Dynamic integration of preference and knowledge status for knowledge concept recommendation

Qingqing Liang, Xuesong Lu, Chunyang Wang, Weining Qian, Aoying Zhou

Abstract: Highlights•We propose a novel DISKRec model to dynamically integrate preference and knowledge status for knowledge concept recommendation in online education systems.•We design a Dual-DGNNs module to disentangle two specific types of status over a continuous-time heterogeneous information network.•We design implicit dual-state integration and explicit status integration to explore the deep correlation between the two types of status.•We develop a live-update training strategy to reduce GPU memory consumption and smooth out varying data distribution in sequence learning of DISKRec.

External IDs:dblp:journals/ijon/LiangLWQZ25