{"title":"Research on Multi-level Location Privacy Protection for Mobile Crowd Sensing based on Blockchain","authors":"Jingpei Chen, Xiaoling Wu, Jie Ling, Hoon Heo","doi":"10.1145/3587716.3587794","DOIUrl":"https://doi.org/10.1145/3587716.3587794","url":null,"abstract":"To address the leakage problem of user location privacy in mobile crowd sensing, a blockchain-based multi-level location privacy protection scheme for mobile crowd sensing is proposed. It uses private chains to decentralize user transaction records. Combined with the order-preserving encryption algorithm and the multi-level privacy cost mechanism, the server can allocate task rewards even when the users’ personalized location privacy is protected. Smart contracts are adopted to realize workers’ autonomous choice of tasks and ensure safe and fair transactions. Through experimental analysis and verification, the proposed scheme not only achieves anonymity, confidentiality, verifiability, and anti-tampering, but also prevents link attacks. Compared with the average reward allocation mechanism, the proposed probability reward distribution mechanism has improved the success rate of task allocation by 15.1%. In addition, the encryption and decryption computation cost is lower than that of related advanced schemes.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116947886","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}
{"title":"Industrial Production Index Forecast based on EMD-TCN","authors":"Kailian Gao","doi":"10.1145/3587716.3587810","DOIUrl":"https://doi.org/10.1145/3587716.3587810","url":null,"abstract":"The industrial production index can be used to reflect the speed of industrial development and economic development. The prediction of industrial production index is of great significance for understanding the law of industrial development, industrial macro-control and guiding production practice. In this paper, we propose a deep learning method based on time empirical mode decomposition (EMD) and time convolution network (TCN). The model can capture the non-stationary characteristics of the data, and obtain the time series degradation signal, namely the intrinsic mode function, through empirical mode decomposition. In addition, the TCN architecture uses causal convolution and dilation convolution. Compared with traditional statistical models such as ARIMA and prophet and deep learning models such as LSTM and TCN, it can better capture the past information of time series. Through experimental verification, the EMD-TCN architecture has smaller prediction error.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"104 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117197724","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}
Y. Ma, Jun-Yong Huang, Jules Harris Danarson, Yi Huang, Xiao-Han Wei
{"title":"Problems and Countermeasures Facing Accounting Informatization in the Era of Big Data","authors":"Y. Ma, Jun-Yong Huang, Jules Harris Danarson, Yi Huang, Xiao-Han Wei","doi":"10.1145/3587716.3587752","DOIUrl":"https://doi.org/10.1145/3587716.3587752","url":null,"abstract":"Data has become an important part of social production factors and is driving society forward at a rapid pace. Accounting informatization has been transformed and updated compared to traditional accounting, and big data needs to be fully used. Companies that value refined management have basically formed a situation where the whole production process is informatized, and accounting informatization is only one part of it. However, the accounting informatization of many enterprises is only based on the extraction and processing of data from internal production processes, and is far from reaching the stage of intelligent analysis of benchmarking and benchmarking for investment and financing decisions, tax management, performance evaluation, etc. based on management accounting. This paper analyzes the problems faced by accounting computerization of financial enterprises and put forward suggestions for countermeasures to solve the problems.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"197 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122368057","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}
{"title":"Multimodal Language Data and Platform Construction for LCTLs Teaching","authors":"Yuan Wang, Kai Jin, Wuying Liu","doi":"10.1145/3587716.3587754","DOIUrl":"https://doi.org/10.1145/3587716.3587754","url":null,"abstract":"The progress of artificial intelligence technology has brought great opportunities for the development of Less Commonly Taught Languages (LCTLs) teaching. However, the scarcity of LCTLs resources still plagues the current teaching platform. This paper is committed to building multimodal language data including foreign language vocabulary and translation, bilingual example sentences, phonetic symbols and audio, morphological similarity recommendation, semantic similarity recommendation, etc., and using full-text index and Web Service technology to build a new LCTLs teaching platform. The detailed steps are: (1) acquiring and processing multimodal language data through web pages such as TED and Opensubtitles; (2) constructing the inverted indexes used the HashMap and storing the index files in the cache of the platform, which avoids the heavy database intervention; (3) using the classical Levenshtein distance in the morphological similar words and semantic similar words recommendation algorithm. The platform implemented based on the above data and technology finally supports Japanese, Italian, Indonesian, Malaysian, Vietnamese, Filipino, and Myanmar seven languages.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"193 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122576144","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}
{"title":"Confidence-based Utterance Selection for a Recognizer-free Spoken Dialogue System","authors":"A. Ito","doi":"10.1145/3587716.3587796","DOIUrl":"https://doi.org/10.1145/3587716.3587796","url":null,"abstract":"This work aims to develop a spoken dialogue system without a speech recognizer, which is used for a CALL system of endangered languages for language revitalization. A problem in realizing such a system is the accuracy of choosing a most-matched example utterance because of the difference in voice characteristics. Thus, this paper proposes a method to choose the selection candidate among multiple candidates selected by different speech features. The selection is based on the confidence measure calculated from the distance values between the input and the database utterances. We conducted an experiment to choose a candidate among those by MFCC and PPG and obtained a 2.6-point accuracy improvement.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127010011","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}
