Pub Date : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00137
Yuchen Jiang, Chang Ji
Predictive maintenance (PdM) has entered into a new era adopting artificial intelligence and Internet-of-Things (IoT) technologies. It is necessary for a manufacturing company to collaborate with other clients using IoT-captured production data. However, training models in a cross-silo manner is still challenging when considering data privacy. In order to tackle these challenges, a personalized cross-silo federated learning mechanism named federated global partners searching (FedGPS) is proposed. Firstly, model parameters for the participating clients are encrypted and uploaded to the central server as input. Next, FedGPS automatically determines the collaboration degrees between clients based on data distribution. After that, personalized model updates are sent back to the clients. Finally, each client conducts local updating after data decryption. The effectiveness of the FedGPS is verified in real-world cases and our method achieves 92.35% Accuracy, 98.55% Precision, 92.90% Recall, and 95.27% F1-Score comparing with other existing models from the literature.
预测性维护(PdM)已经进入了人工智能和物联网(IoT)技术的新时代。制造公司有必要使用物联网捕获的生产数据与其他客户进行协作。然而，在考虑数据隐私时，以跨竖井的方式训练模型仍然具有挑战性。为了解决这些问题，提出了一种个性化的跨竖井联邦学习机制——联邦全局伙伴搜索(federal global partners searching, FedGPS)。首先，对参与客户端的模型参数进行加密，并将其作为输入上传到中央服务器。其次，FedGPS根据数据分布自动确定客户端之间的协作程度。之后，个性化的模型更新被发送回客户端。最后，各客户端在数据解密后进行本地更新。在实际案例中验证了FedGPS的有效性，与文献中已有的模型相比，我们的方法达到了92.35%的准确率、98.55%的精密度、92.90%的召回率和95.27%的F1-Score。
{"title":"FedGPS: Personalized Cross-Silo Federated Learning for Internet of Things-enabled Predictive Maintenance","authors":"Yuchen Jiang, Chang Ji","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00137","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00137","url":null,"abstract":"Predictive maintenance (PdM) has entered into a new era adopting artificial intelligence and Internet-of-Things (IoT) technologies. It is necessary for a manufacturing company to collaborate with other clients using IoT-captured production data. However, training models in a cross-silo manner is still challenging when considering data privacy. In order to tackle these challenges, a personalized cross-silo federated learning mechanism named federated global partners searching (FedGPS) is proposed. Firstly, model parameters for the participating clients are encrypted and uploaded to the central server as input. Next, FedGPS automatically determines the collaboration degrees between clients based on data distribution. After that, personalized model updates are sent back to the clients. Finally, each client conducts local updating after data decryption. The effectiveness of the FedGPS is verified in real-world cases and our method achieves 92.35% Accuracy, 98.55% Precision, 92.90% Recall, and 95.27% F1-Score comparing with other existing models from the literature.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"1 1","pages":"912-920"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90350608","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 : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00112
Wuliang Huang, Xinlong Jiang, Chenlong Gao, Teng Zhang, Yunbing Xing, Yiqiang Chen, Yi Zheng, Jie Li
Attention deficit hyperactivity disorder (ADHD) is a common childhood mental disorder that encompasses three subtypes. Classifying each subtype has practical significance. However, the gold standard for subtype diagnosis depends on face-to-face consultation with psychiatrists, which is limited by medical resources. This paper proposes a graph-based multimodal fusion approach to classify each subtype objectively, alleviating the pressure on psychiatrists. The proposed method leverages heterogeneous signals, including motion and speech, which are significant indicators of ADHD. We construct a personal graph where each child is a vertex, and the similarity of their personal information measures edges. Since the associations between subjects modeled by the personal graph provide rich prior knowledge, we regard the problem of subtype classification as predicting the labels of vertices on a graph. A novel graph neural network model is proposed to enable information passing between children, fusing motion and speech features under the guidance of the personal graph. We design a reading scenario and collect a multimodal dataset containing 56 children with ADHD and 50 typically developing children. Results of ADHD subtype classification demonstrate the practical value of the proposed approach. We also perform ablation studies to verify the validity of the proposed method.
