Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152591
Dapeng Wang, H. Qiu, L. Gao
Direct Monte Carlo Simulation for reliability estimation of rare failure event is challenged by the complicated performance function evaluations and large candidate sample pool. To address these challenges, a subdomain uncertainty- guided Kriging method with subset simulation is proposed. With a concise uncertainty assessment function, efficient subdomain uncertainty-guided sampling strategy is first developed to refine the Kriging model that is used to replace real performance function approximately. Moreover, the number of candidate samples required by subset simulation is also significantly reduced. By sequentially exploiting within the candidate sample pools generated in the first intermediate failure event and other intermediate failure events, an accurate Kriging model can be constructed subsequently. The ingenious method of coupling Kriging and subset simulation can greatly improve the efficiency of reliability estimation. Finally, three classical examples are investigated as benchmark to explore the performance of the proposed method. The comparison results demonstrate the good capability and applicability of the proposed method.
{"title":"A Subdomain Uncertainty-Guided Kriging Method with Subset Simulation for Reliability Estimation","authors":"Dapeng Wang, H. Qiu, L. Gao","doi":"10.1109/CSCWD57460.2023.10152591","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152591","url":null,"abstract":"Direct Monte Carlo Simulation for reliability estimation of rare failure event is challenged by the complicated performance function evaluations and large candidate sample pool. To address these challenges, a subdomain uncertainty- guided Kriging method with subset simulation is proposed. With a concise uncertainty assessment function, efficient subdomain uncertainty-guided sampling strategy is first developed to refine the Kriging model that is used to replace real performance function approximately. Moreover, the number of candidate samples required by subset simulation is also significantly reduced. By sequentially exploiting within the candidate sample pools generated in the first intermediate failure event and other intermediate failure events, an accurate Kriging model can be constructed subsequently. The ingenious method of coupling Kriging and subset simulation can greatly improve the efficiency of reliability estimation. Finally, three classical examples are investigated as benchmark to explore the performance of the proposed method. The comparison results demonstrate the good capability and applicability of the proposed method.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"20 1","pages":"1526-1531"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85501965","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152802
Yanhao Feng, Xiaojie Lin, Zitao Yu
The Internet of Things (IoT) has gained lots of attention in the past decade and has shown huge potential in the integration of power systems to build a smart grid architecture and to achieve a paradigm transformation. This paper first proposed a typical architecture of IoT in power systems. To reveal the research status, literature development, cooperative relationship, hotspots of techniques, and the research trend, a bibliometric analysis from 1196 papers in the WoS database was done. Finally, several research trends were concluded based on the bibliometric analysis. This paper, for the first time, provided a systematic review of the hotspots in the field of IoT in power systems for more than a decade. Additionally, this paper provided a clear understanding of the history and outlook of the development of the IoT in power systems.
{"title":"Internet of Things in Power Systems: A Bibliometric Analysis","authors":"Yanhao Feng, Xiaojie Lin, Zitao Yu","doi":"10.1109/CSCWD57460.2023.10152802","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152802","url":null,"abstract":"The Internet of Things (IoT) has gained lots of attention in the past decade and has shown huge potential in the integration of power systems to build a smart grid architecture and to achieve a paradigm transformation. This paper first proposed a typical architecture of IoT in power systems. To reveal the research status, literature development, cooperative relationship, hotspots of techniques, and the research trend, a bibliometric analysis from 1196 papers in the WoS database was done. Finally, several research trends were concluded based on the bibliometric analysis. This paper, for the first time, provided a systematic review of the hotspots in the field of IoT in power systems for more than a decade. Additionally, this paper provided a clear understanding of the history and outlook of the development of the IoT in power systems.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"26 1","pages":"1427-1432"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84605425","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152613
Xijun Zhang, Jin Su, Hong Zhang, Xianli Zhang, Xuanbing Chen, Yong Cui
Aiming at the shortcomings of the traditional Density-Based Spatial Clustering of Applications with Noise -DBSCAN algorithm such as insignificant clustering effect and the choice of parameter combinations. This paper proposes an AD-DBSCAN algorithm with adaptive parameters, which makes the algorithm more difficult in the selection of the parameters. By establishing a DBSCAN algorithm model to adapt to finding the optimal distance threshold and the minimum number of neighbor points, the clustering is more accurate, and the noise point identified in the data is more accurate. Through the observation of the calculation model of the Calinski-Harabasz index, the evaluation index of the clustering algorithm, the selection of the optimal best distance threshold and the minimum number of neighborhood points, the accuracy of noise point recognition is improved by 5 times in the clustering algorithm, and the Calinski-Harabasz index improved by about 39.84%. The applicability of the algorithm in clustering the locations of urban road traffic accidents is verified.
