MaRDI Services
If you want to add your own service, please refer to this video tutorial or this PDF tutorial [File:Portal add service.pdf].
MaRDI offers a wide range of services.
 |
MediaWiki Math Search ExtensionTool/Application (No status available.)
MediaWiki Math Search Extension: Semantic Formula Discovery Across Wikis The MediaWiki Math Search Extension enables semantic formula search across mathematical wikis, improving the discoverability of mathematical knowledge expressed in TeX or MathML. Developed and maintained by the MaRDI Portal team, it connects wiki-based formulae with the MaRDI Knowledge Graph to foster structured access to mathematical content. What does it do?The extension indexes mathematical expressions semantically, allowing users to search for equivalent or related formulas based on structure rather than plain text. It bridges formula-level content in MediaWiki installations with the semantic relationships captured in the MaRDI Knowledge Graph. How can it be used?
Maintained by: TA5
|
 |
MaRDI Portal Storage (No status available.)
MaRDI Portal Storage: Distributed and FAIR Research Data Preservation The MaRDI Portal Storage provides distributed, content-addressable storage for mathematical data, models, and executable artifacts. Built on the InterPlanetary File System (IPFS), it enables the secure, decentralized sharing of research data within the MaRDI ecosystem, ensuring long-term availability and verifiability. What does it do?It allows MaRDI partners and researchers to store and share datasets, software packages, and Jupyter notebooks in a FAIR and reproducible way. Each object is identified by a unique content hash, ensuring versioning and integrity across distributed systems. How can it be used?
Maintained by: TA5
|
 |
MaRDI Packaging SystemTool/Application (stable version)
MaRDI Packaging System: Simplifying Software PackagingMaPS, short for MaRDI Packaging System is the working name for the software system created by TA1 Computer Algebra for Measure 1.4 Predefined software environments. MaPS provides a unified interface both, to package a software inside a functional environment (called a runtime), and to install a published runtime onto a computer system in a user friendly way. A publication describing this tool can be found at https://arxiv.org/abs/2404.05563 (arxiv) or https://link.springer.com/chapter/10.1007/978-3-031-64529-7_26 (paywall) . How does it work?MaPS deploys packages including their complete library dependencies in an environment which does not conflict with other packages installed on your computer. Deploying runtimes is a one line command. Packaging a runtime is only a small overhead. More information is available at the MaPS wiki or at the MaPS WebsiteMaintained by: TA1
|
 |
MaRDI Knowledge GraphDatabase (stable version)
The MaRDI Knowledge Graph: A New Way to Explore Mathematical ResearchServing the global mathematical community, the MaRDI Knowledge Graph is a growing resource that brings together over 5 million items and 500 million relationships from various open mathematical research data sources. It integrates information from repositories such as DLMF, CRAN, PolyDB, swMATH, zbMATH Open, arXiv, and OpenML, creating connections between datasets, publications, software, and mathematical concepts. What does it do?The MaRDI Knowledge Graph helps organize and link mathematical research data, offering structured insights into how different resources relate to one another. It enables researchers and students to navigate across datasets, software tools, and research papers, making it easier to explore connections between different areas of mathematical research and applications. How can it be used?
Maintained by: TA5
|
 |
MaRDMOTool/Application (stable version)
MaRDMO: Future Gateway to FAIR Mathematical DataMaRDMO is an extension of the Research Data Management Organiser (RDMO) designed to facilitate the documentation, integration, and sharing of mathematical models, interdisciplinary workflows, and algorithms. As a user-friendly interface, it connects researchers with MaRDI services such as the MaRDI, MathModDB and MathAlgoDB knowledge graphs, while also integrating external data sources like Wikidata. By streamlining research data management through guided interviews and encouraging the reuse of existing information, MaRDMO enhances transparency, reproducibility, and interoperability across disciplines. Researchers can document, search, and share mathematical research data seamlessly, ensuring broad accessibility and usability within the scientific community. How to use it?The MaRDMO Plugin and its questionnaires are available on GitHub and can be integrated into any existing RDMO instance, whether local, project-specific, or institute-wide. MaRDMO is also accessible to all researchers via the MaRDI RDMO instance, which is hosted by the basic service DMP4NFDI. Currently, only the questionnaire for the documentation of mathematical models is available, with the others to follow gradually. To document an interdisciplinary workflow, mathematical model, or algorithm—or to search for existing ones—users complete a guided interview in MaRDMO. Once all questions are answered, the documentation can be directly published to the respective knowledge graph. If the search catalog is chosen, MaRDMO suggests suitable interdisciplinary workflows, mathematical models, or algorithms based on the user's input.Maintained by: TA4
|
 |
MaRDI Help DeskOutreach | Support/Consulting (stable version)
The MaRDI Help Desk is your first entry point to MaRDI services, support, and training. Mathematical data consultant Christiane Görgen can help you find out how to handle your research data in practice, what to write in a funding application about it, or how to connect to existing infrastructure and projects. She currently prepares training material for these topics, knows the consortium inside out, and can direct very topic-specific inquiries to the appropriate experts. Dissemination coordinator Tabea Bacher supports the initiative's outreach, community integration, and workshop implementation. She can help with organizational matters, and she is busy building bridges between MaRDI and the library communities. Maintained by: TA6
|
 |
MaRDI Open InterfacesTool/Application (beta version)
MaRDI Open Interfaces: Software project for improving Interoperability in Scientific ComputingMaRDI Open Interfaces is a project that aims to improve interoperability in scientific computing by removing two hurdles that computational scientists commonly face in their daily work. These hurdles are as follows. First, numerical solvers are often implemented in different programming languages. Second, these solvers frequently expose significantly different interfaces---they have different function names, arguments order, and call sequences. As a result, switching from one solver to another can take non-trivial effort for code modifications and correctness testing. The MaRDI Open Interfaces project aims to alleviate these problems by providing automatic data marshalling between languages and a set of interfaces for typical numerical problems, such as integration of differential equations and optimization. How can it be used?
