A Stock Price Trend Prediction Model Based on Tweet Sentiment Analysis and Graph Convolutional Network

Yuning Zhu

2025

Abstract

Public sentiment significantly affects investors’ decisions, often interfering with existing stock trends. In recent years, indicators reflecting public sentiment have been introduced to assist in predicting stock movements. Many studies have incorporated Graph Convolutional Network (GCN) to integrate this influential factor for prediction, achieving highly competitive results. However, existing studies have predominantly focused on official communication channels such as financial news, while neglecting in-depth exploration of public attention dynamics and textual data from general netizens. This study analyzes 1,317,352 Twitter posts to extract their textual characteristics and sentiment attributes, evaluates influence factors through interaction metrics, and constructs a knowledge graph integrated with stock market data. Leveraging GCN's superior capability in modeling node relationships, this paper have effectively achieved stock price trend prediction, demonstrating novel potential for knowledge graph applications in financial forecasting. These findings suggest the potential benefits of incorporating diverse public sentiment sources into stock prediction models and provide a foundation for further exploration of integrating social media dynamics with financial forecasting.

Paper Citation

in Harvard Style

Zhu Y. (2025). A Stock Price Trend Prediction Model Based on Tweet Sentiment Analysis and Graph Convolutional Network. In Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE; ISBN 978-989-758-765-8, SciTePress, pages 448-454. DOI: 10.5220/0013699200004670

in Bibtex Style

@conference{icdse25,
author={Yuning Zhu},
title={A Stock Price Trend Prediction Model Based on Tweet Sentiment Analysis and Graph Convolutional Network},
booktitle={Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE},
year={2025},
pages={448-454},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013699200004670},
isbn={978-989-758-765-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 2nd International Conference on Data Science and Engineering - Volume 1: ICDSE
TI - A Stock Price Trend Prediction Model Based on Tweet Sentiment Analysis and Graph Convolutional Network
SN - 978-989-758-765-8
AU - Zhu Y.
PY - 2025
SP - 448
EP - 454
DO - 10.5220/0013699200004670
PB - SciTePress