{"title":"Traffic accident location study based on AD-DBSCAN Algorithm with Adaptive Parameters","authors":"Xijun Zhang, Jin Su, Hong Zhang, Xianli Zhang, Xuanbing Chen, Yong Cui","doi":"10.1109/CSCWD57460.2023.10152613","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152613","url":null,"abstract":"Aiming at the shortcomings of the traditional Density-Based Spatial Clustering of Applications with Noise -DBSCAN algorithm such as insignificant clustering effect and the choice of parameter combinations. This paper proposes an AD-DBSCAN algorithm with adaptive parameters, which makes the algorithm more difficult in the selection of the parameters. By establishing a DBSCAN algorithm model to adapt to finding the optimal distance threshold and the minimum number of neighbor points, the clustering is more accurate, and the noise point identified in the data is more accurate. Through the observation of the calculation model of the Calinski-Harabasz index, the evaluation index of the clustering algorithm, the selection of the optimal best distance threshold and the minimum number of neighborhood points, the accuracy of noise point recognition is improved by 5 times in the clustering algorithm, and the Calinski-Harabasz index improved by about 39.84%. The applicability of the algorithm in clustering the locations of urban road traffic accidents is verified.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"32 1","pages":"1160-1165"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84688528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152700
Shuang Qiao, Chenhong Cao, Haoquan Zhou, Wei Gong
With the rapid advancement of smartphones and other mobile devices, an ever-increasing desire for wireless indoor localization has emerged. This technology is capable of determining the position of a user or device in an indoor setting and facilitating an array of captivating applications. Due to the low cost and wide availability of WiFi, WiFi-based indoor localization has received considerable attention and has become a prominent research focus in recent times. We have distilled that an ideal WiFi-based indoor localization system is anticipated to meet three criteria: high-accuracy, pervasiveness, and easy-deployment. Nevertheless, it is not a trivial task to satisfy all three criteria simultaneously. This document scrutinizes the key issues, basic models, and current methods for WiFi indoor localization with the objective of highlighting the underlying principles and challenges. Finally, this manuscript pinpoints the prospective research paths for WiFi indoor localization.
{"title":"The trip to WiFi indoor localization across a decade — A systematic review","authors":"Shuang Qiao, Chenhong Cao, Haoquan Zhou, Wei Gong","doi":"10.1109/CSCWD57460.2023.10152700","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152700","url":null,"abstract":"With the rapid advancement of smartphones and other mobile devices, an ever-increasing desire for wireless indoor localization has emerged. This technology is capable of determining the position of a user or device in an indoor setting and facilitating an array of captivating applications. Due to the low cost and wide availability of WiFi, WiFi-based indoor localization has received considerable attention and has become a prominent research focus in recent times. We have distilled that an ideal WiFi-based indoor localization system is anticipated to meet three criteria: high-accuracy, pervasiveness, and easy-deployment. Nevertheless, it is not a trivial task to satisfy all three criteria simultaneously. This document scrutinizes the key issues, basic models, and current methods for WiFi indoor localization with the objective of highlighting the underlying principles and challenges. Finally, this manuscript pinpoints the prospective research paths for WiFi indoor localization.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"R-27 1","pages":"642-647"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84747513","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152667
Rui Huang, Huan Lu, Yan Xing, Wei Fan
Deep learning technology has promoted the performance of defocus blur detection. However, blur detectors suffer from background clutter, scale ambiguity and blurred boundaries of the defocus blur regions. To conquer these issues, previous methods propose to use multi-scale image patches or images for blur detection, which costs much computation time. In this paper, we propose a deep neural network that takes a single-scale image as input to generate robust defocus blur detection. Specifically, we first extract multi-scale convolutional features by a feature extraction network. And then we resize the convolutional features of each layer by a fixed ratio to approximate convolutional features that extracted from a resized image with the same ratio. By approximation, it not only generates features extracted from a scaled image but also reduces the computation of feature extraction from multi-scale images. We concatenate the features extracted from the original image with the approximated features at the corresponding layers by convolutional layers to increase the blur distinguish ability. We gradually fuse the convolutional features from top-to-bottom by Conv-LSTMs to refine the blur predictions. We compare our method with nine state-of-the-art defocus blur detectors on two defocus blur detection benchmark datasets. Experiment results demonstrate the effectiveness of our proposed defocus blur detector.
