ColdNAS: Search to Modulate for User Cold-Start Recommendation
This work addresses the cold-start user problem in recommendation systems, offering an incremental improvement by automating modulation design instead of relying on fixed functions or expertise.
The paper tackles the problem of making personalized recommendations for cold-start users with few interaction histories by proposing ColdNAS, a neural architecture search framework that automatically finds optimal modulation structures, achieving state-of-the-art performance on benchmark datasets.
Making personalized recommendation for cold-start users, who only have a few interaction histories, is a challenging problem in recommendation systems. Recent works leverage hypernetworks to directly map user interaction histories to user-specific parameters, which are then used to modulate predictor by feature-wise linear modulation function. These works obtain the state-of-the-art performance. However, the physical meaning of scaling and shifting in recommendation data is unclear. Instead of using a fixed modulation function and deciding modulation position by expertise, we propose a modulation framework called ColdNAS for user cold-start problem, where we look for proper modulation structure, including function and position, via neural architecture search. We design a search space which covers broad models and theoretically prove that this search space can be transformed to a much smaller space, enabling an efficient and robust one-shot search algorithm. Extensive experimental results on benchmark datasets show that ColdNAS consistently performs the best. We observe that different modulation functions lead to the best performance on different datasets, which validates the necessity of designing a searching-based method.