Installation and documentationPlease refer to the documentation at https://mardi4nfdi.github.io/open-interfaces/ for deeper view on the goals and implementation details as well as installation instructions, tutorials, and API reference. The MaRDI Open Interfaces package is currently in beta and is still evolving; however, it can already be used in computational experiments requiring time integration. PublicationsMaintained by: TA2
|
 |
mlr3Tool/Application (stable version)
mlr3: Machine Learning in RThe mlr3 framework is a collection of more than 30 R packages that provides unified access to machine learning methods in R. At its core, it offers fundamental building blocks for statistical learning through standardized interfaces, enhancing interoperability between independently developed R packages. What does it do?It allows users to easily train, tune, evaluate, or benchmark a wide range of machine learning algorithms or whole ML pipelines. It supports a wide variety of ML problems, including classification, regression, and survival analysis. By establishing standardized interfaces, methods from many different R packages can be accessed in a unified way and be combined seamlessly. On the one hand, this makes it easy for scientists and ML practitioners to solve real world problems. On the other hand, this also supports researchers in the process of evaluating and benchmarking new algorithms against existing methods. For a list of available learning algorithms, which are annotated with standardized metadata, see our website. How can it be used?
Maintained by: TA3
|
 |
MathAlgoDB Knowledge Graph for Scientific ComputingTool/Application (stable version)
MathAlgoDB connects mathematical algorithms,models, problems, and software in a structured knowledge graph. It makes the landscape of scientific computing searchable, comparable, and reusable — from classical numerical methods to AI-based solvers. Whether you’re exploring new methods orcurating algorithmic metadata: MathAlgoDB will give you a transparent, interconnected view of how mathematics is implemented and applied. Join the community and help shape the future ofalgorithmic knowledge!Your voice matters!Help shape the future coverage of MathAlgoDB. Suggest topics and scientific fields that the database should document next by pointing us to a key reference (for example, a journal article, monograph, or survey) from your area. Based on this reference, we construct a basic structured knowledge graph of the algorithms in that field. You are then invited to review, correct, and extend the graph, and your contribution will be acknowledged in MathAlgoDB’s documentation and outreach material. Suggest a new topic for MathAlgoDBMaintained by: TA2
|
 |
MaRDIFlowTool/Application (beta version)
MaRDIFlow : Design and Description of FAIR CSE WorkflowsThis computational framework abstracts multi-layered components from FAIR computational experiments through an input/output pipeline. By incorporating them into redundant descriptions, we describe arbitrary levels of abstraction ranging from mere I/O data to mathematical models. A recent publication describing the tool can be found at https://arxiv.org/abs/2405.00028 Key FeaturesComprehensive Workflow Framework: Enables the abstraction of computational components to describe FAIR workflows. Multi-Layered Component Support: Ensures the integration of various layers such as numerical methods, mathematical models, and I/O data. Output Formats: Supports HTML, PDF, TeX, and Jupyter Notebook outputs for enforced documentation and reporting. What does it do?MaRDIFlow provides a structured environment for designing, executing, and documenting FAIR CSE workflows.It captures metadata and dependencies automatically, ensuring that computational experiments are reusable, reproducible, and well-documented across varying computational setups. How can it be used?Using this working prototype (two use-cases), users can execute the workflow through a user-friendly command-line interface, enabling interaction with the framework's features. Users can : 1. Define workflow components, inputs, and outputs using structured metadata descriptions. 2. Execute workflows while automatically capturing provenance information. 3. Generate detailed reports and provenance documentation in chosen output formats (HTML, PDF, or TeX). Further installation and documentation details can be found in : https://zenodo.org/records/7863520. PublicationsPavan L. Veluvali, Jan Heiland, Peter Benner, "Bridging Ontologies and Computational Workflows: A Framework for Semantic Enrichment and Reproducibility", 2nd Conference on Research Data Infrastructure (CoRDI) 2025. Pavan L. Veluvali, Jan Heiland, Peter Benner, "MaRDIFlow: A CSE workflow framework for abstracting meta-data from FAIR computational experiments", e-print arXiv:2405.00028, arXiv, 2024.Maintained by: TA2
|
 |
MathModDBDatabase (stable version)
