CHIMERA: A Knowledge Base of Scientific Idea Recombinations for Research Analysis and Ideation

TL;DR

CHIMERA is a large-scale knowledge base of scientific idea recombinations mined from literature, enabling analysis of innovation patterns and training models for cross-disciplinary research suggestions.

Contribution

It introduces a new information extraction task, curates an expert-annotated dataset, and demonstrates applications in analyzing recombination patterns and generating research hypotheses.

Findings

Successfully extracted recombination instances from scientific papers.

Analyzed cross-disciplinary recombination patterns in AI.

Generated research ideas rated as inspiring by researchers.

Abstract

A hallmark of human innovation is recombination -- the creation of novel ideas by integrating elements from existing concepts and mechanisms. In this work, we introduce CHIMERA, the first large-scale Knowledge Base (KB) of recombination examples automatically mined from the scientific literature. CHIMERA enables empirical analysis of how scientists recombine concepts and draw inspiration from different areas, and enables training models that propose cross-disciplinary research directions. To construct this KB, we define a new information extraction task: identifying recombination instances in papers. We curate an expert-annotated dataset and use it to fine-tune an LLM-based extraction model, which we apply to a broad corpus of AI papers. We also demonstrate generalization to a biological domain. We showcase the utility of CHIMERA through two applications. First, we analyze patterns of recombination across AI subfields. Second, we train a scientific hypothesis generation model using the KB, showing that it can propose directions that researchers rate as inspiring.