User Embedding Model for Personalized Language Prompting
This work addresses personalized recommendation systems by enabling language models to incorporate user history embeddings, though it appears incremental as an extension of existing prompting techniques.
The researchers tackled the challenge of modeling long user histories for personalized recommendations by introducing a User Embedding Module (UEM) that compresses free-form text histories into embeddings used as soft prompts for language models, resulting in substantial improvements in predictive performance compared to conventional methods.
Modeling long histories plays a pivotal role in enhancing recommendation systems, allowing to capture user's evolving preferences, resulting in more precise and personalized recommendations. In this study we tackle the challenges of modeling long user histories for preference understanding in natural language. Specifically, we introduce a new User Embedding Module (UEM) that efficiently processes user history in free-form text by compressing and representing them as embeddings, to use them as soft prompts to a LM. Our experiments demonstrate the superior capability of this approach in handling significantly longer histories compared to conventional text based prompting methods, yielding substantial improvements in predictive performance. The main contribution of this research is to demonstrate the ability to bias language models with user signals represented as embeddings.