MathModDB: A database for Mathematical ModelsMathModDB defines a data model with classes (Mathematical Model, Mathematical Expression, Academic Discipline, Research Problem, Quantity (Kind), Computational Task, Scholarly Article), object properties/relations, data properties and annotation properties as an ontology. This ontology is populated with individuals/data from various fields of applied mathematics, making it a knowledge graph. For more information, see the official MathModDB website at https://portal.mardi4nfdi.de/wiki/MathModDB. What does it do?The MathModDB Knowledge Graph helps organize and link mathematical models, offering structured insights into mathematical modelling and related computational tasks. It enables researchers and students to navigate across mathematical models and expressions, as well as scholarly articles, making it easier to explore connections between research problems from different academic disciplines. How can it be used?
To learn more about the MathModDB Knowledge Graph, please visit its official website https://portal.mardi4nfdi.de/wiki/MathModDB. The MathModDB Knowledge Graph is still evolving, and its coverage is not yet complete. However, it provides a foundation for discovering research on mathematical modelling in a structured way. As the graph grows, contributions and feedback from the community will help refine and expand its usefulness over time.Maintained by: TA4
|
 |
mrdi File FormatTool/Application (stable version)
The mrdi File FormatDue to the nature of data in computer algebra, the storage of such data requires a more sophisticated format. The mrdi file format is a JSON based file format with the necessary structure for saving and loading common types among computer algebra software. A publication describing our file format can be found here. SpecificationYou can find the format specification here. We have a paper describing the specification available on the arxiv and published in the proceedings of the icms here. UsabilitySerializing to our file format is available from the Oscar.jl computer algebra system. There is also an implementation for some of the types available in the Magma computer algebra system here. We are also working on an implementation for the Macaulay2 and Sage computer algebra systems. ExamplesWe list some examples where The file format has been used in practice. -Combinatorics of slices of cubes -The F-theory geometry with most flux vacua -Finite groups of symplectic birational transformations of IHS manifolds of OG10 type. - The file format is an essential building block for the prototype OscarDB.- A project that interfaces Oscar and Lean using the file form can be found here. Maintained by: TA1
|
 |
Community - Graphical Modelling and Causal InferenceCurated Collection (stable version)
The Dataset collection on Graphical Modelling and Causal InferenceWe host the Zenodo community Graphical Modelling and Causal Inference. On this platform, we curate and present topical datasets, dataset collections, and metadata. Exemplary statistical notebooks showcase advances in methodology and present new applications. The community supports content moderation by TA3, and we will encourage and solicit submissions of datasets and notebooks by researchers from the broader academic community. See also our Library Website for renderings of the notebooks. What does it provide?
How to use it?
Maintained by: TA3
|
 |
MaRDI StationOutreach | Tool/Application (stable version)
The MaRDI station offers an educational, gamified approach to research data management. It comes in three versions. There are two multiplayer versions: a portable one, consisting of three notepads and a projector, and a version with exhibition furniture, consisting of three consoles installed on a desk alongside a projector. In both multiplayer versions, the MaRDI station includes one of the following open-source games: A game designed for the general public about freedom of data or a game for mathematicians of all kinds, exploring research data topics in a university campus setting. The third version is a single-player adventure game, also dealing with research data topics. Get a sneak peek here. The latter is ideal for mathematicians seeking to familiarize themselves with research data topics. If you want to rent and showcase the MaRDI station at your workshop, conference, or event, you can contact us via station@mardi4nfdi.de. Since the game engine is open source, you may also create your own adventure with stories of any topic you like, and set in a world of your choice. Maintained by: TA6
|
 |
Best PracticesSupport/Consulting (stable version)
MaRDI offers support and consultancy for making your own mathematics FAIR. One example is the project "small phylogenetic trees": an early 2000's mathematical library transformed into a modern website, a software package, and a best practice report. Another example is the FAIRified mathematical research-data repository MathRepo. Maintained by: TA6
|
 |
MaRDI Knowledge Graph Query ServiceWeb application | Tool/Application (stable version)
MaRDI Knowledge Graph Query Service: Explore Relationships via SPARQL The MaRDI Knowledge Graph Query Service offers a user-friendly interface for exploring the MaRDI Knowledge Graph using structured SPARQL queries. It allows users to access, filter, and visualize relationships between mathematical entities such as formulas, datasets, algorithms, and publications. What does it do?It provides an interactive web-based SPARQL endpoint that exposes the rich semantic network of the MaRDI Knowledge Graph. Researchers can formulate precise queries to extract mathematical connections and generate structured insights. How can it be used?
