Connected Papers: Find and Map Related Research Papers
Enter a paper title, DOI, URL, keyword, or research topic to discover related studies, influential publications, cited works, newer research, and important connections within your field. Move beyond long lists of search results and understand how academic research actually fits together.
Enter a paper title, DOI, URL, keyword, or research topic, then build a visual map of the citations, related studies, and connected authors around it.
What Is a Connected Papers Tool?
A connected papers tool starts from one paper — the seed, or origin paper — and builds a visual graph of the research most strongly related to it. Instead of a ranked list of keyword matches, you get a map: nodes for papers, lines for relationships, and clusters for shared subfields.
Two papers can sit close together on the map even when neither cites the other directly, because closeness is also shaped by shared references and overlapping citations, not citation links alone. That's what lets visual exploration surface papers a keyword search would miss — while it supports, rather than replaces, your own critical reading of what you find.
Connected Papers is a visual research tool that turns one known paper into a graph of related work, using citation and similarity signals to place strongly connected studies near each other — so you can explore a field visually instead of scrolling through search results.
A search list vs. a research map
- Keyword search — fast, but limited to matching terms
- Citation list — shows what's cited, not what's similar
- Similarity graph — surfaces related work by shared context
- Visual clusters — related subfields appear at a glance
See How Connected Papers Works
Answering: "How does Connected Papers work?" — from a single paper to a full research map in a few clicks.
- Enter a known academic paper
- Generate its research map
- Select a connected paper
- Review its title, author, year, abstract, and citations
- Explore earlier or newer related work
- Save, collect, or export useful papers
Every click deepens the map
Selecting any connected paper regenerates the graph around it — so you can walk forward into newer research or backward into the foundational work it builds on, one paper at a time.
How Does Connected Papers Find Related Research?
Five steps, from a single input to a graph you can keep exploring.
Start with a paper or research topic
Search using a paper title, DOI, academic URL, keyword, author, or a supported identifier such as a PubMed or arXiv ID.
Analyze academic relationships
Signals like shared references, direct citations, co-citation patterns, topic and abstract similarity, and research-field relevance are used to judge how closely two papers relate.
Build a connected research graph
Papers become nodes, relationships become links, and closely related work forms visible clusters — with size, colour, and distance each carrying meaning (more on that below).
Recommend the most relevant papers
Foundational, highly cited, closely related, and recently published papers all surface within the same map, instead of separate searches.
Continue exploring from any paper
Select another node and the graph regenerates around it — building a deeper research path with every click.
How to Understand the Paper Connection Map
Every part of the graph is deliberate. Here's what each element means.
Paper Nodes
Each node represents a single academic paper — its position, size, and colour are never arbitrary.
Connecting Lines
A line indicates a calculated relationship or similarity between two papers, not necessarily a direct citation.
Research Clusters
Groups of nearby nodes often share a topic, method, theory, or subfield worth exploring together.
Node Size
Larger nodes generally reflect higher citation counts or stronger influence within the map.
Node Colour
Colour reflects publication recency, helping you spot older foundational work versus newer studies at a glance.
Distance Between Papers
Closer papers are more strongly related; distance grows as similarity and shared context decrease.
The origin paper is the seed you searched for — it's marked distinctly on the map, usually near the centre, since every other node's position is calculated relative to it.
Explore More Than a Standard Paper Search
Every capability below is part of the Connected Papers workflow.
Visual Paper Mapping
View connected research as a graph rather than a basic ranked list.
Related Paper Finder
Find studies that examine similar questions, methods, topics, or datasets.
Citation Mapping
Follow citations to understand how papers influence one another over time.
Papers That Cite a Paper
Discover newer articles that cited a selected publication.
References and Prior Research
Trace the important studies that came before your selected paper.
Recent and Derivative Research
Find newer research, reviews, replications, and developments.
Multi-Paper Connections
Add two or more origin papers to find the research they share.
Filters and Sorting
Narrow by publication year, author, citation count, journal, field, and more.
Paper Viewer
Preview essential details without ever leaving the graph.
Save and Collect Papers
Create a research reading list or project collection as you explore.
Export Citations
Export references in formats like BibTeX, RIS, or CSV.
Research History
Return to previously generated maps and paper searches any time.
Choose the Right Connected Papers Plan
Start free, and upgrade when your research needs larger maps or deeper history. Current limits are always shown on the pricing page.
Free
Explore Connected Papers and see how it fits your research.
- Limited paper maps per month
- Related-paper discovery
- Basic citation mapping
- Full graph exploration
- Paper previews
- Standard filters and exports
Premium
More maps, deeper history, and room for ongoing projects.
- Increased or unlimited paper maps
- Larger research graphs
- Multi-origin graphs
- Advanced filters and bulk export
- Unlimited saved papers and full history
- Team collaboration and priority support
Why researchers use paper mapping
A graph doesn't just show what exists — it shows how it relates, so field orientation takes minutes instead of days.
