Publications by authors named "Sam Yu-Te Lee"
HINTs: Sensemaking on large collections of documents with Hypergraph visualization and INTelligent agents.
IEEE Trans Vis Comput Graph
September 2024
Sensemaking on a large collection of documents (corpus) is a challenging task often found in fields such as market research, legal studies, intelligence analysis, political science, or computational linguistics. Previous works approach this problem from topic- and entity-based perspectives, but the capability of the underlying NLP model limits their effectiveness. Recent advances in prompting with LLMs present opportunities to enhance such approaches with higher accuracy and customizability.
View Article and Find Full Text PDFTowards Dataset-scale and Feature-oriented Evaluation of Text Summarization in Large Language Model Prompts.
IEEE Trans Vis Comput Graph
September 2024
Article Synopsis
- * Our study identifies five key challenges in prompt evaluation and proposes a feature-oriented workflow, focusing on summary characteristics like complexity and naturalness rather than traditional metrics like ROUGE.
- * We introduce Awesum, a visual analytics system that helps users refine summarization prompts and found that it simplifies systematic evaluation for non-technical users; the workflow could also apply to other natural language generation tasks.