{"title":"A Graph-Based Information Fusion Approach for ADHD Subtype Classification","authors":"Wuliang Huang, Xinlong Jiang, Chenlong Gao, Teng Zhang, Yunbing Xing, Yiqiang Chen, Yi Zheng, Jie Li","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00112","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00112","url":null,"abstract":"Attention deficit hyperactivity disorder (ADHD) is a common childhood mental disorder that encompasses three subtypes. Classifying each subtype has practical significance. However, the gold standard for subtype diagnosis depends on face-to-face consultation with psychiatrists, which is limited by medical resources. This paper proposes a graph-based multimodal fusion approach to classify each subtype objectively, alleviating the pressure on psychiatrists. The proposed method leverages heterogeneous signals, including motion and speech, which are significant indicators of ADHD. We construct a personal graph where each child is a vertex, and the similarity of their personal information measures edges. Since the associations between subjects modeled by the personal graph provide rich prior knowledge, we regard the problem of subtype classification as predicting the labels of vertices on a graph. A novel graph neural network model is proposed to enable information passing between children, fusing motion and speech features under the guidance of the personal graph. We design a reading scenario and collect a multimodal dataset containing 56 children with ADHD and 50 typically developing children. Results of ADHD subtype classification demonstrate the practical value of the proposed approach. We also perform ablation studies to verify the validity of the proposed method.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"14 1","pages":"714-723"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88926073","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 : 2022-12-01DOI: 10.1109/smartworld-uic-atc-scalcom-digitaltwin-pricomp-metaverse56740.2022.00162
Yong Song, Zhiwei Yan, Yukun Qin, Xiaozhou Ye, Ye Ouyang
Named Entity Recognition (NER) is an important down-streaming task in natural language processing. Span-based methods are applicable to both flat and nested entities. However, they lack explicit boundary supervision. To address this issue, we propose a multi-task and self-distilled model which combines biaffine span classification and entity boundary detection tasks. Firstly, the boundary detection and biaffine span classification models are jointly trained under a multi-task learning framework to address the problem of lacking supervision of boundaries. Then, self-distillation technique is applied on the model to reassign entity probabilities from annotated spans to surrounding spans and more entity types, further improving the accuracy of NER by soft labels that contain richer knowledge. Experiments were based on a high-density entity text dataset of the commodity titles from an e-commerce company. Finally, the experimental results show that our model exhibited a better F1 score than the existing common models.
{"title":"Self-distilled Named Entity Recognition Based on Boundary Detection and Biaffine Attention","authors":"Yong Song, Zhiwei Yan, Yukun Qin, Xiaozhou Ye, Ye Ouyang","doi":"10.1109/smartworld-uic-atc-scalcom-digitaltwin-pricomp-metaverse56740.2022.00162","DOIUrl":"https://doi.org/10.1109/smartworld-uic-atc-scalcom-digitaltwin-pricomp-metaverse56740.2022.00162","url":null,"abstract":"Named Entity Recognition (NER) is an important down-streaming task in natural language processing. Span-based methods are applicable to both flat and nested entities. However, they lack explicit boundary supervision. To address this issue, we propose a multi-task and self-distilled model which combines biaffine span classification and entity boundary detection tasks. Firstly, the boundary detection and biaffine span classification models are jointly trained under a multi-task learning framework to address the problem of lacking supervision of boundaries. Then, self-distillation technique is applied on the model to reassign entity probabilities from annotated spans to surrounding spans and more entity types, further improving the accuracy of NER by soft labels that contain richer knowledge. Experiments were based on a high-density entity text dataset of the commodity titles from an e-commerce company. Finally, the experimental results show that our model exhibited a better F1 score than the existing common models.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"1 1","pages":"1132-1137"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90184901","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 : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00197
Lu Ren, Jianwei Niu, Zhenchao Ouyang, Zhibin Zhang, Siyi Zheng
Dynamic scene understanding based on LiDAR point clouds is one of the critical perception tasks for self-driving vehicles. Among these tasks, point cloud semantic segmentation is highly challenging. Some existing work ignores the loss of crucial information caused by sampling and projecting. Others use modules with high computational complexity because of the pursuit of precision, challenging to deploy in the vehicle platform with limited computing power. This paper proposes Fusedown/Fuse-up modules for efficient down-sampling/up-sampling feature extraction. The modules combine the transformer in vision integrating the global information of the feature map with the CNN extracting local feature information. Based on these two modules, we built the transformer and CNN fusion network called TCFNet for point cloud semantic segmentation. Experiments on the SemanticKITTI show that our suitable combination of transformer and CNN is necessary for semantic segmentation accuracy, and the mIoU of our model can reach 82.7% at 10 FPS. The code can be accessed at https://github.com/donkeyofking/TCFNet.git.
