finally, we compare our work with a reference system that only uses non-direct NLOS transmission.
{"title":"Power Optimization for Intelligent Reconfigurable Surfaces in Indoor Environment Using Discrete Phase and Amplitude Shifts","authors":"Emad Naji, Bin Dai","doi":"10.1109/ICCC56324.2022.10065691","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10065691","url":null,"abstract":"Reconfigurable Intelligent Surfaces (RISs) have the ability to make the concept of smart radio environments a reality, by utilizing the special characteristics of meta-surfaces. In this paper, we discuss how an IRS-assisted enhance link quality and coverage between an access point (AP) located on a wall and an antenna user in an indoor environment. specifically, we formulate and solve a non-convex constraint issue to minimize transmit power at the antenna of the transmitter and maximize the received power at user-end by optimizing both phase/amplitude shifts, as well as maximizing Energy Efficiency (EE) by proposing an Optimizing Alternating (OA) technique to solve that issue. The result of simulation show that IRS helps the indoor environment to gain a strong signal and make a virtual link between the AP and USER. Moreover, it is verified that the IRS by joint amplitude/phase shifts and OA are able to make a significant improvement of about 10 dBm by maximizing both discrete phase/amplitude shifts. Also, the IRS be able to create “signal hot-spots” in some points between the IRS and USER to deliver a strong signal and produce 30% improvement as well as maximizing the energy efficiency and keep it highest until 30 dB of (signal to noise ratio) SNR. In this paper, we assume that the user is in a bad situation and not be able to receive a good signal from the base station in which by helping RIS the user receives a higher SNR. Finally, we compare our work with a reference system that only uses non-direct NLOS transmission.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127206930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10065699
YiYan Duan, Yong Liu
Due to the advent of the big data era and the redundancy of industrial data, people use a series of information technologies to continuously promote the transformation of traditional industry, and the development of industry is bound to move into the era of intelligence. The article carries out the construction of process knowledge mapping on the existing industrial process design scheme, constructs a process ontology model through industrial specifications, starts from acquiring data from industrial archives, identification of industrial entities, extraction of relationships between industrial entities and fusion of process knowledge, and constructs a process assembly knowledge mapping in the forging field by storing the acquired process data into the graph database, and finally The visualization display of data storage based on database Neo4j is realized. The experimental results have verified the feasibility of the designed diagram.
{"title":"Ontology-Based Knowledge Graph Construction and Application for Large Workpiece Forging","authors":"YiYan Duan, Yong Liu","doi":"10.1109/ICCC56324.2022.10065699","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10065699","url":null,"abstract":"Due to the advent of the big data era and the redundancy of industrial data, people use a series of information technologies to continuously promote the transformation of traditional industry, and the development of industry is bound to move into the era of intelligence. The article carries out the construction of process knowledge mapping on the existing industrial process design scheme, constructs a process ontology model through industrial specifications, starts from acquiring data from industrial archives, identification of industrial entities, extraction of relationships between industrial entities and fusion of process knowledge, and constructs a process assembly knowledge mapping in the forging field by storing the acquired process data into the graph database, and finally The visualization display of data storage based on database Neo4j is realized. The experimental results have verified the feasibility of the designed diagram.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127293171","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10066011
Xiang Li, Zhi Zeng, Mingmin Wu, Zhongqiang Huang, Ying Sha, Lei Shi
Various forms of social interactions are often char-acterized by toxic or offensive words that can be collectively referred to as offensive languages, which has become a unique linguistic phenomenon in social media platforms. How to detect and identify these offensive languages in social media platforms has become one of the important research in the field of natural language processing. Existing methods utilize machine learning algorithms or text representation models based on deep learning to learn the features of offensive languages and identify them, which have achieved good performances. However, traditional machine learning-based methods mainly rely on keyword identi-fication and blocking, deep learning-based methods do not ade-quately explore the fused deep semantic features of the content by combining word-level embeddings and sentence-level deep semantic feature representations of sentences, which cannot ef-fectively identify offensive languages that do not contain common offensive words but indicate offensive meanings. In this research, we propose a novel offensive language identification model based on deep semantic feature fusion, which uses the pre-trained model Bert to obtain word-level embedding representations of offensive languages, and then integrates the RCNN that combines with the attention mechanism to extract the fused deep semantic feature representations of offensive languages, and label encoder and offensive predictor to improve the identification accuracy and generalization ability of the model so that the performances of the model do not rely on the offensive language lexicon entirely and can identify offensive languages that do not contain common offensive words but indicate offensive meanings. Experimental results on Wikipedia and Twitter comment datasets show that our proposed model can better understand the context and discover potential offensive meanings, and outperforms existing methods.
