Multi agent reasoning predicts breakthrough scientific concept pairs
Bridging Concepts: Multi-Agent Reasoning over Scientific Concept Graphs for High-Impact Research Prediction
Artificial IntelligenceInformation RetrievalMachine Learning
Summary
Sometimes new scientific ideas come from connecting two different concepts that researchers hadn't studied together before. This paper looks at how to predict which pairs of concepts, not yet linked, will become important when combined. The authors build special graphs from research papers to represent these concepts and their connections. They introduce MiroGraph, a system where different intelligent agents create and measure features from these graphs to help predict high-impact connections in science. Their method works better than previous ways of forecasting important new research topics.
What this means in practice
- •For academic conference organizers: Identify upcoming high-impact research topics by forecasting promising new concept combinations for timely session planning.
- •For science funding agencies: Prioritize research proposals by predicting which novel concept pairs will likely generate influential scientific advances.
Authors
Donggyu Yoon, Minsung Hwang, Juneha Baek, Donghyuk Shin, Joyce Jiyoung Whang
Abstract
Within a research community, new topics often emerge when previously unconnected concepts are first studied together. We construct scientific concept graphs (SCGs)--domain knowledge graphs whose nodes are fine-grained research concepts extracted from paper text by LLMs and whose edges capture their co-occurrence in papers--and study the task of forecasting high-impact concept pairs, i.e., discovering concept pairs that are currently unconnected but will have high impact when connected. To this end, we introduce new benchmark datasets constructed from CIKM, KDD, and SIGIR conference papers, with varied prediction settings, including impact forecasting for emerging concepts unseen during training. We propose MiroGraph, a Multi-agent reasoning framework with scientific concept Graphs, in which a feature designer agent designs predictive features for a given concept pair from structural, semantic, and trend perspectives, and then three LLM agents compute the actual feature values from the enclosing subgraph of the concept pair. Using these feature values, a classifier designer agent constructs a classifier that predicts the final impact labels. MiroGraph outperforms various impact forecasting and temporal link prediction methods.