{"title":"TCFNet: Transformer and CNN Fusion Model for LiDAR Point Cloud Semantic Segmentation","authors":"Lu Ren, Jianwei Niu, Zhenchao Ouyang, Zhibin Zhang, Siyi Zheng","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00197","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00197","url":null,"abstract":"Dynamic scene understanding based on LiDAR point clouds is one of the critical perception tasks for self-driving vehicles. Among these tasks, point cloud semantic segmentation is highly challenging. Some existing work ignores the loss of crucial information caused by sampling and projecting. Others use modules with high computational complexity because of the pursuit of precision, challenging to deploy in the vehicle platform with limited computing power. This paper proposes Fusedown/Fuse-up modules for efficient down-sampling/up-sampling feature extraction. The modules combine the transformer in vision integrating the global information of the feature map with the CNN extracting local feature information. Based on these two modules, we built the transformer and CNN fusion network called TCFNet for point cloud semantic segmentation. Experiments on the SemanticKITTI show that our suitable combination of transformer and CNN is necessary for semantic segmentation accuracy, and the mIoU of our model can reach 82.7% at 10 FPS. The code can be accessed at https://github.com/donkeyofking/TCFNet.git.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"63 1","pages":"1366-1372"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86865277","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 : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00195
L. Weigang, L. Martins, Nikson Ferreira, Christian Miranda, Lucas S. Althoff, Walner Pessoa, Mylène C. Q. Farias, Ricardo Jacobi, Mauricio Rincon
Few-shot learning is an important mechanism to minimize the need for the labeling of large amounts of data and taking advantage of transfer learning. To identify image/text input with duality property, this research proposes a “Heuristic once learning (HOL)” mechanism to investigate multi-modal input processing similar to human-like behavior. First, we create an image/text data set of big Latin letters composed of small letters and another data set composed of Arabic, Chinese and Roman numerals. Secondly, we use Convolutional Neural Networks (CNN) for pre-training the dataset of letters to get structural features. Thirdly, using the acquired knowledge, a Self-organizing Map (SOM) and Contrastive Language-Image Pretraining (CLIP) are tested separately using zero-shot learning. Siamese Networks and Vision Transformer (ViT) are also tested using one-shot learning by knowledge transfer to identify the features of unknown characters. The research results show the potential and challenges to realize HOL and make a useful attempt for the development of general agents.