{"title":"An Offensive Language Identification Based on Deep Semantic Feature Fusion","authors":"Xiang Li, Zhi Zeng, Mingmin Wu, Zhongqiang Huang, Ying Sha, Lei Shi","doi":"10.1109/ICCC56324.2022.10066011","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10066011","url":null,"abstract":"Various forms of social interactions are often char-acterized by toxic or offensive words that can be collectively referred to as offensive languages, which has become a unique linguistic phenomenon in social media platforms. How to detect and identify these offensive languages in social media platforms has become one of the important research in the field of natural language processing. Existing methods utilize machine learning algorithms or text representation models based on deep learning to learn the features of offensive languages and identify them, which have achieved good performances. However, traditional machine learning-based methods mainly rely on keyword identi-fication and blocking, deep learning-based methods do not ade-quately explore the fused deep semantic features of the content by combining word-level embeddings and sentence-level deep semantic feature representations of sentences, which cannot ef-fectively identify offensive languages that do not contain common offensive words but indicate offensive meanings. In this research, we propose a novel offensive language identification model based on deep semantic feature fusion, which uses the pre-trained model Bert to obtain word-level embedding representations of offensive languages, and then integrates the RCNN that combines with the attention mechanism to extract the fused deep semantic feature representations of offensive languages, and label encoder and offensive predictor to improve the identification accuracy and generalization ability of the model so that the performances of the model do not rely on the offensive language lexicon entirely and can identify offensive languages that do not contain common offensive words but indicate offensive meanings. Experimental results on Wikipedia and Twitter comment datasets show that our proposed model can better understand the context and discover potential offensive meanings, and outperforms existing methods.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130756313","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10066003
Jie Li, Yanxiang Gong, Zheng Ma, M. Xie
At present, object detection performance can meet some routine tasks' requirements. However, the detection performance for small-sized objects is far from satisfactory. Therefore, we propose the feature layer attention module and nonlinear positioning loss penalty based on size to improve small object detection performance. Our work proposes the feature layer attention module, which introduces an attention mechanism in the feature layer to enhance the model's attention to small objects. Through the feature fusion scheme proposed in this paper, we solve the problem of insufficient features of small objects to a certain extent and reduce the difficulty of model training. Besides, we introduce a size-based nonlinear penalty in the loss function, which can enhance the penalty for small object positioning errors. The effectiveness of our method has been demonstrated on small object data sets. On VisDrone2019 dataset, the proposed method improves the detection's AP by 2.2%. On TT100k dataset, the proposed method improves the detection's AP by 1.0%.
{"title":"Enhancing Feature Fusion Using Attention for Small Object Detection","authors":"Jie Li, Yanxiang Gong, Zheng Ma, M. Xie","doi":"10.1109/ICCC56324.2022.10066003","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10066003","url":null,"abstract":"At present, object detection performance can meet some routine tasks' requirements. However, the detection performance for small-sized objects is far from satisfactory. Therefore, we propose the feature layer attention module and nonlinear positioning loss penalty based on size to improve small object detection performance. Our work proposes the feature layer attention module, which introduces an attention mechanism in the feature layer to enhance the model's attention to small objects. Through the feature fusion scheme proposed in this paper, we solve the problem of insufficient features of small objects to a certain extent and reduce the difficulty of model training. Besides, we introduce a size-based nonlinear penalty in the loss function, which can enhance the penalty for small object positioning errors. The effectiveness of our method has been demonstrated on small object data sets. On VisDrone2019 dataset, the proposed method improves the detection's AP by 2.2%. On TT100k dataset, the proposed method improves the detection's AP by 1.0%.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132500495","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10065801
Shenghong Qin, Laixian Peng, Renhui Xu, Bili Wang
This paper considers an Aerial 3D Mesh Network (A3DMN) with unmanned aerial vehicles (UAVs). The spatial distribution of UAVs in A3DMN plays a crucial role in evaluating mutual interference and connectivity performance. This paper analyzes how antenna directionality affects network connectivity under interference constraints without any channel contention or transmission power control. We introduce an intermediate class ß-Ginibre Point Processes (ß-GPP) between the Poisson point process (PPP) and the GPP as a model for the A3DMN when UAVs exhibit repulsion. The closed-form expressions for connection probability and network coverage radius are derived by stochastic-geometry tools. Simulation results show that the derived theoretical expressions accurately reflect the influence of antenna directivity on the performance of the A3DMN. The results indicate that the directionality of transmitter antennas and the regularity of transmitter location distribution help to improve network connectivity.
