SEMINAR: Search Enhanced Multi-modal Interest Network and Approximate Retrieval for Lifelong Sequential Recommendation

Understand

The modeling of users' behaviors is crucial in modern recommendation systems.

• A lot of research focuses on modeling users' lifelong sequences, which can be extremely long and sometimes exceed thousands of items.
• These models use the target item to search for the most relevant items from the historical sequence.
• However, training lifelong sequences in click through rate (CTR) prediction or personalized search ranking (PSR) is extremely difficult due to the insufficient learning problem of ID embedding, especially when the IDs in the lifelong sequence features do not exist in the samples of training dataset.