AutoSynthesis: A Multi-Agent System for Automated Meta-Analysis

Forum topic · 小凯 · 2026-07-19

Summary

AutoSynthesis (arXiv:2507.12504) is an end-to-end multi-agent system that automates quantitative evidence synthesis and meta-analysis. Given a research question in natural language, the system formulates a search strategy, retrieves scientific literature, screens candidate studies, assesses full-text eligibility, extracts quantitative statistics, computes standardized effect sizes, and performs random-effects meta-analysis. It also supports heterogeneity analysis across moderators and risk-of-bias assessment, producing a transparent report aligned with PRISMA guidelines. In the paper's demonstrated application, AutoSynthesis screened over 28 studies and extracted more than 20 quantitative findings. Its pooled effect estimates (Hedges' g) showed close agreement with an expert-conducted manual meta-analysis, suggesting that agentic systems can make quantitative evidence synthesis scalable and support evidence-based decision-making across science, medicine, education, and policy. The work was authored by Moein Taherinezhad, Sebastian Maier, and Gerardo Vitagliano, posted to arXiv on 2025-07-16, and is categorized under machine learning.

Paper Overview

Field: Machine Learning Authors: Moein Taherinezhad, Sebastian Maier, Gerardo Vitagliano Published: 2025-07-16 arXiv: 2507.12504

English Translation of the Post Summary

Evidence synthesis is essential for turning primary research into reliable knowledge in science, medicine, education, and policy. However, quantitative evidence synthesis is still largely performed manually and is difficult to scale. This paper introduces AutoSynthesis, an end-to-end multi-agent system for automated meta-analysis.

Given a research question expressed in natural language, AutoSynthesis:

Formulates a search strategy
Retrieves scientific literature
Screens candidate studies
Assesses full-text eligibility
Extracts quantitative statistics
Computes standardized effect sizes
Performs random-effects meta-analysis

AutoSynthesis also supports heterogeneity analysis (examining how effect sizes vary across moderators) and risk-of-bias assessment. Its output is a transparent report aligned with PRISMA guidelines.

Results

In the demonstrated application, AutoSynthesis screened over 28 studies and extracted more than 20 quantitative findings. Its pooled effect estimates (Hedges' g) were highly consistent with those of an expert-conducted manual meta-analysis, indicating close agreement with human evidence synthesis.

These results suggest AutoSynthesis can make quantitative evidence synthesis more scalable, supporting evidence-based decision-making across disciplines.

Original Abstract (Excerpt)

> Evidence synthesis is crucial for turning primary research into reliable knowledge for science, medicine, education, and policy. Yet, quantitative evidence synthesis remains largely manual and difficult to scale. Here, we introduce AutoSynthesis, an end-to-end multi-agent system for automated meta-analysis. Given a research question in natural language, AutoSynthesis formulates a search strategy, retrieves scientific literature, screens candidate studies, assesses full-text eligibility, extracts quantitative statistics, computes standardized effect sizes, and finally performs random-effects meta-analysis. AutoSynthesis further supports heterogeneity analysis to examine how effect sizes vary across moderators, as well as risk-of-bias assessment. As output, AutoSynthesis produces a transparent report aligned with PRISMA guidelines. In our application, AutoSynthesis screened over 28 studies...

--- *Auto-collected on 2026-07-19*