{"title":"Enhanced Connectivity of Aerial 3D Mesh Network with Directional Antennas","authors":"Shenghong Qin, Laixian Peng, Renhui Xu, Bili Wang","doi":"10.1109/ICCC56324.2022.10065801","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10065801","url":null,"abstract":"This paper considers an Aerial 3D Mesh Network (A3DMN) with unmanned aerial vehicles (UAVs). The spatial distribution of UAVs in A3DMN plays a crucial role in evaluating mutual interference and connectivity performance. This paper analyzes how antenna directionality affects network connectivity under interference constraints without any channel contention or transmission power control. We introduce an intermediate class ß-Ginibre Point Processes (ß-GPP) between the Poisson point process (PPP) and the GPP as a model for the A3DMN when UAVs exhibit repulsion. The closed-form expressions for connection probability and network coverage radius are derived by stochastic-geometry tools. Simulation results show that the derived theoretical expressions accurately reflect the influence of antenna directivity on the performance of the A3DMN. The results indicate that the directionality of transmitter antennas and the regularity of transmitter location distribution help to improve network connectivity.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131948990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-09DOI: 10.1109/ICCC56324.2022.10065991
Songze Li, Guoliang Xu, Yang Liu
In recent years, fraudulent methods are constantly updated, criminal information is more hidden, and there is a problem of subjectivity of artificial feature design in traditional model feature engineering. To address this problem, a model based on broad learning and dual-channel convolutional neural network is proposed (BLS-DCCNN). First, the broad learning system is transformed from a supervised prediction method to an integrated feature generation method to generate mapped features and enhanced features for the original data. Then, the generated features are reconstructed to integrate the module to reconstruct the data distribution. Finally, a dual-channel convolutional neural network is combined with the shallow and deep layer network structure to extract global and local features, predict the final category labels, and introduce the Focal Loss function is introduced to solve the problem of positive and negative sample imbalance. Experiments and model comparisons are conducted on real telecommunication datasets, and the exper-imental results show that the model has significantly improved both accuracy, recall and F1 scores compared with traditional machine learning models such as support vector machines and random forests, and deep learning models such as long and short term memory networks.
{"title":"Fraud Call Identification Based on Broad Learning System and Convolutional Neural Networks","authors":"Songze Li, Guoliang Xu, Yang Liu","doi":"10.1109/ICCC56324.2022.10065991","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10065991","url":null,"abstract":"In recent years, fraudulent methods are constantly updated, criminal information is more hidden, and there is a problem of subjectivity of artificial feature design in traditional model feature engineering. To address this problem, a model based on broad learning and dual-channel convolutional neural network is proposed (BLS-DCCNN). First, the broad learning system is transformed from a supervised prediction method to an integrated feature generation method to generate mapped features and enhanced features for the original data. Then, the generated features are reconstructed to integrate the module to reconstruct the data distribution. Finally, a dual-channel convolutional neural network is combined with the shallow and deep layer network structure to extract global and local features, predict the final category labels, and introduce the Focal Loss function is introduced to solve the problem of positive and negative sample imbalance. Experiments and model comparisons are conducted on real telecommunication datasets, and the exper-imental results show that the model has significantly improved both accuracy, recall and F1 scores compared with traditional machine learning models such as support vector machines and random forests, and deep learning models such as long and short term memory networks.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131014627","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study the problem of multiple resource allocation in cloud computing systems. Existing fairness-efficiency scheduling procedures can relax fairness constraints by using a knob to improve efficiency. However, these approaches do not take into account users with special needs, i.e., the same resource (meta-type, e.g., CPU) contains different types (e.g., Intel's CPU, AMD's CPU) and the user can only use a specific type of resources (e.g., Intel's CPU). We propose a new allocation mechanism called Fairness-Efficiency Tradeoff Allocation with Meta-Types (FET-MT), which introduces the concept of meta-types. FET-MT not only meets specific requirements proposed by users but also allows users to flexibly balance fairness and efficiency by adjusting the knob values. Finally, we implemented the FET-MT method using GUROBI, and our experiments show that the running time of FET-MT is reduced by approximately a factor of 7 with respect to Maximum Nash Welfare (MNW) and discrete MNW and that FET-MT can still maintain good running efficiency as the number of users increases. The experimental results also show that FET-MT can obtain nearly twice the social welfare of MNW and DRF-MT, and the utilization of meta-types in the system is close to 100%.
