Utilizing Pre-trained and Large Language Models for 10-K Items Segmentation

원문 초록

Extracting specific items from 10-K reports remains challenging due to variations in document formats and item presentation. Traditional rule-based item segmentation approaches often yield suboptimal results. This study introduces two advanced item segmentation methods leveraging language models: (1) GPT4ItemSeg, using a novel line-ID-based prompting mechanism to utilize GPT4 for item segmentation, and (2) BERT4ItemSeg, combining BERT embeddings with a Bi-LSTM model in a hierarchical structure to overcome context window constraints. Trained and evaluated on 3,737 annotated 10-K reports, BERT4ItemSeg achieved a macro-F1 of 0.9825, surpassing GPT4ItemSeg (0.9567), conditional random field (0.9818), and rule-based methods (0.9048) for core items (1, 1A, 3, and 7). These approaches enhance item segmentation performance, improving text analytics in accounting and finance. BERT4ItemSeg offers satisfactory item segmentation performance, while GPT4ItemSeg can easily adapt to regulatory changes. Together, they offer practical benefits for researchers and practitioners, enabling reliable empirical studies and automated 10-K item segmentation functionality.