Why Researchers Use Connected Paper Mapping
Move beyond exact keyword matching
Understand a new research field faster
Find papers you might otherwise miss
Identify foundational research
Discover recent research developments
Save time during literature reviews
Strengthen academic bibliographies
Explore interdisciplinary connections
Connected Papers for Different Users
Undergraduate students
- Find credible academic sources
- Understand unfamiliar topics
- Prepare research assignments
Master's and PhD students
- Conduct literature reviews
- Build thesis bibliographies
- Locate research gaps
- Trace theoretical foundations
Researchers and scientists
- Explore related methods
- Track field development
- Find recent, influential papers
- Investigate interdisciplinary work
Professors and lecturers
- Prepare reading lists
- Introduce a research field
- Find teaching materials
Academic writers
- Find supporting evidence
- Verify related work
- Expand reference coverage
R&D teams and librarians
- Review scientific literature
- Explore technical approaches
- Teach literature-search techniques
Find Connected Research Across Academic Fields
From machine learning to philosophy, the graph works the same way everywhere.
Artificial Intelligence & ML
Trace model architectures back to their foundational papers.
Medicine & Healthcare
Follow clinical evidence from trial to trial.
Biology & Biotechnology
Map methods and findings across related studies.
Engineering
Explore technical approaches to a shared problem.
Computer Science
Connect algorithms, benchmarks, and systems work.
Environmental Science
Trace findings across ecosystems and datasets.
Psychology
Follow theories and replications through the field.
Education
Compare pedagogical studies and outcomes.
Economics & Finance
Connect models, data, and policy analysis.
Social Sciences
Map qualitative and quantitative research together.
Business & Management
Trace frameworks back to their origins.
Law, Philosophy & Humanities
Explore how ideas and arguments build on one another.
What Can You Do With Connected Papers?
How to Find Connected Papers in Five Steps
Enter a paper or topic
Search using a title, DOI, URL, keyword, author, or identifier.
Choose the correct paper
Confirm the author, year, journal, and title before building the map.
Generate the paper map
Build a visual map around your selected origin paper.
Explore related research
Open nodes, examine clusters, apply filters, and follow citation paths.
Save or export useful papers
Create a collection and export citations for later review.
Connected Papers Example
A single origin paper, mapped into its closest related studies, its foundational source, a recent follow-up, and a review — all from one search.
| Role in the map | Paper | Year | Relationship |
|---|---|---|---|
| Origin paper | Deep Reinforcement Learning for Robotic Grasping | 2020 | — |
| Closely related | Sim-to-Real Transfer for Manipulation Policies | 2021 | Shared method and dataset |
| Closely related | Vision-Based Grasp Planning with CNNs | 2019 | Overlapping citations |
| Foundational | Playing Atari with Deep Reinforcement Learning | 2013 | Cited by origin paper |
| Recent follow-up | Generalizable Grasping via Foundation Models | 2024 | Cites origin paper |
| Review paper | A Survey of Robot Learning for Manipulation | 2023 | Synthesizes the cluster |
Connected Papers vs Keyword-Based Paper Search
Most researchers benefit from using both — neither fully replaces the other.
| Capability | Connected Paper Mapping | Traditional Paper Search |
|---|---|---|
| Search starting point | Known paper or topic | Mainly keywords |
| Results format | Visual graph and lists | Ranked result list |
| Relationship discovery | Strong | Limited |
| Topic clusters | Visible | Usually not visible |
| Research timeline | Easy to observe | Requires manual review |
| Interdisciplinary discovery | Easier | Dependent on terminology |
| Best use | Exploration and mapping | Precise database retrieval |
Paper Similarity Mapping Is Not the Same as a Citation Tree
A citation tree only shows direct citation relationships — what actually cites what. Connected Papers is not a citation tree: it also uses co-citation and shared-reference signals, so two papers can appear close together even if neither one cites the other.
Citation search is best when you need to trace an exact citation trail. Similarity mapping is best when you want to discover conceptually related work that a citation list alone wouldn't surface.
Two related, but different, signals
- Direct citation — paper A explicitly cites paper B
- Co-citation — other papers cite A and B together
- Shared references — A and B draw on the same sources
- Topical similarity — A and B address a similar question
Choose the Right Research Discovery Method
Connected paper mapping is strongest for discovery and field orientation; other methods remain important for exhaustive, reproducible searching.
| Method | Strongest for | Weaker for |
|---|---|---|
| Connected Papers | Visual discovery, field orientation, timelines | Exhaustive systematic search |
| Google Scholar | Broad keyword search, citation counts | Visualizing relationships |
| Research databases | Reproducible, exhaustive retrieval | Discovering unexpected connections |
| Reference managers | Organizing sources you already have | Finding new sources |
| AI research assistants | Summarizing and answering questions | Mapping the wider research landscape |
| Manual reference checking | Verifying a specific citation trail | Speed and scale |
How to Evaluate Papers Found Through a Research Graph
A connected map helps you discover papers — it doesn't judge their quality for you.
Check the Publication Source
Review the journal, conference, repository, or publisher behind the paper.
