Toward Byzantine-Robust Distributed Learning for Sentiment Classification on Social Media Platform

被引:1
作者
Zhang, Heyi [1 ]
Wu, Jun [2 ]
Pan, Qianqian [3 ]
Bashir, Ali Kashif [4 ,5 ,6 ]
Omar, Marwan [7 ]
机构
[1] Shanghai Jiao Tong Univ, Sch Elect Informat & Elect Engn, Shanghai 200240, Peoples R China
[2] Waseda Univ, Grad Sch Informat Prod & Syst, Tokyo 1698050, Japan
[3] Univ Tokyo, Sch Engn, Tokyo 1130033, Japan
[4] Manchester Metropolitan Univ, Dept Comp & Math, Manchester M15 6BH, England
[5] Woxsen Univ, Woxsen Sch Business, Hyderabad 502345, India
[6] Lebanese Amer Univ, Dept Comp Sci & Math, Beirut 11022801, Lebanon
[7] Illinois Inst Technol, Dept Informat Technol & Management, Chicago, IL 60616 USA
来源
IEEE TRANSACTIONS ON COMPUTATIONAL SOCIAL SYSTEMS | 2024年
基金
中国国家自然科学基金;
关键词
Blockchains; Training; Blockchain; Byzantine robust; coded computing; distributed learning; sentiment classification; social media platform; BLOCKCHAIN;
D O I
10.1109/TCSS.2024.3361465
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
Distributed learning empowers social media platforms to handle massive data for image sentiment classification and deliver intelligent services. However, with the increase of privacy threats and malicious activities, three major challenges are emerging: securing privacy, alleviating straggler problems, and mitigating Byzantine attacks. Although recent studies explore coded computing for privacy and straggler problems, as well as Byzantine-robust aggregation for poisoning attacks, they are not well-designed against both threats simultaneously. To tackle these obstacles and achieve an efficient Byzantine-robust and straggler-resilient distributed learning framework, in this article, we present Byzantine-robust and cost-effective distributed machine learning (BCML), a codesign of coded computing and Byzantine-robust aggregation. To balance the Byzantine resilience and efficiency, we design a cosine-similarity-based Byzantine-robust aggregation method tailored for coded computing to filter out malicious gradients efficiently in real time. Furthermore, trust scores derived from similarity are published to the blockchain for the reliability and traceability of social users. Experimental results show that our BCML can tolerate Byzantine attacks without compromising convergence accuracy with lower time consumption, compared with the state-of-the-art approaches. Specifically, it is 6x faster than the uncoded approach and 2x faster than the Lagrange coded computing (LCC) approach. Besides, the cosine-similarity-based aggregation method can effectively detect and filter out malicious social users in real time.
引用
收藏
页码:1 / 11
页数:11
相关论文
共 48 条
  • [41] Detection and classification of social media-based extremist affiliations using sentiment analysis techniques
    Ahmad, Shakeel
    Asghar, Muhammad Zubair
    Alotaibi, Fahad M.
    Awan, Irfanullah
    HUMAN-CENTRIC COMPUTING AND INFORMATION SCIENCES, 2019, 9
  • [42] Social media-based COVID-19 sentiment classification model using Bi-LSTM
    Arbane, Mohamed
    Benlamri, Rachid
    Brik, Youcef
    Alahmar, Ayman Diyab
    EXPERT SYSTEMS WITH APPLICATIONS, 2023, 212
  • [43] A Two-Stage Voting-Boosting Technique for Ensemble Learning in Social Network Sentiment Classification
    Cui, Su
    Han, Yiliang
    Duan, Yifei
    Li, Yu
    Zhu, Shuaishuai
    Song, Chaoyue
    ENTROPY, 2023, 25 (04)
  • [44] Deep learning-based natural language sentiment classification model for recognizing users' sentiments toward residential space
    Chang, Sun-Woo
    Dong, Won-Hyeok
    Rhee, Deuk-Young
    Jun, Han-Jong
    ARCHITECTURAL SCIENCE REVIEW, 2021, 64 (05) : 410 - 421
  • [45] Sentiment Classification from Social Media Data with Gaussian Conditional Rule-Based Deep Convolutional Neural Network
    Vinod, Priya
    Sheeja, S.
    INTERNATIONAL JOURNAL OF IMAGE AND GRAPHICS, 2025,
  • [46] DEA-RNN: A Hybrid Deep Learning Approach for Cyberbullying Detection in Twitter Social Media Platform
    Murshed, Belal Abdullah Hezam
    Abawajy, Jemal
    Mallappa, Suresha
    Saif, Mufeed Ahmed Naji
    Al-Ariki, Hasib Daowd Esmail
    IEEE ACCESS, 2022, 10 : 25857 - 25871
  • [47] Depression detection via a Chinese social media platform: a novel causal relation-aware deep learning approach
    Yang Liu
    The Journal of Supercomputing, 2024, 80 : 10327 - 10356
  • [48] Depression detection via a Chinese social media platform: a novel causal relation-aware deep learning approach
    Liu, Yang
    JOURNAL OF SUPERCOMPUTING, 2024, 80 (08) : 10327 - 10356