{"title":"Heuristic Once Learning for Image & Text Duality Information Processing","authors":"L. Weigang, L. Martins, Nikson Ferreira, Christian Miranda, Lucas S. Althoff, Walner Pessoa, Mylène C. Q. Farias, Ricardo Jacobi, Mauricio Rincon","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00195","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00195","url":null,"abstract":"Few-shot learning is an important mechanism to minimize the need for the labeling of large amounts of data and taking advantage of transfer learning. To identify image/text input with duality property, this research proposes a “Heuristic once learning (HOL)” mechanism to investigate multi-modal input processing similar to human-like behavior. First, we create an image/text data set of big Latin letters composed of small letters and another data set composed of Arabic, Chinese and Roman numerals. Secondly, we use Convolutional Neural Networks (CNN) for pre-training the dataset of letters to get structural features. Thirdly, using the acquired knowledge, a Self-organizing Map (SOM) and Contrastive Language-Image Pretraining (CLIP) are tested separately using zero-shot learning. Siamese Networks and Vision Transformer (ViT) are also tested using one-shot learning by knowledge transfer to identify the features of unknown characters. The research results show the potential and challenges to realize HOL and make a useful attempt for the development of general agents.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"20 1","pages":"1353-1359"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72708078","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 : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00221
Huaicheng Fang, Fuqing Zhu, Jizhong Han, Songlin Hu
A large amount of hateful speech exist on the Internet in the form of text and images uploaded by social media users. Recently, multimodal hateful speech detection task has attracted more and more researchers to invest, producing some representative work for perceiving the negative samples. For this special multimodal task, the ability of multimodal semantic information understanding is particularly crucial. However, the existing models have insufficient understanding ability of image modality semantic compared with the text modality, due to the appearance complexity of each image. Therefore, this paper utilizes the text modality which is well understood by the model to improve understanding ability of image modality semantic. Specifically, this paper proposes an image caption supervision (ICS) auxiliary method for multimodal hateful speech detection, where the image caption is designed to supervise the feature learning of images for further understanding the semantic information. On the Facebook Hateful Memes dataset, the proposed ICS method outperforms some state-of-the-art unimodal and multimodal baselines, demonstrating the effectiveness of ICS.
{"title":"Multimodal Hateful Memes Detection via Image Caption Supervision","authors":"Huaicheng Fang, Fuqing Zhu, Jizhong Han, Songlin Hu","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00221","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00221","url":null,"abstract":"A large amount of hateful speech exist on the Internet in the form of text and images uploaded by social media users. Recently, multimodal hateful speech detection task has attracted more and more researchers to invest, producing some representative work for perceiving the negative samples. For this special multimodal task, the ability of multimodal semantic information understanding is particularly crucial. However, the existing models have insufficient understanding ability of image modality semantic compared with the text modality, due to the appearance complexity of each image. Therefore, this paper utilizes the text modality which is well understood by the model to improve understanding ability of image modality semantic. Specifically, this paper proposes an image caption supervision (ICS) auxiliary method for multimodal hateful speech detection, where the image caption is designed to supervise the feature learning of images for further understanding the semantic information. On the Facebook Hateful Memes dataset, the proposed ICS method outperforms some state-of-the-art unimodal and multimodal baselines, demonstrating the effectiveness of ICS.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"2 2","pages":"1530-1537"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72471719","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 : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00086
Qinglong Peng, Bin Tang, Jinhuan Liu, Shuang Cui, Junwei Du, Yan Lu, Feng Jiang, Xu Yu
Recently, cross-domain collaborative filtering (CDCF) has been widely used to solve the data sparsity problem in recommendation systems. Therein, the dual-target cross-domain recommendation becomes a research hotspot, which aims to improve the recommendation performance of both target and source domains. Most existing approaches tend to use fixed weights or self-attention in a single representation space for the bi-directional inter-domain transfer of the user representation. However, a single representation space leads to limited representation capability, which makes the transfer of the user representation coarse-grained and inaccurate. In this paper, Multi-head Attention Based Dual Target Graph Collaborative Filtering Network (MA-DTGCF) is proposed. The core of the model is the bi-directional transfer graph convolution layer, consisting of a graph convolution layer and a bi-directional transfer layer based on a multi-head attention mechanism. The latter can achieve fine-grained and adaptive transfer of user features in multiple representation subspaces. It is worth noting that by stacking multiple bi-directional transfer graph convolutional layers, we can get high-order user and item features and achieve adaptive transfer of each order user features. Experimental results on three real datasets show that the proposed MA-DTGCF model significantly outperforms the state-of-the-art models in terms of HR and NDCG.