Confirm the Publication Type
Differentiate peer-reviewed articles, preprints, reviews, and conference papers.
Read the Methodology
Don't rely only on titles, abstracts, or citation numbers.
Review Publication Date
Consider whether your topic needs current or foundational research.
Examine Citations Carefully
Highly cited doesn't automatically mean methodologically strong.
Look for Corrections or Retractions
Verify the paper's current publication status before relying on it.
Build a Better Literature Review With Connected Papers
Define your research question, select several trusted seed papers, and generate individual and multi-origin maps to identify major clusters, foundational works, and newer research worth including.
A connected graph can support literature discovery, but systematic reviews still require transparent database searches, documented inclusion criteria, and human evaluation of every source.
From map to bibliography
- Select several trusted seed papers
- Generate multi-origin maps
- Identify clusters and conflicting findings
- Apply inclusion and exclusion criteria
- Search academic databases separately
- Record and cite your final sources
Can Connected Papers Help Identify Research Gaps?
It can help you notice possible gaps — but a graph alone can't prove that a true research gap exists.
Small or isolated clusters
May point to an underexplored subtopic.
Missing theory connections
Gaps between related theories worth bridging.
Understudied populations
Groups rarely represented in the sample data.
Limited recent research
A topic that's gone quiet in recent years.
Repeated methods, few alternatives
A field that could use methodological diversity.
Contradictory findings
Disagreements worth resolving with new work.
Foundational, little follow-up
Influential papers with surprisingly few descendants.
Emerging intersections
New links forming between separate fields.
Common Mistakes to Avoid When Finding Research Papers
Starting with an unrepresentative paper
Your origin paper shapes the entire map.
Treating citation count as quality
Popularity and rigor aren't the same thing.
Using only one graph
Different seed papers surface different clusters.
Ignoring recent papers
Older, highly cited work can crowd out newer findings.
Depending only on AI recommendations
Use the map to guide reading, not replace it.
Skipping full-text review
Abstracts and previews aren't a substitute for reading.
Assuming every result is peer reviewed
Preprints and conference papers appear too.
Using a graph as a complete search
Systematic work still needs database searches.
Failing to record search decisions
Document what you searched and why.
Citing a paper without reading it
A node on a map is a lead, not a citation.
Where Does Our Research Data Come From?
Coverage, freshness, and known limitations — stated plainly rather than oversold.
DATA COVERAGE
Spans multiple academic disciplines through partnered academic indexes and open metadata sources; coverage varies by field and publisher.
UPDATE FREQUENCY
Indexes refresh regularly, though newly published papers may take time to appear.
METADATA & DOI MATCHING
Titles, authors, and citations are matched against DOIs where available, with manual and automated de-duplication.
ALGORITHM LIMITATIONS
Similarity scoring is a modeled estimate, not a guarantee — always verify important connections yourself.
No database is complete. Coverage is broad across many disciplines but varies by field, publisher, and how recently a paper was indexed — treat any single map as a strong starting point, not an exhaustive record.
Research Privacy and Responsible AI
Searches & saved maps
Your searches and saved research maps are tied to your account and can be managed or removed from your settings.
Human responsibility
The graph supports discovery; evaluating a source's quality and relevance is always the researcher's judgment call.
Data encryption
Account data and saved research are protected in line with our published security practices.
Citation & academic integrity
Cite the original paper you've read and verified — never a generated map or summary in its place.
Built to Support Serious Academic Research
Transparent about data sources, algorithm limitations, and what the tool can and can't tell you — because good research tools earn trust rather than claim it.
How Researchers Use Connected Papers
Mapping my seed paper surfaced a foundational 2014 study my keyword searches had missed entirely — it reshaped my whole related-work section.
I built a reading list for a new seminar in an afternoon by walking outward from one paper instead of screening fifty search results.
Moving into an adjacent field, the clusters showed me which subtopics mattered before I'd read a single abstract.
Building my thesis bibliography, the map kept surfacing recent papers that cited my key sources — sources I'd have found weeks later otherwise.
Our team compared four technical approaches to the same problem in one sitting, side by side on the same graph.
The origin-paper marker made it easy to teach students how citation and similarity actually differ, in one visual.
Turn One Research Paper Into a Complete Research Map
Start with a paper you already know and discover the studies, references, authors, topics, and research developments connected to it.
Frequently Asked Questions About Connected Papers
Answering the first five below reveals the next five — keep going until you've seen them all.
Connected Papers is a visual research tool that maps the studies most closely related to a single paper you search for, using citation and similarity signals rather than a simple keyword match.
You enter a paper, DOI, or topic, and the tool analyzes shared references, citations, and topical similarity to place related papers on a visual graph around your origin paper.
Yes — it uses AI-assisted similarity analysis to judge how closely papers relate, alongside citation data, to build each graph.
A free plan is available with a limited number of paper maps per month. Premium plans increase your map limits and unlock deeper features.
A visual map where nodes represent papers, lines represent relationships, and clusters represent related subfields — built around the paper you searched for.