Connected Papers

Connected Papers is a visual academic research and literature-discovery platform that helps students, researchers, lecturers, scientists, and other professionals find relevant scholarly publications and understand how research papers relate to one another. Rather than relying entirely on traditional keyword searches or manually following reference lists, users can begin with a known research paper and generate a visual graph of related publications.

The platform is designed to make literature exploration more intuitive. When starting a dissertation, preparing a research proposal, writing a journal article, or investigating an unfamiliar subject, researchers often need to identify foundational studies, related research, and newer publications that build on previous findings. Connected Papers helps users explore these relationships through an interactive visual map.

The service is available through its website and focuses specifically on academic literature discovery. Its primary purpose is to help users understand the landscape of a research topic, discover potentially relevant publications, and develop a more comprehensive bibliography. It is particularly useful during the early stages of a research project, when identifying the most relevant papers can be challenging.

The main feature of Connected Papers is its similarity graph. Users start with an existing academic paper, known as the origin or seed paper, and the platform generates a graph containing other publications that are closely related to it.

Each node in the graph represents a research paper. The placement of the nodes reflects their similarity according to the platform’s underlying analysis. Papers that are more closely related tend to appear near one another, creating visible clusters of research within a broader subject area.

This visual structure allows users to identify connections that may not be immediately apparent from a conventional list of search results. Instead of reviewing papers one at a time, researchers can examine the overall distribution of related work and select promising publications for further reading.

Connected Papers explains that its algorithm analyses a large set of papers and selects the publications with the strongest connections to the original paper. Its similarity calculations draw on concepts including co-citation and bibliographic coupling. The resulting graph is designed to show relationships between research publications, rather than simply displaying a chronological list of citations.

The platform’s workflow is relatively straightforward. A user selects a relevant academic paper and supplies its title, DOI, or another supported identifier. Depending on the available input options, users can also start with a supported paper link or search term.

Connected Papers then analyses the selected paper against its indexed research corpus and constructs a graph of related publications. Users can explore the graph, select individual nodes, inspect paper details, and follow relevant links to publisher websites, repositories, or other sources.

The underlying data is connected to the Semantic Scholar Paper Corpus, which contains research publications from numerous scientific fields. This broad coverage allows Connected Papers to support literature discovery across areas such as computer science, engineering, medicine, environmental science, economics, and social sciences.

However, the results depend on the quality and coverage of the underlying database. A relevant paper may be absent if its metadata is incomplete or it has not been indexed. Researchers should therefore treat the graph as a discovery aid rather than a complete representation of all available literature.

Connected Papers includes two complementary views that help researchers explore the development of a research topic.

Prior Works helps users identify earlier publications associated with the research in a graph. These papers can provide historical context, introduce important concepts, and reveal the foundational studies on which later research builds. This feature is particularly useful when investigating the origins of a theory, method, or research problem.

Derivative Works helps users explore later publications connected to the papers in the graph. These may include newer studies, reviews, or research that extends existing findings. This can help users discover how a field has developed and identify publications worth investigating for a current project.

Together, these views provide a practical way to explore both the background and subsequent development of a research area. They can complement conventional citation searching, although users should still verify publication dates, citation relationships, and the relevance of individual studies.

Connected Papers also provides multi-origin graph functionality. This allows users to introduce additional origin papers into an existing exploration, helping them examine relationships between multiple starting points. It can be useful when a research question spans several related themes or when a researcher has identified multiple important papers.

The platform also offers features for saving papers and retaining graph history, depending on the account and current plan. These capabilities help users return to previous explorations rather than rebuilding every graph from scratch.

Saved graphs can be useful for organising early-stage research, revisiting a dissertation topic, and keeping track of literature discovered during an investigation. However, Connected Papers should not be mistaken for a full reference-management system. Researchers may still need a dedicated tool to manage complete bibliographic records, annotations, citations, and manuscript references.

Connected Papers is particularly valuable for students and researchers conducting literature reviews. A postgraduate student can start with a relevant journal article and use the graph to identify additional studies that may help define a research problem. A doctoral researcher can investigate foundational publications and explore related work that may have been overlooked during an initial search.

Lecturers and research supervisors can use the platform to explore reading lists and identify related publications for students. Academic writers can use it to investigate the background of a subject, discover related methods, and identify papers that deserve closer examination.

The tool can also assist researchers working on systematic or scoping reviews, but it should not be used as the only search method for such projects. Formal evidence reviews generally require documented database searches, explicit inclusion and exclusion criteria, reproducible screening procedures, and careful recording of the search process. Connected Papers can supplement those activities by revealing additional connections between known publications.

Many research questions involve multiple academic disciplines. For example, a project on sustainable urban development might require literature from environmental science, urban planning, economics, public policy, and real estate. Traditional searches can make it difficult to identify relevant work that uses different terminology across disciplines.

Connected Papers can help researchers explore related publications around a known study and identify adjacent areas worth investigating. Its visual approach may reveal clusters of related literature that a researcher would not have considered when using only a narrow keyword list.

This can support the development of research questions, theoretical frameworks, and literature-review structures. Nevertheless, researchers must assess whether each suggested paper genuinely contributes to their subject rather than assuming that visual proximity automatically means direct relevance.

Connected Papers is a browser-based service that does not require users to install a conventional desktop research application. Its graph-based interface provides an alternative to navigating long lists of search results and citation records.

The service is designed for researchers who already have at least one relevant publication or a useful starting point. Users who are completely unfamiliar with a topic may first need to locate an introductory paper through a general academic search engine or scholarly database.

The interface is primarily available in English. Researchers can still explore publications written in other languages where those publications are indexed, but interface language and paper-language coverage should be considered separately.

Connected Papers has several limitations. Its free plan restricts the number of new graphs users can generate each month, so researchers conducting extensive literature exploration may need a paid subscription. The quality of the graphs also depends on the coverage and metadata available in the underlying corpus.

The platform is not a direct replacement for academic search engines, citation databases, or reference managers. Its similarity graphs do not function as a complete citation tree, and a paper’s position in a graph should not be interpreted as proof of quality, reliability, or importance.

Researchers should read the original papers, evaluate research methods, verify findings, and assess publication quality before relying on them. Graphs are most effective when used alongside other research methods, including database searches, backward and forward citation tracking, and consultation of relevant scholarly literature.

Connected Papers is a useful research-discovery tool for students, academics, scientists, and professionals who want to explore relationships between academic publications visually. Its similarity graphs, Prior Works and Derivative Works views, multi-origin graphs, and saved research features make it particularly helpful for building bibliographies and gaining an overview of an unfamiliar research field.

Its principal strengths are ease of use, visual exploration, and the ability to uncover potentially relevant literature from a known paper. Its main drawbacks are the monthly free graph limit, dependence on indexed academic metadata, and the need to combine it with other research methods for comprehensive literature reviews.

Overall, Connected Papers is worth considering as a supplementary tool for academic research. It can make literature discovery more efficient, but the responsibility for evaluating sources and producing accurate scholarly work remains with the researcher.