{"title":"Multi-scale Convolutional Feature Approximation for Defocus Blur Detection","authors":"Rui Huang, Huan Lu, Yan Xing, Wei Fan","doi":"10.1109/CSCWD57460.2023.10152667","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152667","url":null,"abstract":"Deep learning technology has promoted the performance of defocus blur detection. However, blur detectors suffer from background clutter, scale ambiguity and blurred boundaries of the defocus blur regions. To conquer these issues, previous methods propose to use multi-scale image patches or images for blur detection, which costs much computation time. In this paper, we propose a deep neural network that takes a single-scale image as input to generate robust defocus blur detection. Specifically, we first extract multi-scale convolutional features by a feature extraction network. And then we resize the convolutional features of each layer by a fixed ratio to approximate convolutional features that extracted from a resized image with the same ratio. By approximation, it not only generates features extracted from a scaled image but also reduces the computation of feature extraction from multi-scale images. We concatenate the features extracted from the original image with the approximated features at the corresponding layers by convolutional layers to increase the blur distinguish ability. We gradually fuse the convolutional features from top-to-bottom by Conv-LSTMs to refine the blur predictions. We compare our method with nine state-of-the-art defocus blur detectors on two defocus blur detection benchmark datasets. Experiment results demonstrate the effectiveness of our proposed defocus blur detector.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"156 1","pages":"1172-1177"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79876407","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152575
Hanqing Li, Niannian Chen
Social relations, as the basic relationships in our daily life, are a phenomenon unique to human society that shows how people interact in society. Social relations understanding is to infer the existing social relationships between individuals in a given scenario, which is crucial for us to analyze social behavior. Existing research methods are usually limited to extracting features of characters and related entities, which limits the scope of attention and may miss important clues such as interactions between characters. In this paper, we propose a global attention mechanism that adaptively grasps scenes, objects, and human interactions for reasoning about social relationships. We propose an end-to-end global attention network, which consists of three modules, namely, a convolutional attention module, a graph inference module, and an attentional inference module. The visual and location information is first extracted by the convolutional attention module as the feature information of the person pairs, then it is made to process the relationships between character nodes on the graph inference network, and finally, the attention is fully utilized to classify the social relationships. Extensive experiments on the PISC and PIPA datasets show that our proposed method outperforms the state-of-the-art methods in terms of accuracy.
{"title":"Understanding Social Relations with Graph-Based and Global Attention","authors":"Hanqing Li, Niannian Chen","doi":"10.1109/CSCWD57460.2023.10152575","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152575","url":null,"abstract":"Social relations, as the basic relationships in our daily life, are a phenomenon unique to human society that shows how people interact in society. Social relations understanding is to infer the existing social relationships between individuals in a given scenario, which is crucial for us to analyze social behavior. Existing research methods are usually limited to extracting features of characters and related entities, which limits the scope of attention and may miss important clues such as interactions between characters. In this paper, we propose a global attention mechanism that adaptively grasps scenes, objects, and human interactions for reasoning about social relationships. We propose an end-to-end global attention network, which consists of three modules, namely, a convolutional attention module, a graph inference module, and an attentional inference module. The visual and location information is first extracted by the convolutional attention module as the feature information of the person pairs, then it is made to process the relationships between character nodes on the graph inference network, and finally, the attention is fully utilized to classify the social relationships. Extensive experiments on the PISC and PIPA datasets show that our proposed method outperforms the state-of-the-art methods in terms of accuracy.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"1 1","pages":"41-46"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80530826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152736
Yibo Xie, Gaopeng Gou, G. Xiong, Zhuguo Li, Mingxin Cui
Snowflake is a special proxy system against IP-based network blocking. As its IP addresses refresh frequently, faster than IP blacklist’s update, users can exploit it to access blocked websites. To block snowflake, existing methods focus on detecting snowflake proxies. But they are susceptible to various factors, for example, proxy’s location and version. In the paper, we propose a new manner to block snowflake. We observe that to adapt fast IP changes, users need to request latest proxies from proxy database before using snowflake. Thus, adversaries can block snowflake by detecting proxy request instead of proxy itself. To verify our method, we analyse covertness of snowflake proxy requests, that has been protected by imitating normal web requests. After comparing with typical web requests, we find the imitation is vulnerable in packet size, direction, time and network speed, such as, the latency time is higher than normal obviously. Using the four vulnerabilities, we train machine learning algorithm to detect snowflake proxy requests in reality. Experimental results demonstrate that proxy request can be detected accurately across different versions at the beginning of connection. In conclusion, our work paves a new way to block snowflake.
