Textual analysis of insurance claims with large language models

Journal of Risk & Insurance, Год журнала: 2025, Номер unknown

Опубликована: Март 26, 2025

Abstract This study proposes a comprehensive and general framework for examining discrepancies in textual content using large language models (LLMs), broadening application scenarios the insurtech risk management fields, conducting empirical research based on actual needs real‐world data. Our integrates OpenAI's interface to embed texts project them into external categories while utilizing distance metrics evaluate discrepancies. To identify significant disparities, we design prompts analyze three types of relationships: identical information, logical relationships potential relationships. analysis shows that 22.1% samples exhibit substantial semantic discrepancies, 38.1% with differences contain at least one identified The average processing time each sample does not exceed 4 s, all processes can be adjusted needs. Backtesting results comparisons traditional NLP methods further demonstrate our proposed method is both effective robust.

Язык: Английский

Adverse impacts of metaverse-induced cognitive biases on the immersive shopping experience: A conceptual model developed from a qualitative approach

Technology in Society, Год журнала: 2025, Номер unknown, С. 102916 - 102916

Опубликована: Апрель 1, 2025

Язык: Английский

Generative AI Adoption and Barriers: The Case of the Insurance Industry

Опубликована: Янв. 1, 2025

Язык: Английский