SynGraph: A Dynamic Graph-LLM Synthesis Framework for Sparse Streaming User Sentiment Modeling

2025-03-06

AI Summary

This work addresses temporal, spatial, and holistic data sparsity in e-commerce streaming reviews caused by heterogeneous user activity levels. We propose a dynamic sentiment analysis framework based on a dynamic heterogeneous graph neural network, integrating large language model (LLM)-driven semantic completion and pseudo-label generation, and introducing a novel user stratification mechanism tailored to medium–long-tail and extremely sparse scenarios—enabling co-optimization of graph structural evolution and LLM semantics. Our key contributions are: (i) the first integration of dynamic graph modeling with LLM enhancement for sparse streaming sentiment forecasting; and (ii) a spatiotemporally aware hierarchical strategy to accommodate user heterogeneity. Evaluated on real-world streaming review datasets, our method achieves an 8.2% improvement in sentiment prediction accuracy and a 14.7% gain in F1-score for sparse users, demonstrating both effectiveness and cross-scenario generalizability.

Technology Category

• Data Mining & Knowledge Management: Data Stream Mining
• Natural Language Processing: Sentiment Analysis, Stylistic Analysis, and Argument Mining
• Machine Learning: Time-Series/Data Streams

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

User reviews on e-commerce platforms exhibit dynamic sentiment patterns driven by temporal and contextual factors. Traditional sentiment analysis methods focus on static reviews, failing to capture the evolving temporal relationship between user sentiment rating and textual content. Sentiment analysis on streaming reviews addresses this limitation by modeling and predicting the temporal evolution of user sentiments. However, it suffers from data sparsity, manifesting in temporal, spatial, and combined forms. In this paper, we introduce SynGraph, a novel framework designed to address data sparsity in sentiment analysis on streaming reviews. SynGraph alleviates data sparsity by categorizing users into mid-tail, long-tail, and extreme scenarios and incorporating LLM-augmented enhancements within a dynamic graph-based structure. Experiments on real-world datasets demonstrate its effectiveness in addressing sparsity and improving sentiment modeling in streaming reviews.

Problem

• Dynamic sentiment patterns in e-commerce user reviews
• Data sparsity in streaming sentiment analysis
• Temporal and contextual sentiment evolution modeling

Innovation

• Dynamic graph structure for sentiment modeling
• LLM-augmented enhancements for data sparsity
• User categorization into mid-tail, long-tail scenarios