Maintained by: TA5
|
 |
Library - Graphical Modelling and Causal InferenceCurated Collection (stable version)
The Library for Graphical Modelling and Causal InferenceWe host the Library Website, extending our Graphical Modelling and Causal Inference Zenodo Community to provide a place for FAIR (= Findable, Accessible, Interoperable, and Reusable) GMCI research and analyses. On this website, we (A) extensively present the community and its scope, (B) visualize exemplary statistical notebooks that showcase advances in methodology, present new dataset applications, or introduce programming packages, and (C) give detailed contributing instructions. The community supports content moderation by TA3, and we will encourage and solicit submissions of datasets and notebooks by researchers from the broader academic community. What does it provide?
How to use it?
Maintained by: TA3
|
 |
MaRDI PortalWeb application (stable version)
The MaRDI Portal: Your Gateway to FAIR Mathematical Research Data The MaRDI Portal is the central access point to MaRDI’s digital research ecosystem. It connects researchers, data, models, algorithms, and software tools through a unified interface, promoting FAIR (Findable, Accessible, Interoperable, and Reusable) mathematical research. What does it do?It brings together all MaRDI services - such as the Knowledge Graph, MaRDMO, and MaRDIFlow - into a coherent platform. The portal enables discovery, documentation, and integration of mathematical research data and workflows. How can it be used?
Maintained by: TA5
|
 |
MaRDI Knowledge Graph APIAPI (stable version)
MaRDI Knowledge Graph API: Programmatic Access to Mathematical Research Data The MaRDI Knowledge Graph API provides structured, machine-readable access to interconnected mathematical research data. It empowers developers, data scientists, and digital library providers to integrate MaRDI data into applications and analytical environments. What does it do?It exposes the contents of the MaRDI Knowledge Graph via RESTful endpoints, allowing retrieval of entities, metadata, and relations between mathematical objects such as models, algorithms, and datasets. How can it be used?
Maintained by: TA5
|
 |
MediaWiki Math Rendering ExtensionTool/Application (stable version)
MediaWiki Math Rendering Extension: High-Quality and Accessible Mathematical Display The MediaWiki Math Rendering Extension enhances the display of mathematical expressions across the web. Upgraded by the MaRDI team, it introduces a new HTML5-based rendering mode that improves performance, accessibility, and visual consistency across devices and browsers. What does it do?It converts TeX input into modern HTML5 and MathML, ensuring that formulas are rendered quickly and accurately while remaining accessible to assistive technologies. The improvements made by MaRDI are now deployed across Wikipedia and numerous scientific wikis. How can it be used?
Maintained by: TA5
|
|
|
Software ReviewingSupport/Consulting (stable version)
MaRDI makes an effort to make software reviewing a normal part of mathematical peer reviewing. In order to do this we offer to review papers with code ourselves, write guidelines on how to write and review software and hold talks on software reviewing. What services do we provide?
Where to find the guidelines?Maintained by: TA1
|
 |
OpenMLDatabase (stable version)
OpenMLOpenML is a open, community-driven platform for sharing machine learning research data in a FAIR way. It enables users to discover, share, and organize datasets, tasks, models, and experiment results via a web interface and APIs. It integrates with major ML toolchains in Python, R, and other languages. While the main platform is developed by an open source community, with the center being located in TU Eindhoven, TA3 is responsible for maintaining the R integration and has also shared various curated benchmarking suites on the platform. What does it do?
How can I use it?
Maintained by: TA3
|
|
|
OpenMLDatabase (stable version)
OpenMLOpenML is a open, community-driven platform for sharing machine learning research data in a FAIR way. It enables users to discover, share, and organize datasets, tasks, models, and experiment results via a web interface and APIs. It integrates with major ML toolchains in Python, R, and other languages. While the main platform is developed by an open source community, with the center being located in TU Eindhoven, TA3 is responsible for maintaining the R integration and has also shared various curated benchmarking suites on the platform. What does it do?
How can I use it?
Maintained by: OpenML Team
|
