Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

LabelCraft: Empowering Short Video Recommendations with Automated Label Crafting

Yimeng Bai, Yang Zhang, Jing Lu, Jianxin Chang, Xiaoxue Zang, Yanan Niu, Yang Song, Fuli Feng

TL;DR

LabelCraft is introduced, a novel automated label generation method explicitly optimizing pivotal operational metrics for platform success and effectively addresses the bi-level optimization hurdle posed by the recommender and labeling models, enabling the automatic acquisition of intricate label generation mechanisms.

Abstract

Short video recommendations often face limitations due to the quality of user feedback, which may not accurately depict user interests. To tackle this challenge, a new task has emerged: generating more dependable labels from original feedback. Existing label generation methods rely on manual rules, demanding substantial human effort and potentially misaligning with the desired objectives of the platform. To transcend these constraints, we introduce LabelCraft, a novel automated label generation method explicitly optimizing pivotal operational metrics for platform success. By formulating label generation as a higher-level optimization problem above recommender model optimization, LabelCraft introduces a trainable labeling model for automatic label mechanism modeling. Through meta-learning techniques, LabelCraft effectively addresses the bi-level optimization hurdle posed by the recommender and labeling models, enabling the automatic acquisition of intricate label generation mechanisms. Extensive experiments on real-world datasets corroborate LabelCraft's excellence across varied operational metrics, encompassing usage time, user engagement, and retention. Codes are available at https://github.com/baiyimeng/LabelCraft.

LabelCraft: Empowering Short Video Recommendations with Automated Label Crafting

TL;DR

LabelCraft is introduced, a novel automated label generation method explicitly optimizing pivotal operational metrics for platform success and effectively addresses the bi-level optimization hurdle posed by the recommender and labeling models, enabling the automatic acquisition of intricate label generation mechanisms.

Abstract

Short video recommendations often face limitations due to the quality of user feedback, which may not accurately depict user interests. To tackle this challenge, a new task has emerged: generating more dependable labels from original feedback. Existing label generation methods rely on manual rules, demanding substantial human effort and potentially misaligning with the desired objectives of the platform. To transcend these constraints, we introduce LabelCraft, a novel automated label generation method explicitly optimizing pivotal operational metrics for platform success. By formulating label generation as a higher-level optimization problem above recommender model optimization, LabelCraft introduces a trainable labeling model for automatic label mechanism modeling. Through meta-learning techniques, LabelCraft effectively addresses the bi-level optimization hurdle posed by the recommender and labeling models, enabling the automatic acquisition of intricate label generation mechanisms. Extensive experiments on real-world datasets corroborate LabelCraft's excellence across varied operational metrics, encompassing usage time, user engagement, and retention. Codes are available at https://github.com/baiyimeng/LabelCraft.
Paper Structure (24 sections, 11 equations, 3 figures, 3 tables, 1 algorithm)

This paper contains 24 sections, 11 equations, 3 figures, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Task Formulation
  3. Label generation
  4. Recommendation
  5. Methodology
  6. Overview
  7. Bi-level Optimization Problem
  8. Learning Strategy
  9. Platform Objectives
  10. Objective Representation
  11. Objective Balancing
  12. Experiment
  13. Experimental Settings
  14. Datasets.
  15. Baselines.
  16. ...and 9 more sections

Figures (3)

  • Figure 1: An overview of LabelCraft, which utilizes the platform objectives to guide the learning of the labeling model $g_{\phi}$ with the recommender model $f_{\theta}$ as the intermediary bridge.
  • Figure 2: Distribution of video duration in the top-$k$ recommended video lists generated by the PC, D2Q, DVR, and our LabelCraft for Kuaishou.
  • Figure 3: Performance of LabelCraft across different values of $\tau$, compared to the best results achieved by baselines.