Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Modeling User Intent Beyond Trigger: Incorporating Uncertainty for Trigger-Induced Recommendation

Jianxing Ma, Zhibo Xiao, Luwei Yang, Hansheng Xue, Xuanzhou Liu, Wen Jiang, Wei Ning, Guannan Zhang

TL;DR

The paper addresses the limitation of trigger-centric Trigger-Induced Recommendation by modeling both explicit and latent user intents while accounting for uncertainty in intent intensity. It introduces the Deep Uncertainty Intent Network (DUIN), comprising three modules: Explicit Intent Exploit Module for discriminative explicit intent via contrastive learning, Latent Intent Explore Module for multi-view latent intent through a graph-based relational framework and attention, and Intent Uncertainty Measurement Module for Gaussian-distributed intent intensity. DUIN trains with a combination of prediction loss and a contrastive objective, and demonstrates superior performance on three real-world datasets as well as online A/B testing on Alibaba.com. The work shows practical impact by deploying DUIN across all TIR scenarios with measurable gains in CTR and CVR, highlighting the value of uncertainty-aware intent modeling for enhanced recommendation diversity and long-term user experience.

Abstract

To cater to users' desire for an immersive browsing experience, numerous e-commerce platforms provide various recommendation scenarios, with a focus on Trigger-Induced Recommendation (TIR) tasks. However, the majority of current TIR methods heavily rely on the trigger item to understand user intent, lacking a higher-level exploration and exploitation of user intent (e.g., popular items and complementary items), which may result in an overly convergent understanding of users' short-term intent and can be detrimental to users' long-term purchasing experiences. Moreover, users' short-term intent shows uncertainty and is affected by various factors such as browsing context and historical behaviors, which poses challenges to user intent modeling. To address these challenges, we propose a novel model called Deep Uncertainty Intent Network (DUIN), comprising three essential modules: i) Explicit Intent Exploit Module extracting explicit user intent using the contrastive learning paradigm; ii) Latent Intent Explore Module exploring latent user intent by leveraging the multi-view relationships between items; iii) Intent Uncertainty Measurement Module offering a distributional estimation and capturing the uncertainty associated with user intent. Experiments on three real-world datasets demonstrate the superior performance of DUIN compared to existing baselines. Notably, DUIN has been deployed across all TIR scenarios in our e-commerce platform, with online A/B testing results conclusively validating its superiority.

Modeling User Intent Beyond Trigger: Incorporating Uncertainty for Trigger-Induced Recommendation

TL;DR

The paper addresses the limitation of trigger-centric Trigger-Induced Recommendation by modeling both explicit and latent user intents while accounting for uncertainty in intent intensity. It introduces the Deep Uncertainty Intent Network (DUIN), comprising three modules: Explicit Intent Exploit Module for discriminative explicit intent via contrastive learning, Latent Intent Explore Module for multi-view latent intent through a graph-based relational framework and attention, and Intent Uncertainty Measurement Module for Gaussian-distributed intent intensity. DUIN trains with a combination of prediction loss and a contrastive objective, and demonstrates superior performance on three real-world datasets as well as online A/B testing on Alibaba.com. The work shows practical impact by deploying DUIN across all TIR scenarios with measurable gains in CTR and CVR, highlighting the value of uncertainty-aware intent modeling for enhanced recommendation diversity and long-term user experience.

Abstract

To cater to users' desire for an immersive browsing experience, numerous e-commerce platforms provide various recommendation scenarios, with a focus on Trigger-Induced Recommendation (TIR) tasks. However, the majority of current TIR methods heavily rely on the trigger item to understand user intent, lacking a higher-level exploration and exploitation of user intent (e.g., popular items and complementary items), which may result in an overly convergent understanding of users' short-term intent and can be detrimental to users' long-term purchasing experiences. Moreover, users' short-term intent shows uncertainty and is affected by various factors such as browsing context and historical behaviors, which poses challenges to user intent modeling. To address these challenges, we propose a novel model called Deep Uncertainty Intent Network (DUIN), comprising three essential modules: i) Explicit Intent Exploit Module extracting explicit user intent using the contrastive learning paradigm; ii) Latent Intent Explore Module exploring latent user intent by leveraging the multi-view relationships between items; iii) Intent Uncertainty Measurement Module offering a distributional estimation and capturing the uncertainty associated with user intent. Experiments on three real-world datasets demonstrate the superior performance of DUIN compared to existing baselines. Notably, DUIN has been deployed across all TIR scenarios in our e-commerce platform, with online A/B testing results conclusively validating its superiority.
Paper Structure (25 sections, 14 equations, 6 figures, 4 tables)

This paper contains 25 sections, 14 equations, 6 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Method
  4. Problem Setup
  5. Deep Uncertainty Intent Network
  6. Explicit Intent Exploit Module
  7. Latent Intent Explore Module
  8. Intent Uncertainty Measurement Module
  9. Prediction and Optimization
  10. Evaluation
  11. Experimental Setup
  12. Datasets.
  13. Compared Methods.
  14. Implementation details.
  15. Evaluation Metrics.
  16. ...and 10 more sections

Figures (6)

  • Figure 1: Recommendation scenarios of an e-commerce platform. Left: Just for You. Middle: Mini Detail. Right: Detail Recommendation.
  • Figure 2: Differences in user intent modeling between existing TIR methods and DUIN. The black dashed frame are similar items, red indicates complementary items, and blue denotes trending items. Ratio denotes the proportion of purchasing users. Best viewed in color.
  • Figure 3: The architecture of DUIN consists of three modules, EIEM, LIEM and IUMM.
  • Figure 4: Parameters and Time Analysis on Alibaba.com.
  • Figure 5: The online serving architecture of DUIN.
  • ...and 1 more figures