{"title":"Covertness Analysis of Snowflake Proxy Request","authors":"Yibo Xie, Gaopeng Gou, G. Xiong, Zhuguo Li, Mingxin Cui","doi":"10.1109/CSCWD57460.2023.10152736","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152736","url":null,"abstract":"Snowflake is a special proxy system against IP-based network blocking. As its IP addresses refresh frequently, faster than IP blacklist’s update, users can exploit it to access blocked websites. To block snowflake, existing methods focus on detecting snowflake proxies. But they are susceptible to various factors, for example, proxy’s location and version. In the paper, we propose a new manner to block snowflake. We observe that to adapt fast IP changes, users need to request latest proxies from proxy database before using snowflake. Thus, adversaries can block snowflake by detecting proxy request instead of proxy itself. To verify our method, we analyse covertness of snowflake proxy requests, that has been protected by imitating normal web requests. After comparing with typical web requests, we find the imitation is vulnerable in packet size, direction, time and network speed, such as, the latency time is higher than normal obviously. Using the four vulnerabilities, we train machine learning algorithm to detect snowflake proxy requests in reality. Experimental results demonstrate that proxy request can be detected accurately across different versions at the beginning of connection. In conclusion, our work paves a new way to block snowflake.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"21 1","pages":"1802-1807"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73063725","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152771
Renfang Wang, Zijian Yang, Hong Qiu, X. Liu, Dun Wu
Change detection is an important branch in remote sensing image processing. Deep learning has been widely used in this field. In particular, a wide variety of attention mechanisms have made great achievements. However, some models have become increasingly complex and large, often unfeasible for edge applications. This poses a major obstacle to industrial applications. In this paper, to solve the above challenges, we propose a Lightweight network structure to improve results while taking into account efficiency. Specifically, first, the shallow features are extracted by using the spatial exchange and change exchange of the down-sampling bi-temporal channel of the three-layer EfficientNet backbone network, and then the shallow features are used for low-dimensional skip-connection. After that, a hybrid dual-temporal data module is designed to mix the dual-temporal phase into a single image, then the high-dimensional low-pixel image is restored through the up-sampling. Finally the final change map is generated through the pixel-level classifier. Our method was evaluated on public datasets by evaluation indicators such as OA, IoU, F1, Recall, Precision.
