CORD-19 Demos and Resources

The community is creating many engaging experiences to explore the CORD-19 dataset. This is a place to list your project so others can discover it!

  • NLP-based Semantic Visualization of Covid-19 Dataset
    Over the past year, Brandeis has been developing semantic analysis tools for identifying and interpreting information from natural language text. Recently, we have applied these semantic data models for data visualization of complex networks of relations. We believe such methods might be useful for knowledge discovery and novel inferencing and generalization in the current drive to understand the underlying cause and spread of Covid-19. We present our initial analysis of two datasets: Blender’s Cord-19 dataset analysis; and Harvard INDRA’s Protein-protein-causal-analysis dataset. We also present our LAPPS Grid Covid-QA natural language query system for the CORD-19 literature, AskMe.

  • Doing a review on the CORD-19 dataset
    We took the metadata and put it in a public review in our systematic review system Rayyan. One can go there and export the data and then use the exported file to create a new review in Rayyan (many of our users are already doing that). You can work with multiple collaborators in a blinded or unblinded mode while chatting with them. More about Rayyan can be found here

  • Covid19 Literature Monkey
    A personalized semantic search engine and more. By Yale University Section for Pulmonary, Critical Care and Sleep Medicine

  • Covidex
    A CORD-19 Search Engine using NLP and IR Components by the University of Waterloo and NYU

  • AUEB NLP Group Covid-19 Search Engine
    An Experimental Document and Snippet Retrieval Search Engine for CORD-19 by the AUEB’s NLP Group

  • Covid Search
    A CORD-19 Search engine from Sinequa

  • Covidexplorer
    A CORD-19 search and visualization engine to understand current state-of-the-art COVID-19 research using NLP by the IIT Gandhinagar’s Lingo Group

  • Covid AI-powered Search
    A CORD-19 Search Engine and Knowledge Graph powered by Curiosity

  • CORD-19 Search
    A CORD-19 Search Engine powered by Vespa

  • Covid-19 Search
    A CORD-19 Search Engine powered by Azure Cognitive Search

  • Covid 19 Corpus
    The CORD-19 dataset loaded into the Sketch Engine

  • covidsearch
    Real-time question answering on 31K COVID-19 related articles by the DMIS Lab of Korea University

  • Coronavirus Research Repository
    The Elsevier Coronavirus Research Repository, with scholarly articles on COVID-19, SARS, MERS and other coronaviruses

  • [CovidSeer] ( - Multi Faceted CORD-19 Search Engine based on ElasticSearch by the College of Information Sciences and Technology in Penn State University


  • DOC Search
    Semantic search of CORD-19, PubMed, ClinicalTrialsgov, WHO, FDA, news, and more. Real-time Rapid Reviews with full-text PICO annotation, interactive visuals, RCT/RWE tagging and more.

  • Ludwig Initiative Against COVID-19
    LIA COVID-19 is a search engine that helps researchers find scientific COVID-19 related content and produce scientific breakthroughs in English by Ludwig.


  • COVID-19 Concept Embeddings; vector representations for SNOMED concepts and terms trained on CORD-19.

  • An Innovative Search and Discovery Engine

  • - open fulltext search from the Internet Archive, built on existing open scholarly web catalog

  • COVID-19 Knowledge Discovery Platform provides a unique search on gene and drug targets using proprietary knowledge mapping technology from Epistemic AI.

  • Kahun’s COVID-19 knowledge graph is freely available here: [ ]

  • CORD-19 Data Evaluation shows evaluations of the structure and uniformity of the dataset, to highlight any idiosyncrasies in the data an understand any gaps to be addressed to make it fit for a particular purpose.

  • Topic Forest is an Exploration Engine that uses interactive concept hierarchies to help researchers explore large documents sets. I’ve applied it to publications on Covid-19 and the general Coronavirus family. Hopefully researchers will find this useful. Any feedback to help improve the prototype is highly appreciated.

  • tmCOVID: a bioconcept extraction and summarization tool for COVID-19 literature in PubMed and PMC. A third database to support CORD-19 is under development.

  • CORD-19 Search Platform: A free platform to assist clinical and scientific researchers and public health authorities in their COVID-19 response efforts. We look forward to your feedback! by Element AI

  • DISCOVID: Discover relevant publications in CORD-19 with our user-friendly search and recommendation engine! Our approach combines a keyword search with a topic model, so you can easily find what you are looking for, and also explore related work to spark new insights.

