83 Free Datasets for Your Next Data Science Project

Data science projects start with quality data. You’ve heard it before: garbage in, garbage out. Fortunately, there’s a wealth of free datasets online, from governmental and economic data to narrower topics like MLB stats and video game sales.

Whether you’re putting together a data science project to land a job or just want to brush up on your SQL or data analyst skills, we’ve selected some of our favorite sources of free data and useful datasets you can use for your next project.

• Dataset Search Engines
• Machine Learning Datasets
• Data Visualization Datasets
• EDA & Analytics Datasets
• NLP and Sentiment Analysis Datasets
• Computer Vision Datasets

You can collect data from just about everywhere, from Wikipedia to your own personal Facebook data. But if you don’t know what kind of data you are looking for, these dataset search engines and data repositories are the places to start:

1. Google Dataset Search

Google’s data search engine is useful for finding datasets in a particular niche. This is a great starting point for both paid and free datasets from top sources around the web. Other useful Google sources are Google Trends and Google’s Public Data Directory.

2. Data.gov

Find all of the U.S. government’s free and open datasets here. This is a rich source for public economic data—like housing, wages, and inflation—as well as education, health, agriculture, and census data. With more than 300,000 datasets available, this repository is extremely helpful.

3. FiveThirtyEight

FiveThirtyEight might be best known for its data journalism. Fortunately, the site also makes most of the data it uses in its reporting open to the public. This is a great source for a wide range of data with a focus on politics, sports and culture.

4. Kaggle

Kaggle is one of the most popular communities for data scientists, and the site’s user-published datasets are great for self-guided machine learning or analysis projects. You’ll find a wide range of data from movie reviews to customer sales data and, fortunately, most have some of the preprocessing done. This is a useful source of data for sentiment analysis projects as well as data analysis and visualization projects.

5. UCI Machine Learning Repository

Check out the University of California Irvine’s repository, which features nearly 500 public datasets. This is a great source for clean, ready-to-model data in a wide range of niches from a dataset of chickenpox cases, to bank marketing data.

6. GitHub’s Awesome-Public-Datasets

This regularly updated library of datasets is a great place to start. The data is organized by category, with options like machine learning and software, and you’ll find quick links to sources.

7. Amazon Web Service Open Data Registry

Amazon’s registry provides public access to data from a range of organizations from the 1000 Genomes Project to NASA. You’ll also find helpful usage examples for many of the datasets, as well as project links for various organizations and groups.

8. Pew Internet

The Pew Research Center’s data repository focuses mainly on culture and media. In particular, you’ll find datasets and surveys covering media consumption, social media use, and demographic trends like this 2018 Twitter Survey.

9. data.world

data.world calls itself a “collaborative data community,” and the site has built a dedicated audience of data scientists who have collaborated on projects like social bot detection and data journalism. You’ll find datasets in a range of categories from crime, to Twitter.

10. COVID datasets

There’s a plethora of regularly updated public COVID data available online. Some of the best sources include CDC COVID Data Tracker and Our World In Data. For more niche projects, try the Coronavirus Tweets Database, featuring more than 1 billion Tweets, as well as The Marshall Project’s COVID cases in prisons datasets.

11. WHO Health Statistics

If you’re looking for healthcare data, start with the WHO’s Global Health Observation repository. The platform features a variety of health-related statistics such as HIV/AIDs, vaccination rates, and malaria. If you want to build a machine learning health project, this is the source to utilize

12. Academic Torrents

Academic Torrents is a database for large-scale datasets for research projects. The data is shared by researchers, and there’s a variety of interesting sources, including the classic Enron email dataset or the annotated New York Times text corpus, which contains 1.8 million articles.

13. Nasdaq Data Link

For FinTech machine learning projects, you’ll find a variety of finance-related datasets on Nasdaq Data Link. The site features both paid and free data. Some free datasets of note include Zillow Real Estate Data and Federal Reserve Economic Data. To access the site’s free datasets, you’ll need to create an account to access the 20+ free sources. However, there are numerous premium datasets available as well. This is a great data source for a real estate data science project.

14. Data Is Plural

Each week, Jeremy Singer-Vine compiles a newsletter of useful and curious datasets. Since 2015, he’s published more than 300 newsletters, and you can access the full archive on his site. The latest edition featured a dataset of restaurant “chainness” as well as an energy-demand dataset. If you’re interested in data, you should subscribe to the weekly newsletter.

