Enhancing Travel Decision-Making: A Contrastive Learning Approach for Personalized Review Rankings in Accommodations

TL;DR

The paper tackles personalized review ranking in accommodations by introducing a large-scale, authentic dataset and a contrastive-learning framework that pairs reviewer context with reviews. It encodes context and review separately, computes an interaction matrix $F$ with $f_{i,j} = \sigma(c_i^T r_j)$, and optimizes using InfoNCE or BCE losses, with in-accommodation sampling to reflect live ranking. The approach outperforms vote-based baselines on metrics such as Mean Reciprocal Rank and Precision@k, while also offering interpretability through topic overlap analyses. This work enables more tailored decision aids for travelers and has potential for cross-domain adoption in e-commerce and other review-centric settings.

Abstract

User-generated reviews significantly influence consumer decisions, particularly in the travel domain when selecting accommodations. This paper contribution comprising two main elements. Firstly, we present a novel dataset of authentic guest reviews sourced from a prominent online travel platform, totaling over two million reviews from 50,000 distinct accommodations. Secondly, we propose an innovative approach for personalized review ranking. Our method employs contrastive learning to intricately capture the relationship between a review and the contextual information of its respective reviewer. Through a comprehensive experimental study, we demonstrate that our approach surpasses several baselines across all reported metrics. Augmented by a comparative analysis, we showcase the efficacy of our method in elevating personalized review ranking. The implications of our research extend beyond the travel domain, with potential applications in other sectors where personalized review ranking is paramount, such as online e-commerce platforms.

Figures (4)

• Figure 1: A guest review example. The respective names of the fields in our dataset are mentioned in green rectangles.
• Figure 2: Histogram of number of reviews per accommodation
• Figure 3: The proposed approach: (1) user-generated review and (2) user context are transformed into text and passed to (3) encoding layers that are fine-tuned to optimize $f_{i,j}$ diagonal to 1 and the rest to 0. (4) In inference time our model generates $f_{i,j}$ scores for a user context and a set of reviews, and ranks based on the descending order of scores.
• Figure 4: Precision@k over $k \in [1,10]$