{"title":"A Multi-Head Attention Based Dual Target Graph Collaborative Filtering Network","authors":"Qinglong Peng, Bin Tang, Jinhuan Liu, Shuang Cui, Junwei Du, Yan Lu, Feng Jiang, Xu Yu","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00086","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00086","url":null,"abstract":"Recently, cross-domain collaborative filtering (CDCF) has been widely used to solve the data sparsity problem in recommendation systems. Therein, the dual-target cross-domain recommendation becomes a research hotspot, which aims to improve the recommendation performance of both target and source domains. Most existing approaches tend to use fixed weights or self-attention in a single representation space for the bi-directional inter-domain transfer of the user representation. However, a single representation space leads to limited representation capability, which makes the transfer of the user representation coarse-grained and inaccurate. In this paper, Multi-head Attention Based Dual Target Graph Collaborative Filtering Network (MA-DTGCF) is proposed. The core of the model is the bi-directional transfer graph convolution layer, consisting of a graph convolution layer and a bi-directional transfer layer based on a multi-head attention mechanism. The latter can achieve fine-grained and adaptive transfer of user features in multiple representation subspaces. It is worth noting that by stacking multiple bi-directional transfer graph convolutional layers, we can get high-order user and item features and achieve adaptive transfer of each order user features. Experimental results on three real datasets show that the proposed MA-DTGCF model significantly outperforms the state-of-the-art models in terms of HR and NDCG.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"107 1","pages":"476-483"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74476078","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}
Service recommendation is important in creating composite services, workflows, e-business solutions, etc. It often takes developers a long time to Figure out what the next service is. A lot of researchers have used collaborative filtering-based or content-based approaches to recommend services for developers. However, failing to model the co-occurrence relationships between services, current approaches cannot recommend the next services for service composition. This leads to a decrease in the accuracy of service composition recommendations. To tackle this problem, this paper proposes an Encoder-Decoder-based Recommender named EDeR, which transforms the service recommendation problem into a generation problem. First, we employ a self-supervised graph embedding method to fully learn the representation of each service according to both structural and descriptive information. Then, based on the co-occurrence relationships between services, we propose an encoder-decoder model to sequentially recommend services in a way that translates user requirements into a composite service. The results obtained from experiments conducted on a real-world dataset show that EDeR outperforms the state-of-the-art approaches significantly.
{"title":"What Is Next? A Generative Approach for Service Composition Recommendations","authors":"Guodong Fan, Shizhan Chen, Hongyue Wu, Ming Zhu, Xiao Xue, Zhiyong Feng","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00078","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00078","url":null,"abstract":"Service recommendation is important in creating composite services, workflows, e-business solutions, etc. It often takes developers a long time to Figure out what the next service is. A lot of researchers have used collaborative filtering-based or content-based approaches to recommend services for developers. However, failing to model the co-occurrence relationships between services, current approaches cannot recommend the next services for service composition. This leads to a decrease in the accuracy of service composition recommendations. To tackle this problem, this paper proposes an Encoder-Decoder-based Recommender named EDeR, which transforms the service recommendation problem into a generation problem. First, we employ a self-supervised graph embedding method to fully learn the representation of each service according to both structural and descriptive information. Then, based on the co-occurrence relationships between services, we propose an encoder-decoder model to sequentially recommend services in a way that translates user requirements into a composite service. The results obtained from experiments conducted on a real-world dataset show that EDeR outperforms the state-of-the-art approaches significantly.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"20 1","pages":"409-416"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74614627","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 : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00265
Hao Li, P. Yue, Shangcheng Li, Fangqiang Yu, Chenxiao Zhang, Can Yang, Liangcun Jiang
Next Point-of-Interest (POI) recommendation has been applied by many Internet companies to enhance user travel experience. The state-of-the-art deep learning methods in next POI recommendation advocate the self-attention mechanism to model the user long-term check-in sequence. However, the existing methods ignore the interdependence between POI and POI category in the historical interaction. The POI and POI category sequences can be regarded as multi-view information of user check-in behaviors. This paper proposes a multi-view self-attention network (MVSAN) for next POI recommendation. Firstly, MVSAN uses a self-attention layer to update the feature representation of POI and POI category respectively. Then it generates the importance of POI under the condition of the POI category through a co-attention module. To make better use of geospatial information, we design a spatial candidate set filtering module to help the model improve recommendation performance. Experiments on two real check-in datasets show that MVSAN yields outstanding improvements over the state-of-the-art models in terms of recall.
