Natural Language Processing and Multimodal Stock Price Prediction

TL;DR

The paper investigates predicting stock price trends by reframing targets to stock percent-change and leveraging sentiment from news articles. It employs a lightweight Bert-Tiny architecture across multiple data modalities to predict directional trends and magnitude, comparing against LSTM baselines. Key findings show that heads-up signals from headlines and company identity are particularly informative, with sector-specific data offering occasional gains, while date/source inputs may introduce noise. The approach demonstrates competitive long-term trend prediction, suggesting practical value for investors when combined with human judgment, and points to future work integrating additional modalities and raw price factors.

Abstract

In the realm of financial decision-making, predicting stock prices is pivotal. Artificial intelligence techniques such as long short-term memory networks (LSTMs), support-vector machines (SVMs), and natural language processing (NLP) models are commonly employed to predict said prices. This paper utilizes stock percentage change as training data, in contrast to the traditional use of raw currency values, with a focus on analyzing publicly released news articles. The choice of percentage change aims to provide models with context regarding the significance of price fluctuations and overall price change impact on a given stock. The study employs specialized BERT natural language processing models to predict stock price trends, with a particular emphasis on various data modalities. The results showcase the capabilities of such strategies with a small natural language processing model to accurately predict overall stock trends, and highlight the effectiveness of certain data features and sector-specific data.

Figures (13)

• Figure 1: Dataset Table
• Figure 2: The BERT-Tiny Architecture
• Figure 3: BERT vs LSTM Metrics
• Figure 4: Tech Model (Ver. 2) vs. LSTM
• Figure 5: General Model Price Direction Accuracy