مدل محاسباتی جهت ارزیابی عملکرد عامل عملیات نفوذ در شبکه‌های اجتماعی برخط

نوع مقاله : مقاله پژوهشی

نویسندگان

1 دانشجوی دکتری، دانشگاه امام حسین (ع)، تهران، ایران

2 استاد، دانشگاه علم و صنعت، تهران، ایران

چکیده

گسترش شتابان استفاده از شبکه‌های اجتماعی برخط در میان جامعه، زمینه مستعدی برای اجرای عملیات‌های نفوذ شناختی و اجتماعی را فراهم آورده است. طرح‌ریزی و اجرای بهینه عملیات نفوذ، وابسته به داشتن یک چارچوب مناسب جهت ارزیابی این عملیات است. ارزیابی عامل‌ها و بازیگران مؤثر در عملیات نفوذ از ملزومات اصلی ارزیابی عملیات نفوذ است. با توجه به پویایی شبکه‌های اجتماعی برخط و تولید روزافزون داده‌های انبوه در آن، استفاده از رویکرد محاسباتی جهت ارزیابی عملیات نفوذ ضروری است. لذا هدف این تحقیق، یافتن مدلی محاسباتی جهت ارزیابی عامل‌های عملیات نفوذ در شبکه‌های اجتماعی برخط است. به‌طورکلی روش‌های ارزیابی نفوذ عامل را می‌توان به سه دسته کلی ارزیابی کیفی، ارزیابی کمی و ارزیابی محاسباتی تقسیم کرد. روش‌های ارزیابی محاسباتی را می‌توان به دو دسته‌ی روش‌های مبتنی بر یادگیری ماشین و روش‌های مبتنی بر ویژگی‌های دستی تقسیم‌بندی کرد. روش‌های دسته‌ی اول دارای دقت بالاتری هستند اما نیاز به حجم زیادی داده‌ی آموزش دارند. این در حالی است که در مسائلی همچون مسئله رتبه‌بندی نفوذ عامل‌ها، امکان آماده‌سازی داده‌های برچسب‌دار وجود ندارد. یکی دیگر از معایب غالب روش‌های مبتنی بر یادگیری ماشین، عدم امکان قابلیت تفسیر نتایج است. همچنین با بهره‌گیری از سازه‌های نظری مرتبط بانفوذ در شبکه‌های اجتماعی، می‌توان به مولفه‌های مؤثر در محاسبه‌ی نفوذ دست پیدا کرد. در این مقاله با تعریف شاخص‌ها و معیارهای شبکه‌ای فعالیت عامل‌ها متناسب با عملیات نفوذ، نفوذ عامل‌ها محاسبه می‌شود. در روش پیشنهادی، ابتدا مدلی جهت ارزیابی عامل با توجه به ویژگی‌های بااهمیت برای ارزیابی عملیات نفوذ معرفی شده است و سپس با مجموعه داده‌های تولیدی متناسب، ارزیابی شده است. با توجه به شاخص‌های مورد استفاده در این مدل، مجموعه دادگانی که شامل همة این شاخص‌ها باشد، وجود ندارد و ما سه مجموعه دادگان حاوی شاخص‌های مورد نظر از داده‌های توییتر تولید کردیم. نتایج به‌دست‌آمده نشانگر این مطلب است که مدل ارائه شده در کنار قابلیت تفسیرپذیری و عدم نیاز به داده‌های آموزشی، دارای عملکردی قابل مقایسه با روش‌های قبلی است.

