Large Language Models for Semantic Monitoring of Corporate Disclosures: A Case Study on Korea's Top 50 KOSPI Companies

In the rapidly advancing domain of artificial intelligence, state-of-the-art language models such as OpenAI's GPT-3.5-turbo and GPT-4 offer unprecedented opportunities for automating complex tasks. This research paper delves into the capabilities of these models for semantically analyzing corporate disclosures in the Korean context, specifically for timely disclosure. The study focuses on the top 50 publicly traded companies listed on the Korean KOSPI, based on market capitalization, and scrutinizes their monthly disclosure summaries over a period of 17 months. Each summary was assigned a sentiment rating on a scale ranging from 1(very negative) to 5(very positive). To gauge the effectiveness of the language models, their sentiment ratings were compared with those generated by human experts. Our findings reveal a notable performance disparity between GPT-3.5-turbo and GPT-4, with the latter demonstrating significant accuracy in human evaluation tests. The Spearman correlation coefficient was registered at 0.61, while the simple concordance rate was recorded at 0.82. This research contributes valuable insights into the evaluative characteristics of GPT models, thereby laying the groundwork for future innovations in the field of automated semantic monitoring.

A New Approach for Measuring Sentiment Orientation based on Multi-Dimensional Vector Space

This study implements a vector space model approach to measure the sentiment orientations of words. Two representative vectors for positive/negative polarity are constructed using high-dimensional vec-tor space in both an unsupervised and a semi-supervised manner. A sentiment ori-entation value per word is determined by taking the difference between the cosine distances against the two reference vec-tors. These two conditions (unsupervised and semi-supervised) are compared against an existing unsupervised method (Turney, 2002). As a result of our experi-ment, we demonstrate that this novel ap-proach significantly outperforms the pre-vious unsupervised approach and is more practical and data efficient as well.