Saved in:
Bibliographic Details
Main Authors: Long, Boyuan, Wang, Yueqi, Mehta, Hiloni, Zomnir, Mick, Pathak, Omkar, Meng, Changping, Jia, Ruolin, Peng, Yajun, Hong, Dapeng, Wu, Xia, Gao, Mingyan, Dalal, Onkar, Han, Ningren
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.06657
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909830996819968
author Long, Boyuan
Wang, Yueqi
Mehta, Hiloni
Zomnir, Mick
Pathak, Omkar
Meng, Changping
Jia, Ruolin
Peng, Yajun
Hong, Dapeng
Wu, Xia
Gao, Mingyan
Dalal, Onkar
Han, Ningren
author_facet Long, Boyuan
Wang, Yueqi
Mehta, Hiloni
Zomnir, Mick
Pathak, Omkar
Meng, Changping
Jia, Ruolin
Peng, Yajun
Hong, Dapeng
Wu, Xia
Gao, Mingyan
Dalal, Onkar
Han, Ningren
contents This paper presents a case study on deploying Large Language Models (LLMs) as an advanced "annotation" mechanism to achieve nuanced content understanding (e.g., discerning content "vibe") at scale within a large-scale industrial short-form video recommendation system. Traditional machine learning classifiers for content understanding face protracted development cycles and a lack of deep, nuanced comprehension. The "LLM-as-annotators" approach addresses these by significantly shortening development times and enabling the annotation of subtle attributes. This work details an end-to-end workflow encompassing: (1) iterative definition and robust evaluation of target attributes, refined by offline metrics and online A/B testing; (2) scalable offline bulk annotation of video corpora using LLMs with multimodal features, optimized inference, and knowledge distillation for broad application; and (3) integration of these rich annotations into the online recommendation serving system, for example, through personalized restrict retrieval. Experimental results demonstrate the efficacy of this approach, with LLMs outperforming human raters in offline annotation quality for nuanced attributes and yielding significant improvements of user participation and satisfied consumption in online A/B tests. The study provides insights into designing and scaling production-level LLM pipelines for rich content evaluation, highlighting the adaptability and benefits of LLM-generated nuanced understanding for enhancing content discovery, user satisfaction, and the overall effectiveness of modern recommendation systems.
format Preprint
id arxiv_https___arxiv_org_abs_2510_06657
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LLM-Powered Nuanced Video Attribute Annotation for Enhanced Recommendations
Long, Boyuan
Wang, Yueqi
Mehta, Hiloni
Zomnir, Mick
Pathak, Omkar
Meng, Changping
Jia, Ruolin
Peng, Yajun
Hong, Dapeng
Wu, Xia
Gao, Mingyan
Dalal, Onkar
Han, Ningren
Information Retrieval
This paper presents a case study on deploying Large Language Models (LLMs) as an advanced "annotation" mechanism to achieve nuanced content understanding (e.g., discerning content "vibe") at scale within a large-scale industrial short-form video recommendation system. Traditional machine learning classifiers for content understanding face protracted development cycles and a lack of deep, nuanced comprehension. The "LLM-as-annotators" approach addresses these by significantly shortening development times and enabling the annotation of subtle attributes. This work details an end-to-end workflow encompassing: (1) iterative definition and robust evaluation of target attributes, refined by offline metrics and online A/B testing; (2) scalable offline bulk annotation of video corpora using LLMs with multimodal features, optimized inference, and knowledge distillation for broad application; and (3) integration of these rich annotations into the online recommendation serving system, for example, through personalized restrict retrieval. Experimental results demonstrate the efficacy of this approach, with LLMs outperforming human raters in offline annotation quality for nuanced attributes and yielding significant improvements of user participation and satisfied consumption in online A/B tests. The study provides insights into designing and scaling production-level LLM pipelines for rich content evaluation, highlighting the adaptability and benefits of LLM-generated nuanced understanding for enhancing content discovery, user satisfaction, and the overall effectiveness of modern recommendation systems.
title LLM-Powered Nuanced Video Attribute Annotation for Enhanced Recommendations
topic Information Retrieval
url https://arxiv.org/abs/2510.06657