Connected Papers is an AI tool that visualises the connections between academic papers, enabling researchers to explore the evolution and relationships within the scientific literature. It uses an intuitive graph-based interface to show how papers are connected through citations and thematic similarity, helping users to efficiently discover new and relevant research.

Connected Papers generates graphs by analysing around 50,000 papers to find the few dozen most strongly connected to a selected source paper. These graphs arrange papers based on similarity rather than direct citations, allowing closely related papers to be placed close together even without direct citation links. The similarity metric is based on co-citation and bibliographic coupling, suggesting that papers with similar citations and references are likely to cover related topics. The algorithm uses a Force Directed Graph to visually cluster similar papers, highlighting the interconnectedness of research within the Semantic Scholar Paper Corpus database.

Researchers can use Connected Papers for a variety of purposes, including literature reviews, identifying research gaps, and finding opportunities for collaboration. It supports cross-disciplinary exploration and provides a comprehensive view of related studies.

The free version allows users to create and download up to five graphs per month. Connected Papers also offers an academic and business plan that allows users to create an unlimited number of graphs.

Personal observation: The tool uses Semantic Scholar as its data source. You can add any paper from a graph as a source to further refine your search.

Let’s give it a try: