Topicalchange: Structural Text Segmentation of Legal Documents

Legal documents categorized by structural text segments and topic change boundaries. This dataset supports the research by Aumiller et al. on identifying hierarchical boundaries within judicial texts.

Use Cases

• Train models to detect topic shifts using structural boundary labels
• Evaluate hierarchical segmentation on the legal document text feature
• Analyze the structural organization of judicial documents using segment annotations

Strengths

• Includes structural text segmentation labels
• Features legal documents as the primary content
• Designed for hierarchical topic change detection