Heuristics for Opinion Diffusion via Local Elections

Rica Gonen, Martin Koutecký, Roei Menashof, Nimrod Talmon

نتاج البحث: فصل من :كتاب / تقرير / مؤتمرمنشور من مؤتمرمراجعة النظراء


Most research on influence maximization considers asimple diffusion model, in which binary information is being diffused (i.e., vertices – corresponding to agents – are either active or passive). Here we consider a more involved model of opinion diffusion: In our model, each vertex in the network has either approval-based or ordinal-based preferences and we consider diffusion processes in which each vertex is influenced by its neighborhood following a local election, according to certain “local” voting rules. We are interested in externally changing the preferences of certain vertices (i.e., campaigning) in order to influence the resulting election, whose winner is decided according to some “global” voting rule, operating after the diffusion converges. As the corresponding combinatorial problem is computationally intractable in general, and as we wish to incorporate probabilistic diffusion processes, we consider classic heuristics adapted to our setting: A greedy heuristic and a local search heuristic. We study their properties for plurality elections, approval elections, and ordinal elections, and evaluate their quality experimentally. The bottom line of our experiments is that the heuristics we propose perform reasonably well on both the real world and synthetic instances. Moreover, examining our results in detail also shows how the different parameters (ballot type, bribery type, graph structure, number of voters and candidates, etc.) influence the run time and quality of solutions. This knowledge can guide further research and applications.

اللغة الأصليةالإنجليزيّة
عنوان منشور المضيفSOFSEM 2023
العنوان الفرعي لمنشور المضيفTheory and Practice of Computer Science - 48th International Conference on Current Trends in Theory and Practice of Computer Science, SOFSEM 2023, Proceedings
المحررونLeszek Gasieniec
ناشرSpringer Science and Business Media Deutschland GmbH
عدد الصفحات15
رقم المعيار الدولي للكتب (المطبوع)9783031231001
المعرِّفات الرقمية للأشياء
حالة النشرنُشِر - 2023
الحدث48th International Conference on Current Trends in Theory and Practice of Computer Science, SOFSEM 2023 - Nový Smokovec, سلوفاكيا
المدة: ١٥ يناير ٢٠٢٣١٨ يناير ٢٠٢٣

سلسلة المنشورات

الاسمLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
مستوى الصوت13878 LNCS
رقم المعيار الدولي للدوريات (المطبوع)0302-9743
رقم المعيار الدولي للدوريات (الإلكتروني)1611-3349


!!Conference48th International Conference on Current Trends in Theory and Practice of Computer Science, SOFSEM 2023
المدينةNový Smokovec

ملاحظة ببليوغرافية

Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.


أدرس بدقة موضوعات البحث “Heuristics for Opinion Diffusion via Local Elections'. فهما يشكلان معًا بصمة فريدة.

قم بذكر هذا