{"title":"Spatial and Channel Exchange based on EfficientNet for Detecting Changes of Remote Sensing Images","authors":"Renfang Wang, Zijian Yang, Hong Qiu, X. Liu, Dun Wu","doi":"10.1109/CSCWD57460.2023.10152771","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152771","url":null,"abstract":"Change detection is an important branch in remote sensing image processing. Deep learning has been widely used in this field. In particular, a wide variety of attention mechanisms have made great achievements. However, some models have become increasingly complex and large, often unfeasible for edge applications. This poses a major obstacle to industrial applications. In this paper, to solve the above challenges, we propose a Lightweight network structure to improve results while taking into account efficiency. Specifically, first, the shallow features are extracted by using the spatial exchange and change exchange of the down-sampling bi-temporal channel of the three-layer EfficientNet backbone network, and then the shallow features are used for low-dimensional skip-connection. After that, a hybrid dual-temporal data module is designed to mix the dual-temporal phase into a single image, then the high-dimensional low-pixel image is restored through the up-sampling. Finally the final change map is generated through the pixel-level classifier. Our method was evaluated on public datasets by evaluation indicators such as OA, IoU, F1, Recall, Precision.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"8 1","pages":"1595-1600"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"73225305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10151999
Hugo Guercio, Victor Ströele, José Maria N. David, R. Braga
Development communities and the software industry increasingly adopt the Software Ecosystems approach (SECO). This approach can provide advantages but add additional complexity to resource management, affecting the software supply network. Observing SECOs, we can see through three dimensions: business, technical, and social. The social dimension focuses on stakeholders and how they interact with other dimensions. This paper presents a process for analyzing the social dimension of Software Ecosystems, supported by Complex Networks metrics, which allow the presentation of existing SECO relationships’ through visualizations and the use of complex networks. A preliminary evaluation with real data was carried out. The results point to the solution’s viability.
{"title":"A Process to Analyze Software Ecosystem Social Dimension Through a Collaboration Perspective","authors":"Hugo Guercio, Victor Ströele, José Maria N. David, R. Braga","doi":"10.1109/CSCWD57460.2023.10151999","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10151999","url":null,"abstract":"Development communities and the software industry increasingly adopt the Software Ecosystems approach (SECO). This approach can provide advantages but add additional complexity to resource management, affecting the software supply network. Observing SECOs, we can see through three dimensions: business, technical, and social. The social dimension focuses on stakeholders and how they interact with other dimensions. This paper presents a process for analyzing the social dimension of Software Ecosystems, supported by Complex Networks metrics, which allow the presentation of existing SECO relationships’ through visualizations and the use of complex networks. A preliminary evaluation with real data was carried out. The results point to the solution’s viability.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"50 1","pages":"1202-1207"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75813146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-05-24DOI: 10.1109/CSCWD57460.2023.10152760
Jing Tian, Juan Chen, Ningjiang Chen, Lin Bai, Suqun Huang
The pre-training language model BERT has brought significant performance improvements to a series of natural language processing tasks, but due to the large scale of the model, it is difficult to be applied in many practical application scenarios. With the continuous development of edge computing, deploying the models on resource-constrained edge devices has become a trend. Considering the distributed edge environment, how to take into account issues such as data distribution differences, labeling costs, and privacy while the model is shrinking is a critical task. The paper proposes a new BERT distillation method with source-free unsupervised domain adaptation. By combining source-free unsupervised domain adaptation and knowledge distillation for optimization and improvement, the performance of the BERT model is improved in the case of cross-domain data. Compared with other methods, our method can improve the average prediction accuracy by up to around 4% through the experimental evaluation of the cross-domain sentiment analysis task.
{"title":"Knowledge Distillation with Source-free Unsupervised Domain Adaptation for BERT Model Compression","authors":"Jing Tian, Juan Chen, Ningjiang Chen, Lin Bai, Suqun Huang","doi":"10.1109/CSCWD57460.2023.10152760","DOIUrl":"https://doi.org/10.1109/CSCWD57460.2023.10152760","url":null,"abstract":"The pre-training language model BERT has brought significant performance improvements to a series of natural language processing tasks, but due to the large scale of the model, it is difficult to be applied in many practical application scenarios. With the continuous development of edge computing, deploying the models on resource-constrained edge devices has become a trend. Considering the distributed edge environment, how to take into account issues such as data distribution differences, labeling costs, and privacy while the model is shrinking is a critical task. The paper proposes a new BERT distillation method with source-free unsupervised domain adaptation. By combining source-free unsupervised domain adaptation and knowledge distillation for optimization and improvement, the performance of the BERT model is improved in the case of cross-domain data. Compared with other methods, our method can improve the average prediction accuracy by up to around 4% through the experimental evaluation of the cross-domain sentiment analysis task.","PeriodicalId":51008,"journal":{"name":"Computer Supported Cooperative Work-The Journal of Collaborative Computing","volume":"1 1","pages":"1766-1771"},"PeriodicalIF":2.4,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75299806","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","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：