Develop adaptive interaction between semantic branches

Develop a learnable fusion gate or a contrastive-alignment mechanism for the long-term and short-term semantic branches of DSRec to enable more adaptive interactions than residual fusion.

Background

DSRec currently coordinates its long-term and short-term semantic branches through residual cross-fusion. The authors explicitly identify learnable fusion gating and contrastive alignment as unresolved alternatives that could produce more adaptive cross-branch interactions.

References

While DSRec achieves advanced performance, several limitations remain open for future research. First, the item polysemous modeling is novel but basic, we plan to extend by context aware dual-tokenization to incorporate item-side semantic roles, such as Variational AutoEncoder (VAE). Second, sparse context information in transactions need the integration of external knowledge (such as graph structure) into the interest embedding, which may further enhance generalization. Finally, exploring a learnable fusion gate or contrastive alignment between the two semantic branches could yield more adaptive interactions beyond residual fusion.

— Dual-Interest Sequential Product Recommendation With Multi-Granular SSM  (2609.21548 - Liao et al., 18 Sep 2026) in Conclusion and Future Work, subsection “Future Work”