کنفرانس بین المللی فناوری اطلاعات و دانش

صفحه اصلی / پانزدهمین کنفرانس بین المللی فناوری اطلاعات و دانش

LLM-Driven Feature Extraction for Stock Market Prediction: A case study of Tehran Stock Exchange

نویسندگان :

Siavash Hosseinpour Saffarian¹ Saman Haratizadeh²

1- University of Tehran 2- University of Tehran

کلمات کلیدی :

Stock Market Prediction،Self-Attention Mechanism،Large Language Models،Information Fusion،Graph Neural Networks

چکیده :

Abstract—Stock market prediction is one of the most challenging research areas in recent years. With the emergence of deep learning and artificial intelligence, researchers have proposed various methods to predict stock market directions, considering different financial variables. One of the most significant variables influencing stock movement is user opinions and social media, which has attracted much attention from researchers in recent years. Although existing studies have introduced various methods to combine stock price and textual features, a reliable and comprehensive method has not yet been established, and there is still room for improvement. In this research, a novel method based on large language models is introduced for feature extraction from financial texts, and a self-attention mechanism is proposed to capture the internal relationships between textual and financial features. The results of the model presented in this study show a 3.10% improvement in accuracy compared to the latest competing models on a newly collected dataset.