Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Hou, Yupeng, Ni, Jianmo, He, Zhankui, Sachdeva, Noveen, Kang, Wang-Cheng, Chi, Ed H., McAuley, Julian, Cheng, Derek Zhiyuan
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2502.13581
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908491255382016
author Hou, Yupeng
Ni, Jianmo
He, Zhankui
Sachdeva, Noveen
Kang, Wang-Cheng
Chi, Ed H.
McAuley, Julian
Cheng, Derek Zhiyuan
author_facet Hou, Yupeng
Ni, Jianmo
He, Zhankui
Sachdeva, Noveen
Kang, Wang-Cheng
Chi, Ed H.
McAuley, Julian
Cheng, Derek Zhiyuan
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2502_13581
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ActionPiece: Contextually Tokenizing Action Sequences for Generative Recommendation
Hou, Yupeng
Ni, Jianmo
He, Zhankui
Sachdeva, Noveen
Kang, Wang-Cheng
Chi, Ed H.
McAuley, Julian
Cheng, Derek Zhiyuan
Information Retrieval
Machine Learning
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.
title ActionPiece: Contextually Tokenizing Action Sequences for Generative Recommendation
topic Information Retrieval
Machine Learning
url https://arxiv.org/abs/2502.13581