ReSpark: Leveraging Previous Data Reports as References to Generate New Reports with LLMs

๐Ÿ“… 2025-02-04
๐Ÿ“ˆ Citations: 0
โœจ Influential: 0
๐Ÿ“„ PDF
๐Ÿค– AI Summary
Large language models (LLMs) struggle to generate logically coherent and factually reliable data reports, while users bear the burden of designing analytical logic. To address this, we propose ReportSynthโ€”a novel framework that leverages historical reports as structured reference sources. It employs LLM-driven similar-report retrieval, paragraph-level analytical logic parsing, cross-report goal transfer, data-aware text rewriting, and interactive feedback to enable dynamic customization and iterative refinement of new reports. The method supports real-time editing and logical consistency verification. Experimental results demonstrate significant improvements: report completeness (+32.7%), factual accuracy (+28.4%), user satisfaction (+41.5%), and a 63.2% reduction in manual analytical effort. ReportSynth establishes a new paradigm for trustworthy, automated data report generation.

Technology Category

Application Category

๐Ÿ“ Abstract
Creating data reports is time-consuming, as it requires iterative exploration and understanding of data, followed by summarizing the insights. While large language models (LLMs) are powerful tools for data processing and text generation, they often struggle to produce complete data reports that fully meet user expectations. One significant challenge is effectively communicating the entire analysis logic to LLMs. Moreover, determining a comprehensive analysis logic can be mentally taxing for users. To address these challenges, we propose ReSpark, an LLM-based method that leverages existing data reports as references for creating new ones. Given a data table, ReSpark searches for similar-topic reports, parses them into interdependent segments corresponding to analytical objectives, and executes them with new data. It identifies inconsistencies and customizes the objectives, data transformations, and textual descriptions. ReSpark allows users to review real-time outputs, insert new objectives, and modify report content. Its effectiveness was evaluated through comparative and user studies.
Problem

Research questions and friction points this paper is trying to address.

Generates data reports using LLMs efficiently
Reduces mental load in creating analysis logic
Leverages existing reports for new report creation
Innovation

Methods, ideas, or system contributions that make the work stand out.

LLM-based method
leverages existing reports
customizes objectives and descriptions
๐Ÿ”Ž Similar Papers
No similar papers found.