{"title":"SATE: A Self-Attention Based Text Topic Extraction Mechanism for Customer Service Dialog System","authors":"Ziqi Xu, Bo Cheng, Meng Niu","doi":"10.1145/3587716.3587802","DOIUrl":"https://doi.org/10.1145/3587716.3587802","url":null,"abstract":"AI customer service system has been widely used in various industries. It is a research hotspot that accurately understanding customers’ intent and extracting key information from customer service dialogues. However, customer service dialog texts are short, specialized, and sparse. Those features lead to the poor performance of existing topic extraction mechanisms. This paper proposed a self-attention based topic extraction mechanism for customer service dialogue text, which is abbreviated as SATE. It consists of a corpus expansion model that can adaptively expand a suitable number of words based on words’ class and frequency. Then, SATE introduces the self-attention mechanism to train the corpus word vectors and mine their potential semantics. The self-attention mechanism can solve the semantic confusion caused by the polysemy of words in word expansion. Finally, SATE adopts the LDA topic model to extract topics of the expanded corpus. The experimental results show that the proposed SATE method can work better than the state-of-art methods.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124210393","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}
{"title":"Parallel Simulation of Lid-Driven Flow with Internal Square Obstacle","authors":"Huaxing Tan, Lianglun Cheng, Zhuowei Wang","doi":"10.1145/3587716.3587719","DOIUrl":"https://doi.org/10.1145/3587716.3587719","url":null,"abstract":"In order to study the fluid flow in a two-dimensional cavity containing an obstacle,Lattice Boltzmann method based on multi-relaxation time model is used in this paper to simulate the flow in lid-driven cavity contains an obstacle for a range of Reynolds numbers, and the implemented method is optimised for parallel simulation using CUDA.For convergence, we select suitable boundary treatment conditions, such as slip-free boundary conditions with fixed avoidance at the sides and bottom and a uniform horizontal velocity at the top of the cavity. Our work focuses on observing the fluid flow in the presence of an obstacle in the cavity below a certain range of Reynolds numbers and on the CUDA parallel optimisation of the lattice Boltzmann method with multi-relaxation time, with an average maximum speedup ratio of 13.72X based on the available experimental conditions.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"363 21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125701659","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}
{"title":"Generative and contrastive based self-supervised learning model for histopathology image analysis","authors":"Hongbo Chu, Fang Li, Yonghong He, Tian Guan","doi":"10.1145/3587716.3587774","DOIUrl":"https://doi.org/10.1145/3587716.3587774","url":null,"abstract":"In recent years, the deep neural network has achieved astonishing results in the automatic Whole slide images(WSI) processing. The current mainstream approach is inseparable from a large number of manual annotations. However, labeling such giant images with billions of pixels is very labor-intensive. So the shortage of annotation has become a bottleneck in developing Whole slide image diagnostic models. Therefore, we propose a new self-supervised learning(SSL) network to solve the problem of insufficient annotation. In our work, massive semantic information can be extracted from a large number of WSI, which significantly gets rid of our dependence on the label. At the same time, the results are further refined by a silhouette-coefficient-based recursive Spectral Clustering Bipartition, which significantly improved the classification accuracy. Moreover, our framework is highly transferable and can take on many downstream tasks in pathology. Our final results are verified on the NCT-CRC-100K and MSHIR datasets. Our code is available at https://github.com/Hongbo-Chu/generative-contrastive","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132013609","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}
{"title":"Preserving model privacy for Federated Reinforcement Learning in Complementary Environments","authors":"Pengyu Xuan, Aiguo Chen, Zexin Sha","doi":"10.1145/3587716.3587815","DOIUrl":"https://doi.org/10.1145/3587716.3587815","url":null,"abstract":"Federated reinforcement learning (FRL) uses data from multiple partners interacting with the environment to train a global decision model while maintaining data privacy. In specific situations, it is necessary to protect not only data privacy but also model privacy. That is, our objective is to collaboratively train a global model that can be used privately and independently by the initiator. We refer to this need for model privacy preservation as model monopolization. It allows the task initiator to ensure the privacy of the model during the training process and to own the trained model alone. To the best of our knowledge, model monopolization protection in Federated Reinforcement Learning has not been studied. We propose monopolized federated reinforcement learning in complementary environments via function transformation. We also provide contribution metrics to ensure the sustainability of monopolized federated reinforcement learning and to provide a quantitative basis for task initiators to give back to task participants. The experimental results show that our method achieved model monopolization by training agents in the Grid-world.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127398712","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}
{"title":"A Self-attention Feature Metric Learning Method for Remote Sensing Image Retrieval","authors":"Jiahui Wu, Zhuowei Wang, Genpin Zhao, Shuo Qu","doi":"10.1145/3587716.3587783","DOIUrl":"https://doi.org/10.1145/3587716.3587783","url":null,"abstract":"Remote sensing image retrieval (RSIR) is a challenging task due to the complex backgrounds of the image and the inclusion of various semantic objects. This paper presents a novel self-attention mechanism for RSIR using deep learning networks. We elaborate on the long-distance and the multi-layer dependence of image regions within a convolutional neural network to extract the salient features of remote sensing image against a complex background. We propose a new learning network structure of a multiple similarity (MS) loss function to further enhance the discriminability of features. The experiments on the typical datasets show that our method can significantly enhance the accuracy of RSIR. Moreover, when self-attention mechanism and MS loss function are simultaneously considered in the network, the retrieval precision is better than the results obtained by the previous methods.","PeriodicalId":238441,"journal":{"name":"Proceedings of the 2023 15th International Conference on Machine Learning and Computing","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-02-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115800816","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}