Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multi-Domain Recommendation to Attract Users via Domain Preference Modeling

Hyunjun Ju, SeongKu Kang, Dongha Lee, Junyoung Hwang, Sanghwan Jang, Hwanjo Yu

TL;DR

DRIP framework is proposed that models users' preference at two levels (i.e., domain and item) and learns various seen-unseen domain mappings in a unified way with masked domain modeling and its ability to capture users' domain-level preferences is demonstrated.

Abstract

Recently, web platforms have been operating various service domains simultaneously. Targeting a platform that operates multiple service domains, we introduce a new task, Multi-Domain Recommendation to Attract Users (MDRAU), which recommends items from multiple ``unseen'' domains with which each user has not interacted yet, by using knowledge from the user's ``seen'' domains. In this paper, we point out two challenges of MDRAU task. First, there are numerous possible combinations of mappings from seen to unseen domains because users have usually interacted with a different subset of service domains. Second, a user might have different preferences for each of the target unseen domains, which requires that recommendations reflect the user's preferences on domains as well as items. To tackle these challenges, we propose DRIP framework that models users' preferences at two levels (i.e., domain and item) and learns various seen-unseen domain mappings in a unified way with masked domain modeling. Our extensive experiments demonstrate the effectiveness of DRIP in MDRAU task and its ability to capture users' domain-level preferences.

Multi-Domain Recommendation to Attract Users via Domain Preference Modeling

TL;DR

DRIP framework is proposed that models users' preference at two levels (i.e., domain and item) and learns various seen-unseen domain mappings in a unified way with masked domain modeling and its ability to capture users' domain-level preferences is demonstrated.

Abstract

Recently, web platforms have been operating various service domains simultaneously. Targeting a platform that operates multiple service domains, we introduce a new task, Multi-Domain Recommendation to Attract Users (MDRAU), which recommends items from multiple ``unseen'' domains with which each user has not interacted yet, by using knowledge from the user's ``seen'' domains. In this paper, we point out two challenges of MDRAU task. First, there are numerous possible combinations of mappings from seen to unseen domains because users have usually interacted with a different subset of service domains. Second, a user might have different preferences for each of the target unseen domains, which requires that recommendations reflect the user's preferences on domains as well as items. To tackle these challenges, we propose DRIP framework that models users' preferences at two levels (i.e., domain and item) and learns various seen-unseen domain mappings in a unified way with masked domain modeling. Our extensive experiments demonstrate the effectiveness of DRIP in MDRAU task and its ability to capture users' domain-level preferences.
Paper Structure (38 sections, 11 equations, 4 figures, 5 tables, 2 algorithms)

This paper contains 38 sections, 11 equations, 4 figures, 5 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Problem Formulation
  4. Notations
  5. MDRAU Task
  6. Proposed Framework
  7. Overview
  8. Domain-Specific Encoder
  9. Multi-Domain Encoder
  10. Constructing Masked Input.
  11. Contextualizing User Embeddings over Multi-Domains.
  12. Preference Modeling
  13. Domain-level preference.
  14. Item-level preference.
  15. Model Learning
  16. ...and 23 more sections

Figures (4)

  • Figure 1: A conceptual illustration of MDRAU task. A platform operates five different service domains, and each user partially interacts with a subset of the entire service domains. MDRAU aims to provide recommendations from each user's unseen domains to attract users.
  • Figure 2: The overview of the proposed DRIP framework.
  • Figure 3: Domain-level preference analysis. KL-divergence scores of each method (Best viewed in color).
  • Figure 4: Sensitivity analysis of DRIP.

Theorems & Definitions (1)

  • Definition 1: Multi-Domain Recommendation to Attract Users