کلیدواژه‌ها

موضوعات

عنوان مقاله [English]

a computational model to evaluate the agent of influence operations in online social networks

نویسندگان [English]

  • gholamreza bazdar 1
  • Mohammad Abdollahi Azgomi 2
1 PhD student, Imam Hossein University (AS), Tehran, Iran
2 Professor, University of Science and Technology, Tehran, Iran
چکیده [English]

The rapid and increasing use of online social networks among the society has provided a suitable field for the running cognitive and social influence operations. Optimal planning and implementation of influence operations depends on having a suitable framework for evaluating these operations. Evaluation of effective agents and actors in influence operations is one of the main requirements for evaluation of influence operations. On the other hand, considering the dynamics of online social networks and the ever-increasing production of mass data in it, it is necessary to use a computational approach to evaluate influence operations. Therefore, the purpose of this research is to find a computational model to evaluate the agents of influence operations in online social networks. In general, the methods of evaluating agent influence can be divided into three categories: qualitative evaluation, quantitative evaluation, and computational evaluation. Computational evaluation methods can be divided into two categories: methods based on machine learning (or deep learning) and methods based on hand-crafted features. The first category methods have higher accuracy but require a large amount of training data. Meanwhile, in issues such as ranking influence of agents, the preparation of labeled data is more complicated. Another disadvantage of methods based on machine learning is the inability to interpret the results. On the other hand, by using the theoretical structures related to influence in social networks, it is possible to achieve effective components in the calculation of influence. Meanwhile, better results can be achieved by combining theoretical and computational approaches. Therefore, in this article, we have presented a method that calculates the influence of agents by considering the network indicators and measures of agents' activity, by presenting the concept of user network power. In the proposed method, firstly, a model to evaluate the agent according to the important features for evaluating the influence operation is introduced based on the concept of network power, and then it is evaluated with the appropriate data set. According to the indicators used in this model, there is no data set that includes all these indicators, also we produced 3 data sets containing the desired indicators. The obtained results indicate that the presented model, in addition to interpretability and no need for training data, has a performance comparable to previous methods.

کلیدواژه‌ها [English]

  • Online social networks
  • influence operation evaluation
  • influence operation agents evaluation

