Following the TRAIL: Predicting and Explaining Tomorrow's Hits with a Fine-Tuned LLM

TL;DR

This paper introduces TRAIL, a fine-tuned large language model that predicts short-term item popularity and generates natural language explanations, improving recommendation accuracy and interpretability.

Contribution

The paper presents a novel LLM-based approach that jointly forecasts item trends and provides explanations, addressing limitations of traditional recommendation systems.

Findings

TRAIL outperforms strong baselines in popularity prediction.

It generates coherent and grounded natural language explanations.

Contrastive learning improves alignment between predictions and explanations.

Abstract

Large Language Models (LLMs) have been widely applied across multiple domains for their broad knowledge and strong reasoning capabilities. However, applying them to recommendation systems is challenging since it is hard for LLMs to extract user preferences from large, sparse user-item logs, and real-time per-user ranking over the full catalog is too time-consuming to be practical. Moreover, many existing recommender systems focus solely on ranking items while overlooking explanations, which could help improve predictive accuracy and make recommendations more convincing to users. Inspired by recent works that achieve strong recommendation performance by forecasting near-term item popularity, we propose TRAIL (TRend and explAnation Integrated Learner). TRAIL is a fine-tuned LLM that jointly predicts short-term item popularity and generates faithful natural-language explanations. It employs contrastive learning with positive and negative pairs to align its scores and explanations with structured trend signals, yielding accurate and explainable popularity predictions. Extensive experiments show that TRAIL outperforms strong baselines and produces coherent, well-grounded explanations.