{"title":"Fairness-Efficiency Tradeoff Allocation with Meta-Types in Cloud Computing","authors":"Feng-Qin Zhang, Xingxi Li, Weidong Li, Xuejie Zhang","doi":"10.1109/ICCC56324.2022.10065880","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10065880","url":null,"abstract":"We study the problem of multiple resource allocation in cloud computing systems. Existing fairness-efficiency scheduling procedures can relax fairness constraints by using a knob to improve efficiency. However, these approaches do not take into account users with special needs, i.e., the same resource (meta-type, e.g., CPU) contains different types (e.g., Intel's CPU, AMD's CPU) and the user can only use a specific type of resources (e.g., Intel's CPU). We propose a new allocation mechanism called Fairness-Efficiency Tradeoff Allocation with Meta-Types (FET-MT), which introduces the concept of meta-types. FET-MT not only meets specific requirements proposed by users but also allows users to flexibly balance fairness and efficiency by adjusting the knob values. Finally, we implemented the FET-MT method using GUROBI, and our experiments show that the running time of FET-MT is reduced by approximately a factor of 7 with respect to Maximum Nash Welfare (MNW) and discrete MNW and that FET-MT can still maintain good running efficiency as the number of users increases. The experimental results also show that FET-MT can obtain nearly twice the social welfare of MNW and DRF-MT, and the utilization of meta-types in the system is close to 100%.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131246200","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The large number and wide variety of hazardous entities is contradicted with the limited law enforcement strength of safety production, resulting in duplicate or missed inspections. In order to realize the key entity recommendation for safety production inspection, we introduce recommendation algorithms into this field. Data sparsity and cold start problems are inevitable in traditional recommendations, while knowledge graphs can be added as side information to solve the problems. Due to the strong sparsity of safety inspection data and the severe overfitting of existing models, we adaptively improve the multi-task learning algorithm by dividing the model into high layers and low layers and designing the structures respectively. A recommendation model based on multi-task learning and convolutional structures (CMKR) is proposed in this paper to provide better hazardous entity recommendations for safety production inspection. To solve the serious problem of over-fitting of the original multi-task learning algorithm, the convolutional neural network with the characteristics of sparse connection and weight sharing displaces a fully-connected multi-layer perceptron (MLP). ConvKB, an embedding model using CNN for the knowledge graph completion task is used at the high layers to improve the generalization ability of the model. In click-through rate prediction, ACC reaches 0.7061 and AUC reaches 0.7112 on hazardous entity recommendations of key sites. Compared with previous algorithms, the proposed method effectively controls the overfitting problem and improves the overall performance.
{"title":"Hazardous Entity Recommendation for Safety Production Inspection Based on Multi-task Learning","authors":"Xinyi Wang, Xinbo Ai, Yaniun Guo, Zhanghui Chen, Yichi Zhang","doi":"10.1109/ICCC56324.2022.10065664","DOIUrl":"https://doi.org/10.1109/ICCC56324.2022.10065664","url":null,"abstract":"The large number and wide variety of hazardous entities is contradicted with the limited law enforcement strength of safety production, resulting in duplicate or missed inspections. In order to realize the key entity recommendation for safety production inspection, we introduce recommendation algorithms into this field. Data sparsity and cold start problems are inevitable in traditional recommendations, while knowledge graphs can be added as side information to solve the problems. Due to the strong sparsity of safety inspection data and the severe overfitting of existing models, we adaptively improve the multi-task learning algorithm by dividing the model into high layers and low layers and designing the structures respectively. A recommendation model based on multi-task learning and convolutional structures (CMKR) is proposed in this paper to provide better hazardous entity recommendations for safety production inspection. To solve the serious problem of over-fitting of the original multi-task learning algorithm, the convolutional neural network with the characteristics of sparse connection and weight sharing displaces a fully-connected multi-layer perceptron (MLP). ConvKB, an embedding model using CNN for the knowledge graph completion task is used at the high layers to improve the generalization ability of the model. In click-through rate prediction, ACC reaches 0.7061 and AUC reaches 0.7112 on hazardous entity recommendations of key sites. Compared with previous algorithms, the proposed method effectively controls the overfitting problem and improves the overall performance.","PeriodicalId":263098,"journal":{"name":"2022 IEEE 8th International Conference on Computer and Communications (ICCC)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129030990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：