PRISM: Prompt-Refined In-Context System Modelling for Financial Retrieval

The article introduces PRISM, a system for financial retrieval that leverages prompt refinement and in-context learning. The focus is on improving the accuracy and efficiency of information retrieval within the financial domain. The use of 'prompt-refined' suggests an emphasis on optimizing the prompts used to query the system, likely to improve the quality of the results. The source being ArXiv indicates this is a research paper, suggesting a novel approach to the problem.