Smiley face

مراجع

[1]          P. D. Allen, Information Operations Planning. Artech House, 2007.
[2]          W. Lu, “Computational social influence : models, algorithms, and applications,” University of British Columbia, 2016.
[3]          J. Stubbs and C. Bing, “Exclusive: Iran-based political influence operation - bigger, persistent, global,” 2018. https://www.reuters.com/article/us-usa-iran-facebook-exclusive/exclusive-iran-based-political-influence-operation-bigger-persistent-global-idUSKCN1LD2R9 (accessed Nov. 01, 2018).
[4]          J. Publication, “Information Operations Joint Publication 3-13,” no. November 2012.
[5]          L. Ben Jabeur, L. Tamine, and M. Boughanem, “Active microbloggers: Identifying influencers, leaders and discussers in microblogging networks,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2012, vol. 7608 LNCS, doi: 10.1007/978-3-642-34109-0_12.
[6]          B. Krishnamurthy, P. Gill, and M. Arlitt, “A few chirps about Twitter,” 2008, doi: 10.1145/1397735.1397741.
[7]          J. J. F. Forest, Influence Warfare: How Terrorists and Governments Fight to Shape Perceptions in a War of Ideas: How Terrorists and Governments Fight to Shape Perceptions in a War of Ideas. ABC-CLIO, 2009.
[8]          L. Page, S. Brin, R. Motwani, and T. Winograd, “The PageRank Citation Ranking: Bringing Order to the Web,” World Wide Web Internet Web Inf. Syst., vol. 54, no. 1999–66, 1998, doi: 10.1.1.31.1768.
[9]          Y. Yamaguchi, T. Takahashi, T. Amagasa, and H. Kitagawa, “TURank: Twitter user ranking based on user-tweet graph analysis,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2010, vol. 6488 LNCS, doi: 10.1007/978-3-642-17616-6_22.
[10]        R. Nagmoti, A. Teredesai, and M. De Cock, “Ranking approaches for microblog search,” in Proceedings - 2010 IEEE/WIC/ACM International Conference on Web Intelligence, WI 2010, 2010, vol. 1, doi: 10.1109/WI-IAT.2010.170.
[11]        T. Majer and M. Šimko, “Leveraging microblogs for resource ranking,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2012, vol. 7147 LNCS, doi: 10.1007/978-3-642-27660-6_42.
[12]        B. Hajian and T. White, “Modelling influence in a social network: Metrics and evaluation,” 2011, doi: 10.1109/PASSAT/SocialCom.2011.118.
[13]        A. Khrabrov and G. Cybenko, “Discovering influence in communication networks using dynamic graph analysis,” 2010, doi: 10.1109/SocialCom.2010.48.
[14]        Z. Ding, Y. Jia, B. Zhou, and Y. Han, “Mining topical influencers based on the multi-relational network in micro-blogging sites,” China Commun., vol. 10, no. 1, 2013, doi: 10.1109/CC.2013.6457533.
[15]        D. M. Romero, W. Galuba, S. Asur, and B. A. Huberman, “Influence and passivity in social media,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2011, vol. 6913 LNAI, no. PART 3, pp. 18–33, doi: 10.1007/978-3-642-23808-6_2.
[16]        Z. Yin and Y. Zhang, “Measuring pair-wise social influence in microblog,” 2012, doi: 10.1109/SocialCom-PASSAT.2012.10.
[17]        D. Gayo-Avello, “Nepotistic relationships in Twitter and their impact on rank prestige algorithms,” Inf. Process. Manag., vol. 49, no. 6, 2013, doi: 10.1016/j.ipm.2013.06.003.
[18]        J. Zhang, R. Zhang, J. Sun, Y. Zhang, and C. Zhang, “TrueTop: A Sybil-Resilient System for User Influence Measurement on Twitter,” IEEE/ACM Trans. Netw., vol. 24, no. 5, 2016, doi: 10.1109/TNET.2015.2494059.
[19]        I. Anger and C. Kittl, “Measuring influence on Twitter,” 2011, doi: 10.1145/2024288.2024326.
[20]        T. Noro, F. Ru, F. Xiao, and T. Tokuda, “Twitter user rank using keyword search,” Front. Artif. Intell. Appl., vol. 251, 2013, doi: 10.3233/978-1-61499-177-9-31.
[21]        M. S. Srinivasan, S. Srinivasa, and S. Thulasidasan, “Exploring celebrity dynamics on twitter,” 2013, doi: 10.1145/2528228.2528242.
[22]        M. S. Srinivasan, S. Srinivasa, and S. Thulasidasan, “A comparative study of two models for celebrity identification on twitter,” Proc. 20th Int. Conf. Manag. Data, 2014.
[23]        S. Kong and L. Feng, “A tweet-centric approach for topic-specific author ranking in micro-blog,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2011, vol. 7120 LNAI, no. PART 1, doi: 10.1007/978-3-642-25853-4_11.
