Most knowledge graphs are evolving over time. Yet, most knowledge graph embedding models are designed to work on static knowledge graphs.
In this thesis, the student will focus on techniques that learn embeddings for different snapshots of KG. Stochastic Weight Average and (Adaptive Stochastic Weight Averaging)[https://arxiv.org/abs/2406.19092] techniques can be explored to learn embeddings in an efficient manner.
The student will closely work on dice-embeddings
In case you have further questions, feel free to contact Caglar Demir.