Connected Papers

Overview

Connected Papers is a web-based visual tool that helps researchers, students, and academics discover, navigate, and understand academic literature through interactive graph visualizations. Rather than scanning through long linear search results, users can generate a structured, visual map of a research field starting from a single seed paper. The interface makes it easy to see how papers are related, which works are central, and where emerging research is heading.

The core workflow is simple: enter a typical paper into the search bar, and Connected Papers builds a graph of similar papers in the same field. This graph provides a bird's-eye view of the academic landscape, highlighting key works, popular research directions, and the overall dynamics of the discipline. Users can click through papers, explore connections, and build further graphs for interesting findings, gradually forming a comprehensive visual understanding of a field. This is especially useful for quickly getting up to speed in a new area of study.

The platform also addresses the practical needs of literature review and academic writing. Its Prior Works view identifies important ancestor papers that establish the foundation of a given research topic, while the Derivative Works view surfaces subsequent literature reviews and recent state-of-the-art contributions that build on the input paper. For thesis writing, researchers can begin with known key references and use the graph to discover missing citations, ensuring their bibliography is complete and well-contextualized. Connected Papers draws on the Semantic Scholar database, which contains hundreds of millions of academic papers from all fields of science—from machine learning and biology to philosophy—making the tool equally useful across disciplines.

For researchers working in fast-moving areas such as machine learning, where new papers are published at an overwhelming rate, Connected Papers offers an alternative to maintaining manual paper lists. By simply searching and visually exploring, users can discover important recent publications while preserving the broader context of the field. The service supports a wide range of scholarly tasks, including field exploration, literature review, citation gap analysis, and tracking the evolution of research topics.

Core capabilities

Visual Graph Interface

Generates an interactive graph of related academic papers for any given seed paper, enabling intuitive exploration of research landscapes.

Seed Paper Search

Users can enter a typical paper by title or identifier to build a graph of similar papers in the same field.

Field Overview

Provides a high-level visual summary of a research area, revealing trends, popular works, and the relationships between papers.

Prior Works View

Focuses on important ancestor papers and foundational works that precede a chosen paper.

Derivative Works View

Highlights literature reviews and recent state-of-the-art publications that follow an input paper.

Bibliography Assistance

Helps researchers fill gaps in thesis reference lists by visually discovering overlooked relevant papers.

Comprehensive Database

Leverages the Semantic Scholar database, covering hundreds of millions of papers across all scientific fields.

Recent Paper Discovery

Makes it easy to search and visually identify important recent publications without maintaining manual lists.

Use cases

Getting a visual overview of a new academic field

Keeping track of important recent papers in fast-moving disciplines

Identifying foundational prior works for a research topic

Discovering literature reviews and state-of-the-art follow-up studies

Completing the bibliography for a thesis or academic paper

Mapping the intellectual structure of a research area before starting a project

Finding related papers that may have been missed during a manual search