[24]        D. Hatcher, G. S. Bawa, and B. de Ville, “How you can identify influencers in SAS R Social media analysis (and why it matters),” in SAS Global Forum, 2011, pp. 4–7.
[25]        Z. Y. Ding, Y. Jia, B. Zhou, Y. Han, L. He, and J. F. Zhang, “Measuring the spreadability of users in microblogs,” J. Zhejiang Univ. Sci. C, vol. 14, no. 9, 2013, doi: 10.1631/jzus.CIIP1302.
[26]        A. Silva, S. Guimarães, W. Meira, and M. Zaki, “ProfileRank: Finding relevant content and influential users based on information diffusion,” 2013, doi: 10.1145/2501025.2501033.
[27]        P. Y. Huang, H. Y. Liu, C. T. Lin, and P. J. Cheng, “A diversity-dependent measure for discovering influencers in social networks,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2013, vol. 8281 LNCS, doi: 10.1007/978-3-642-45068-6_32.
[28]        D. Liu, Q. Wu, and W. Han, “Measuring micro-blogging user influence based on user-tweet interaction model,” in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2013, vol. 7929 LNCS, no. PART 2, doi: 10.1007/978-3-642-38715-9_18.
[29]        J. Weng, E. P. Lim, J. Jiang, and Q. He, “Twitterrank: Finding topic-sensitive influential twitterers,” Proc. 3rd ACM Int. Conf. Web Search Data Min. (WSDM 2010), pp. 261–270, 2010, doi: 10.1145/1718487.1718520.
[30]        A. Aleahmad, P. Karisani, M. Rahgozar, and F. Oroumchian, “OLFinder: Finding opinion leaders in online social networks,” J. Inf. Sci., vol. 42, no. 5, 2016, doi: 10.1177/0165551515605217.
[31]        A. Pal and S. Counts, “Identifying topical authorities in microblogs,” 2011, doi: 10.1145/1935826.1935843.
[32]        M. Montangero and M. Furini, “TRank: Ranking Twitter users according to specific topics,” 2015, doi: 10.1109/CCNC.2015.7158074.
[33]        X. Li, S. Cheng, W. Chen, and F. Jiang, “Novel user influence measurement based on user interaction in microblog,” 2013, doi: 10.1145/2492517.2492635.
[34]        U. Ishfaq, H. U. Khan, S. Iqbal, and M. Alghobiri, “Finding influential users in microblogs: state-of-the-art methods and open research challenges,” Behav. Inf. Technol., 2021, doi: 10.1080/0144929X.2021.1915384.
[35]        L. Qi, Y. Huang, L. Li, and G. Xu, “Learning to rank domain experts in microblogging by combining text and non-text features,” 2015, doi: 10.1109/BESC.2015.7365953.
[36]        M. Yu, W. Yang, W. Wang, and G. W. Shen, “Information influence measurement based on user quality and information attribute in microblogging,” 2016, doi: 10.1109/ICCSN.2016.7586594.
[37]        G. S. Mahalakshmi, K. Koquilamballe, and S. Sendhilkumar, “Influential detection in twitter using tweet quality analysis,” 2017, doi: 10.1109/ICRTCCM.2017.62.
[38]        X. Luo, L. Zhang, Y. Yi, R. Xue, and D. Jiang, “The key user discovery model based on user importance calculation,” Int. J. Comput. Sci. Eng., vol. 21, no. 2, 2020, doi: 10.1504/ijcse.2020.10027436.
[39]        C. Lee, H. Kwak, H. Park, and S. Moon, “Finding influentials based on the temporal order of information adoption in Twitter,” 2010, doi: 10.1145/1772690.1772842.
[40]        C. Sun, L. Zhang, and Q. Li, “Who are influentials on micro-blogging services: Evidence from social network analysis,” 2013.
[41]        J. Yuan, L. Li, L. Luo, and M. Huang, “Topology-based algorithm for users’ influence on specific topics in micro-blog,” J. Inf. Comput. Sci., vol. 10, no. 8, 2013, doi: 10.12733/jics20102229.
[42]         and Y. J. Z. Y. B. Zhuang, Z. H. Li, “Identification of influencers in online social networks: measuring influence considering multidimensional factors exploration,” HELIYON, vol. 7, no. 4, 2021.
[43]        K. K. Darsipudi, “Influential User Detection,” 2017. https://github.com/krishnakartik1/influentialUserDetection.
[44]        “Influence Finder,” 2017. https://github.com/sapansanu/InfluenceFinder.
[45]         and Z. X. G. Wang, W. Jiang, J. Wu, “Fine-grained feature-based social influence evaluation in online social networks,” IEEE Trans. Parallel Distrib. Syst., vol. 25, no. 9, pp. 2286–2296, 2014.
پدافند الکترونیکی و سایبری
دوره 12، شماره 1 - شماره پیاپی 45
شماره پیا پی 45 بهار 1403
خرداد 1403
صفحه 89-107

فایل ها

سابقه مقاله

  • تاریخ دریافت: 11 بهمن 1402
  • تاریخ بازنگری: 31 فروردین 1403
  • تاریخ پذیرش: 14 اردیبهشت 1403
  • تاریخ انتشار: 13 خرداد 1403

هم رسانی

ارجاع به این مقاله

آمار