The central community of Twitter ego-networks as a means for fake influencer detection
| Autor: | Nicolas Tsapatsoulis, Vasiliki Anastasopoulou, Klimis Ntalianis |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Degeneracy
Computer and Information Sciences social networks graph partitioning Computer science Egocentric network 01 natural sciences Social networks 010305 fluids & plasmas genetic algorithms World Wide Web K-core Id ego and super-ego 0103 physical sciences Genetic algorithm degeneracy community detection Social media 010306 general physics Cluster analysis k-core computer.programming_language Community detection Graph partitioning Genetic algorithms Python (programming language) Ego networks Graph (abstract data type) twitter ego- networks Natural Sciences computer Twitter ego networks |
| Zdroj: | The 4th IEEE Cyber Science and Technology Congress (CyberSciTech 2019) DASC/PiCom/DataCom/CyberSciTech |
| DOI: | 10.5281/zenodo.3237771 |
| Popis: | The central community of social networks, usuallyrepresented through the highest degree k-core of the corresponding graph, is proposed here as a compact representationof large social networks. We show that the central communityof egocentric social media networks, such as the ego networkson Twitter and Instagram, tell us much more about the actualinfluence of the ego than the whole egocentric network itself. We also propose a novel genetic algorithm for the identification ofcentral community of egocentric social networks and we examinethe importance of the proper initialisation of this algorithm. Theactual Twitter ego networks we used in our experiments alongwith the corresponding Python code are made publicly availablefor anyone who wishes to use them. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|