{"title":"Multi-view Self-attention Network for Next POI Recommendation","authors":"Hao Li, P. Yue, Shangcheng Li, Fangqiang Yu, Chenxiao Zhang, Can Yang, Liangcun Jiang","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00265","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00265","url":null,"abstract":"Next Point-of-Interest (POI) recommendation has been applied by many Internet companies to enhance user travel experience. The state-of-the-art deep learning methods in next POI recommendation advocate the self-attention mechanism to model the user long-term check-in sequence. However, the existing methods ignore the interdependence between POI and POI category in the historical interaction. The POI and POI category sequences can be regarded as multi-view information of user check-in behaviors. This paper proposes a multi-view self-attention network (MVSAN) for next POI recommendation. Firstly, MVSAN uses a self-attention layer to update the feature representation of POI and POI category respectively. Then it generates the importance of POI under the condition of the POI category through a co-attention module. To make better use of geospatial information, we design a spatial candidate set filtering module to help the model improve recommendation performance. Experiments on two real check-in datasets show that MVSAN yields outstanding improvements over the state-of-the-art models in terms of recall.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"5 1","pages":"1825-1832"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74654920","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 : 2022-12-01DOI: 10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00324
Wanning Bao, Liangmin Wang, Jie Chen
The blockchain system requires every node to preserve a complete copy of data arbitrarily, which exerts tremendous storage pressure on nodes. Some researchers applied the erasure code to reduce storage redundancy. However, code storage schemes have the problem of inefficient data communication while verifying transactions and downloading data. To solve this problem, this paper proposes a lightweight locally repairable code (LRC) storage scheme inspired by the idea of slice strategy from privacy computing. Firstly, partitioning each block into distinct transaction slices substantially reduces the amount of transmitted data required to verify a transaction. Secondly, our scheme can recover single-point data with fewer code data slices by local nodes and with less network communication overhead. At last, we analyze the performance of our scheme from theoretical perspectives and examine the storage performance and computation efficiency of our scheme from experimental perspectives. Results suggest that our scheme can effectively reduce the storage overhead while also decreasing the network communication overhead and improving the data reading efficiency.
{"title":"A Lightweight Locally Repairable Code-based Storage Architecture for Blockchains","authors":"Wanning Bao, Liangmin Wang, Jie Chen","doi":"10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00324","DOIUrl":"https://doi.org/10.1109/SmartWorld-UIC-ATC-ScalCom-DigitalTwin-PriComp-Metaverse56740.2022.00324","url":null,"abstract":"The blockchain system requires every node to preserve a complete copy of data arbitrarily, which exerts tremendous storage pressure on nodes. Some researchers applied the erasure code to reduce storage redundancy. However, code storage schemes have the problem of inefficient data communication while verifying transactions and downloading data. To solve this problem, this paper proposes a lightweight locally repairable code (LRC) storage scheme inspired by the idea of slice strategy from privacy computing. Firstly, partitioning each block into distinct transaction slices substantially reduces the amount of transmitted data required to verify a transaction. Secondly, our scheme can recover single-point data with fewer code data slices by local nodes and with less network communication overhead. At last, we analyze the performance of our scheme from theoretical perspectives and examine the storage performance and computation efficiency of our scheme from experimental perspectives. Results suggest that our scheme can effectively reduce the storage overhead while also decreasing the network communication overhead and improving the data reading efficiency.","PeriodicalId":43791,"journal":{"name":"Scalable Computing-Practice and Experience","volume":"13 1","pages":"2279-2283"},"PeriodicalIF":1.1,"publicationDate":"2022-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77688727","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：