Inferring multilayer spreading networks in online social media

Thumbnail Image

Files

master_Xia_Yan_2021.pdf (2.8 MB)

URL

Journal Title

Journal ISSN

Volume Title

Perustieteiden korkeakoulu | Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.

Authors

Date

2021-08-23

Department

Major/Subject

Machine Learning, Data Science and Artificial Intelligence

Mcode

SCI3044

Degree programme

Master’s Programme in Computer, Communication and Information Sciences

Language

en

Pages

54 + 13

Series

Abstract

Beneath the item sharing activities in online social media underlies a spreading network that connects people of common interests and opinions, where items spread through its edges from one person to another. Often, this spreading network can be decomposed into multiple layers, each representing the spreading dynamics of a certain type of items, as classified by topic, attitude, language, etc. Such a multilayer spreading network is extremely valuable for understanding the system of item spreading, for it not only provides a more fine-grained and more correct model of people's item sharing behavior under different circumstances, but also reveals a division of items by their spreading dynamics. The problem is that even single layer spreading networks are usually not directly observable in online social media, let alone their multilayer decompositions. In this work, therefore, we focus on the task of automatically inferring a multilayer spreading network from the observable spreading logs of items. Building upon the theoretical model of a previous study, we propose a novel computational framework for conducting the inference, which we show to be both effective and efficient. In addition, we systematically analyze the performance of the inference framework under different network and cascade settings, thus providing a comprehensive reference for evaluating the applicability of the method to real-world datasets. We also briefly discuss how our results inspire further improvements of the framework.

Description

Supervisor

Kivelä, Mikko

Thesis advisor

Chen, Ted

Keywords

spreading, multilayer, network, inference

Other note

Citation

Permanent link to this item

https://urn.fi/URN:NBN:fi:aalto-202108298655

Collections

[dipl] Perustieteiden korkeakoulu / SCI