A Graph Neural Network Approach for Identification of Influencers and Micro-Influencers in a Social Network : *Classifying influencers from non-influencers using GNN and GCN
- Resource Type
- Conference
- Authors
- Bhadra, Jayati; Khanna, Amandeep Singh; Beuno H, Alexei
- Source
- 2023 International Conference on Advances in Electronics, Communication, Computing and Intelligent Information Systems (ICAECIS) Advances in Electronics, Communication, Computing and Intelligent Information Systems (ICAECIS), 2023 International Conference on. :66-71 Apr, 2023
- Subject
- Bioengineering
Communication, Networking and Broadcast Technologies
Components, Circuits, Devices and Systems
Computing and Processing
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Graphical models
Social networking (online)
Image edge detection
Predictive models
Graph neural networks
Telecommunication computing
Convolutional neural networks
GNN (Graphical Neural Network)
Social Media
Marketing
Influencer
Deep Learning
GCN (Graph Convolutional Neural Networks)
- Language
Targeting potential influencers for ad campaigns is one of the main challenges in social media marketing. This paper aims to incorporate Artificial Intelligence to segregate influencers from non-influencers. The target groups in focus are potential influencers. The authors build two different kinds of Graph Neural Networks, one is simple Graph Neural Network and the other is Graph Convolutional Network for node classification. A simple GNN model provides a convenient way of expressing node level, edge level and graph level prediction tasks. GCN classifies nodes based on their features. They compare the accuracy of these two models and propose the best model for social media targeting for ad campaigns, influencer detection. Additional research on Graphical models like Linear Threshold Model is being implemented to understand the importance of AI in detecting micro-influencers. Technology stack: Python and tensorflow.