Exploring Forum Post Retrieval with Generative Modeling

2026-09-29

AI Summary

This study addresses the challenge of training generative recommendation models from scratch for Facebook Forum due to data sparsity. To overcome this limitation, the authors propose a cross-platform transfer strategy based on hierarchical prefix semantic IDs. Specifically, the method leverages cross-platform group interaction data and Feed semantic IDs to directly generate recommendations through instruction tuning of a 3B-parameter large language model. Experimental results demonstrate that cross-platform semantic IDs can be effectively transferred to novel interfaces, validating the proposed strategy in cold-start scenarios. Overall, this work provides practical guidelines for the industrial deployment of generative recommendation systems.

Abstract

Generative recommendation (GR) has emerged as an alternative to embedding-based retrieval, building on the success of generative models in language and vision. We are exploring GR on Facebook Forum, a standalone application for medium-to-heavy users of Facebook Groups. Because Forum is a new surface, its own interaction data are too sparse to train a GR model from scratch. We address this with transfer along two axes: we train on a broader corpus of Facebook Groups engagements rather than Forum sessions alone, and we reuse hierarchical, prefix-based semantic IDs (SIDs) learned from cross-platform Facebook Feed data instead of fitting a Forum-specific tokenizer. A 3B-parameter instruction-tuned language model is then supervised-fine-tuned to generate SIDs directly from user context. We systematically ablate the design choices that matter most in practice, including SID construction, the composition and length of user history, and the inclusion of user-profile features. Our results show that cross-platform SIDs transfer to a new recommendation surface, and offer practical guidance for teams deploying GR on real-world social platforms.

Problem

• Generative Recommendation
• Data Sparsity
• Forum Post Retrieval
• Semantic IDs

Innovation

• Generative Recommendation
• Semantic IDs
• Transfer Learning
• Instruction-tuned Language Model
• Cross-platform