  • Observable COVID-19 Pre-print and Review Dashboard: This Observable notebook pulls in the latest Biorxiv/Medrxiv preprints, links with reviews from Mount Sinai immunologists, and clusters papers/terms using Clustergrammer2

  • – a Q&A system to ask questions related to COVID-19.



  • Oumy: Explore CORD-19 papers much faster via query-by-example - simply click on any text and Oumy will find you more of the same.

  • WellAI COVID-19 Research Tool: NLP medical relationship engine, not a search. Significantly saves time for COVID-19 researchers by creating a structured list of medical concepts with ranked probabilities. Besides covering every medical category on the list of UMLS semantic types, another difference with CoViz is the ability of the WellAI tool to understand multi-dimensional relationships between medical concepts.

  • VOSViewer interactive term map of the CORD-19 corpus: allows to explore key terms extracted from noun phrases with a variety of overlays (clustering, year of publication, received citations, number of tweets, etc.). Refer to this pre-print for a full analysis.

  • CAiRE-Covid: A CORD-19 Question Answering Engine with Abstractive Summaries from the Centre for Artificial Intelligence Research (CAiRE), HKUST.

  • AVOBMAT-COVID-19: integration of the CORD-19 in a beta version of the AVOBMAT (Analysis and Visualization of Bibliographic Metadata and Texts) user-friendly text mining tool (e.g. faceted, advanced, proximity, command-line search, topic modeling, Ngram viewer, KWIC, network, frequency & significant text analysis). Provided by the University of Szeged.

  • COVID-19 Navigator: Semantic search of CORD-19, clinical trails, patents, medline abstracts and pubmed docs powered by IBM Watson Insights for Medical Literature and Watson Annotator for Clinical Data.

  • ResearchRabbit Citation Network Visualization Tool: ResearchRabbit enables researchers to search the CORD19 dataset through the search engine. Upon adding articles to the citation graph, researchers can quickly identify related papers based on their citation relationship. For an example, please see our GIF here!

  • database of promising substances against SARS-CoV-2 – Up-to-date mapping of COVID-19 treatment and vaccine development

  • is a free text mining tool for researching Covid-19 scientific papers and generating new insights. This text mining tool is based on the Azure Cognitive search (people, locations, organizations, keyphrases) and BERN a multi-type named entity recognition tool based on BioBERT that also supports normalization of extracted entities for identifying genes/proteins, drugs/chemicals, species, mutations, miRNAs and pathways as concepts. It also includes graphical representation of links between identified concepts.

  • CORD-19 Search: Leverage this AWS tool to query the COVID-19 Open Research Dataset (CORD-19), with natural language questions and answers from Amazon Kendra. CORD-19 Search includes data from CORD-19 which has been processed using Amazon Comprehend Medical.

  • VIDAR-19: Visualization of Diseases At Risk in CORD-19. Dashboard presenting automatic extractions from the CORD-19 corpus of all diseases referenced in the International Classification of Diseases (ICD-11) maintained by the WHO, as well as the diseases which might be considered as risk factors. Refer to this pre-print for more information. Developed by Yotta Conseil. Feedbacks are welcome!

  • COVID-Miner: A search engine for exploring coronavirus literature using clinical Natural Language Processing. The NLP pipeline identifies and standardizes clinical terms (diseases, medications, ICD codes) from research articles and ranks the articles based on degrees of matching with your query. The engine is updated bi-weekly and users can subscribe to receive latest updates. Developed by Cincinnati Children’s Hospital Medical Center.

  • CORD-19 Topic Browser: An interactive collection of concentrated topics for researchers. This is an alternative information seeking strategy to traditional query-based search. The CORD-19 dataset is presented in several views allowing users to see an overview of topics contained in the corpus, find articles related to a topic of interest, and visualize trends in topics over time. Scan the list of topics to find the ones that are most related to your areas of interest. Click a topic to view articles that most exemplify that topic, as well as a timeline. Developed by the MITRE Corporation. Topics are determined algorithmically using the MITRE Topic Modeling Neural Toolkit.



Thanks a lot for listing us!

Just noticed a misspell in our company name on

Sineque should be Sinequa

Would be great if you could update!

Thanks a lot,


@Adrien_Gabeur this post is a wiki and should be editable by anyone. Let me know if you’re unable to edit it directly and I’ll look into it. The “Edit” button is in the lower right of the post.


Thanks for your response!

I was referring to the active website page on CORD-19 explorer where you echo the list in the wiki.



