A Comprehensive Review of Information Diffusion Models in Social Networks with a Proposed Deep Learning-Based Model
محورهای موضوعی : journal of Artificial Intelligence in Electrical Engineering
yeghaneh baybourdi
1
,
jali jabari
2
1 - Ahar Branch, Islamic Azad University, Ahar, Iran
2 - Ahar Branch, Islamic Azad University, Ahar, Iran
کلید واژه: Information diffusion, social networks, diffusion models, deep learning, graph neural networks.,
چکیده مقاله :
Social networks have become the primary platforms for information dissemination, significantly influencing user interactions and shaping information trends. Understanding how information spreads in these networks is crucial not only for optimizing recommendation algorithms and digital marketing strategies but also for identifying fake news and analyzing user influence. This paper provides a comprehensive review of information diffusion models, analyzing the strengths and weaknesses of traditional approaches such as probabilistic, deterministic, and influence-based models. Subsequently, a Hybrid Deep Learning Model (DLHM) is proposed, integrating Graph Neural Networks (GNNs) for modeling user relationships and Reinforcement Learning (RL) for optimizing the diffusion process. This combination enables the model to learn complex network structures while dynamically selecting optimal strategies for maximizing information spread. Experimental results indicate that the proposed model improves prediction accuracy by 25% compared to classical models and performs better in terms of scalability in large-scale networks. These findings demonstrate that combining deep learning with classical models can significantly enhance the analysis and prediction of information diffusion in social networks.
Social networks have become the primary platforms for information dissemination, significantly influencing user interactions and shaping information trends. Understanding how information spreads in these networks is crucial not only for optimizing recommendation algorithms and digital marketing strategies but also for identifying fake news and analyzing user influence. This paper provides a comprehensive review of information diffusion models, analyzing the strengths and weaknesses of traditional approaches such as probabilistic, deterministic, and influence-based models. Subsequently, a Hybrid Deep Learning Model (DLHM) is proposed, integrating Graph Neural Networks (GNNs) for modeling user relationships and Reinforcement Learning (RL) for optimizing the diffusion process. This combination enables the model to learn complex network structures while dynamically selecting optimal strategies for maximizing information spread. Experimental results indicate that the proposed model improves prediction accuracy by 25% compared to classical models and performs better in terms of scalability in large-scale networks. These findings demonstrate that combining deep learning with classical models can significantly enhance the analysis and prediction of information diffusion in social networks.
[1] M. E. J. Newman, "Networks: An Introduction," Oxford University Press, 2018.
[2] J. Leskovec, L. Adamic, and B. Huberman, "The dynamics of viral marketing," ACM Transactions on the Web, 2019.
[3] A. Goyal, F. Bonchi, and L. V. S. Lakshmanan, "Learning influence probabilities in social networks," WSDM, 2020.
[4] X. Wang, "Graph Neural Networks for Social Network Analysis," IEEE Transactions on Knowledge and Data Engineering, 2021.
[5] Y. Li et al., "Deep Learning for Social Network Analysis: A Comprehensive Review," ACM Computing Surveys, 2022.
[6] D. Kempe et al., "Maximizing the spread of influence in social networks," ACM SIGKDD, 2020.
[7] W. Chen et al., "Scalable influence maximization in social networks," IEEE Data Mining Conf., 2021.
[8] C. Wang et al., "Dynamic social influence analysis through factor graphs," Advances in Social Network Analysis, 2022.
[9] Y. Li et al., "Graph neural networks for social network analysis," IEEE TNNLS, 2023.
[10] S. Roch et al., "On the submodularity of influence in social networks," IEEE TKDE, 2020.
[11] M. Granovetter, "Threshold models of collective behavior," American Journal of Sociology, 2021.
[12] Wang, Y., Cong, G., Song, G., & Xie, K. (2010). Community-based greedy algorithm for mining top-k influential nodes in mobile social networks. Proceedings of the 16th ACM SIGKDD international conference on Knowledge discovery and data mining, 1039-1048.
[13] Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of 'small-world' networks. Nature, 393(6684), 440-442.
[14] Li, R., Wang, S., & Chang, E. Y. (2013). Multiple location profiling for users and relationships from social network and content. Proceedings of the VLDB Endowment, 5(11), 1603-1614.
[15] Guille, A., Hacid, H., Favre, C., & Zighed, D. A. (2013). Information diffusion in online social networks: A survey. ACM SIGMOD Record, 42(2), 17-28.
[16] Zhang, J., Tang, J., & Li, J. (2014). Expert finding in a social network. Proceedings of the 13th International Conference on Information and Knowledge Management, 51-58.
[17] Lerman, K., & Ghosh, R. (2010). Information contagion: An empirical study of the spread of news on Digg and Twitter social networks. Proceedings of the Fourth International AAAI Conference on Weblogs and Social Media, 90-97.
[18] Bakshy, E., Rosenn, I., Marlow, C., & Adamic, L. (2012). The role of social networks in information diffusion. Proceedings of the 21st International Conference on World Wide Web, 519-528.
[19] Li, Y., & Zhang, X. (2022). "Convolutional neural networks for information diffusion prediction in social networks." International Journal of Data Science and Analytics, 12(4), 45-60.
[20] Wang, Y., & Zhao, Q. (2021). "Support vector machine for diffusion prediction in complex networks." Journal of Machine Learning Research, 22(1), 115-130.
[21] Liu, Z., & Wang, X. (2020). "Random forests for modeling information diffusion in large-scale social networks." Computational Social Networks, 7(3), 124-138.
