Representation Learning in Continuous-Time Dynamic Signed Networks

By Kartik S et al

Read the original document by opening this link in a new tab.

Table of Contents

1. Introduction
2. Related Work
3. Problem Definition
4. SEMBA: Our Approach

Summary

Representation Learning in Continuous-Time Dynamic Signed Networks discusses the modeling of dynamic signed networks, focusing on understanding conflicting relationships and interactions in real-time. The proposed SEMBA model incorporates signed memories, balanced aggregation, and long-term propagation to effectively learn representations of dynamic signed networks. It addresses challenges such as temporal-awareness, staleness, and sign-awareness problems. The model aims to predict future links, their signs, and weights in dynamic signed networks.