Ah, thanks–sorry I misread. We’re going to take down that list on and redirect here :wink:

Kahun’s COVID-19 knowledge graph is freely available here: [ ]. We have developed a structured representation of COVID-19-related clinical knowledge based on standard ontologies (SNOMED-CT and LOINC). It carries causal and quantitative information about risk factors, clinical manifestations, lab abnormalities and imaging findings in patients with COVID-19. The knowledge graph is live and growing with each publication, and each piece of information presented is referenced.
Hope you find it useful.

1 Like

I created a Powershell script that runs a multiple keyword search function (up to 10 keywords) on the CORD-19 data set. Not “term1 OR term2” but “term1 AND term2” (up to 10 keywords) so that researchers can understand causal relationships:

COVID-19 Keyword Convergence Viewer

I’m not able to edit the wiki due to a message about new users only being allowed to post 2 links (?), but here is one more:

Thanks to others for sharing and for the CORD-19 dataset!

Hi @michaels,

We always get involved in Kaggle challenges, always internally.
Internally, we are working COVID open research questions and we did indexed/processed the dataset.
We made our search engine available publicly here,
We add info about our process and system here,
We will make our Jupyter Notebooks publicly available.
We are compiling taxonomies and ontologies from the research papers, we will make that public available as well.
We are also building a knowledge graph.
I also set aside a budge for this project for the rest of the year. The budge includes GPU’s for Kubernetes nodes where the Jobs and Notebooks are running, to speed up the NLP processing.

We published our own COVID dashboard here,
with some basic graphs for analysis.


Our team in New Jersey Institude of Technology has developed a website for searching the literature in COVID-19. It comes with a cluster map so you can easily navigate all papers just in one page. Specially, you can search for figures, tables in those papers.

Credicts: Jiaxin Du, Saiyang Na, Weimin Li, Xinyue Ye

Our team at has put up a semantic search engine on top of the CORD-19 dataset here

We would love to hear your feedback :smiley:

Hi there!
I’ve developed CoronaSearch, a multilingual search engine for the CORD-19 dataset making use of scispaCy (for preprocessing) and Universal Sentence Encoder. It is my first app (I usually focus on the modeling) so I would really love to have it listed and know your feedback.

Thanks for putting your resources and knowledge at the service of the common good.

Hello all!

We’ve developed a keyword-based search engine for the CORD-19 data set, it’s freely available here:

Thanks for the resources!

Hi Michael, Would it be possible to add Alexandria.Works to the list of CORD-19 exploration tools? Alexandria.Works is an NLP-based search engine that allows researchers to query the dataset in natural language (e.g.: “­ What is the function of the GM-CSF growth factor in the defense against viruses”). Access can be requested at Thanks. Kind regards, Marc

I have created , a Q&A system to ask questions related to COVID-19. You can ask questions in 16 languages and I am working on Neural Machine Translation by the moment to translate the answer to Spanish. So you could have answers in English and Spanish in the next release.

It’s running on my laptop by the moment so some latency is expected.


COVIDExplorer: by the Penn State’s Coronavirus-AI Project
Unsupervised Clustering, feature extraction and interactive visualizations to help researchers find the most relevant and urgent documents around key COVID-research questions.

The COVID-19 Research Pass (CRP) program, launched by ReadCube, provides direct access to over 26 million articles and is available to anyone studying or writing about COVID-19. The program expands on the efforts made by publishers opening up COVID-19 content, to provide researchers with access to a broader set of content needed in the course of their research. Rather than pre-filtering access to specific articles related to COVID-19, the CRP allows researchers to access any article from participating publishers they may need while studying COVID-19, including both open access and content behind paywalls.

​Access content from initial participating publishers: JAMA, Springer Nature, and Wiley​. ​

​You can find more about the program here:

CAiRE-Covid: A CORD-19 Question Answering Engine with Abstractive Summaries from the Centre for Artificial Intelligence Research (CAiRE), HKUST.


We integrated the CORD-19 in a beta version of the AVOBMAT (Analysis and Visualization of Bibliographic Metadata and Texts) user-friendly text mining tool (e.g. faceted, advanced, proximity, command-line search, topic modeling, Ngram viewer, KWIC, network, frequency & significant text analysis).


The tool was briefly introduced in an online NLP meet-up after the talk of Lucy Lu Wang & Kyle Lo of Allen AI who discussed the COVID-19 Open Research Dataset (CORD-19).

Hi all,

I shared a notebook on Kaggle:

How to boostrap a Covid-19 semantic search engine in 30 minutes

Using Python, Pandas, Gensim/Doc2Vec, UMAP & HDBSCAN…

Exemples of result:

This is preliminary work… I am working on an improved NLP model with better querying feature.

Feel free to reach at on Twitter: @bu2adaptive