ActionPiece: Contextually Tokenizing Action Sequences for Generative Recommendation
This work addresses a bottleneck in generative recommendation for improving prediction accuracy, representing an incremental advancement in tokenization methods.
The paper tackles the problem of suboptimal performance in generative recommendation due to context-independent tokenization of actions, proposing ActionPiece to incorporate context by tokenizing action sequences based on feature co-occurrence and set permutation regularization, achieving improved results on benchmark datasets.
Generative recommendation (GR) is an emerging paradigm where user actions are tokenized into discrete token patterns and autoregressively generated as predictions. However, existing GR models tokenize each action independently, assigning the same fixed tokens to identical actions across all sequences without considering contextual relationships. This lack of context-awareness can lead to suboptimal performance, as the same action may hold different meanings depending on its surrounding context. To address this issue, we propose ActionPiece to explicitly incorporate context when tokenizing action sequences. In ActionPiece, each action is represented as a set of item features. Given the action sequence corpora, we construct the vocabulary by merging feature patterns as new tokens, based on their co-occurrence frequency both within individual sets and across adjacent sets. Considering the unordered nature of feature sets, we further introduce set permutation regularization, which produces multiple segmentations of action sequences with the same semantics. Our code is available at: https://github.com/google-deepmind/action_piece.