Rethinking Generative Recommender Tokenizer: Recsys-Native Encoding and Semantic Quantization Beyond LLMs
Authors:
Preprint
Abstract
Semantic ID (SID)-based recommendation is a promising paradigm for scaling sequential recommender systems, but existing methods largely follow a semantic-centric pipeline: item embeddings are learned from foundation models and discretized using generic quantization schemes. This design is misaligned with generative recommendation objectives: semantic embeddings are weakly coupled with collaborative prediction, and generic quantization is inefficient at reducing sequential uncertainty for autoregressive modeling. To address these, we propose ReSID, a recommendation-native, principled SID framework that rethinks representation learning and quantization from the perspective of information preservation and sequential predictability, without relying on LLMs. ReSID consists of two components: (i) Field-Aware Masked Auto-Encoding (FAMAE), which learns predictive-sufficient item representations from structured features, and (ii) Globally Aligned Orthogonal Quantization (GAOQ), which produces compact and predictable SID sequences by jointly reducing semantic ambiguity and prefix-conditional uncertainty. Theoretical analysis and extensive experiments across ten datasets show the effectiveness of ReSID. ReSID consistently outperforms strong sequential and SID-based generative baselines by an average of over 10%, while reducing tokenization cost by up to 122x. Code is available at https://github.com/FuCongResearchSquad/ReSID.Keywords
All Related Versions
This publication has 7 references indexed in Scilit:
- UNGER: Generative Recommendation with A Unified Code via Semantic and Collaborative IntegrationACM Transactions on Information Systems, 2025
- Recommender Systems with Generative Retrieval2023
- Session-Based Recommendation with Graph Neural NetworksProceedings of the AAAI Conference on Artificial Intelligence, 2019
- Billion-Scale Similarity Search with GPUsIEEE Transactions on Big Data, 2019
- Personalized Top-N Sequential Recommendation via Convolutional Sequence Embedding2018
- Neural Attentive Session-based Recommendation2017
- Fast unfolding of communities in large networksJournal of Statistical Mechanics: Theory and Experiment, 2008