15. NASA Earth Data

NASA’s Earth Science Data Systems Program is a repository of the organization’s Earth science data. You’ll find datasets related to sea level rise, wildfire frequency, and tropical storms, among other interesting earth sciences insights. See the Data Pathfinders tool to learn how to source and access science datasets.

16. Datahub.io

Datahub is a wonderful source for open data. Jump to the Collections tab to browse datasets in various categories covering everything from climate change to football. You can also use the Find Data tool to search for relevant datasets.

17. Federal Bureau of Investigation Crime Data Explorer

For open data related to crime and law enforcement, this is one of the best sources for U.S. crime statistics. You can search by state or jump into various datasets including use of force or arrests.

Looking for a specific type of data for your project? We’re surfacing some of the most useful datasets to use in a wide range of data science projects from data analysis and visualization to machine learning and data cleaning.

Whether you want to work with predictions or classification, these datasets are both interesting and helpful for machine learning projects. The data is relatively clean and lends nicely to machine learning, e.g., plenty of variables that can help make predictions for the target column.

1. DoorDash Machine Learning Take-Home

This dataset, used in DoorDash data science take-homes, features user and transaction data and asks you to build a model to predict the time for deliveries.

2. Stroke Prediction Dataset

Build a stroke prediction model with this handy dataset. The CVS contains patient information like gender, age, pre-existing conditions, and smoking status that can help you build a model.

3. Swish Analytics Take-Home Data

The dataset used in this take-home contains all pitches from the 2011 MLB season and asks you to build a model to predict the probability of a type of pitch.

4. Divorce Predictors Dataset

This dataset from the UCI Machine Learning Repository contains survey data from married couples. Use the data to identify predictive indicators of divorce or to build a divorce prediction model.

5. RiskLens Machine Learning Take-Home

The data used in this take-home includes cyber security threats and events faced in the healthcare industry. You can use this data to build a healthcare risk assessment model.

6. January Flight Delay Prediction Dataset

With data from more than 400,000 flights in January 2019 and January 2020, this data from the Bureau of Transportation is well suited for building a model for winter flight delays. This is a useful dataset for a regression data science project.

7. Twitter User Gender Classification

Can you predict gender from a Twitter user’s profile and tweets? Build models to answer that question with this dataset, which contains information on more than 20,000 Twitter users.

8. Mushroom Classification Dataset

This classic dataset from UCI is a great source for a classification data science project. One great project idea is to build a model to identify classifiers for poisonous mushrooms.

9. Credit Card Approval Prediction Dataset

This is a great dataset for a financial prediction model. Use the data to understand if an applicant is “good” or “bad.”

10. Water Quality Dataset

Use water quality metrics from nearly 4,000 bodies of water to predict whether the water is safe for consumption or not.

11. New Yorker Caption Contest Dataset

This dataset features ratings for submitted New Yorker caption contest entries. Get some ideas for using this data here.

12. MovieLens Ratings Database

There are numerous movie rating datasets available here, including one featuring 25 million ratings, making it a great source for building a recommendation engine.

13. San Francisco Restaurant Health Scores Data

This open dataset covers health inspection scores for restaurants in San Francisco. One option is to use the data to build a model to predict a restaurant’s repeat health scores.

14. Capital Bikeshare Dataset

This dataset is useful for demand forecasting projects. It contains bike rental data from a bike sharing program, including travel duration, departure and arrival locations, and weather data. This dataset is similar to one that’s used in the McKinsey data analytics take-home.

15. Cats vs Dogs Dataset

With more than 20,000 images of cats and dogs, this is one of the best datasets for beginner image classification projects. The data is already separated into training and testing datasets, and it’s already labeled.

Build data visualization projects with these helpful datasets. We looked at data that had potential for interesting visualizations as well as datasets that weren’t too messy or overly complex.

1. Supercell Data Analytics Take-home

This dataset includes revenue and sales data from Supercell and asks you to create a visualization of a single aspect of the data that you find important.

2. Twitter Edge Nodes Dataset

With more than 11 million nodes and 85 million edges, this dataset is useful for building graphical relationship models of Twitter users.

3. Hotel Booking Demand Data

This is a great dataset for visualizing hotel bookings. You’ll be able to build visualizations that answer questions like:

• When’s the best time of year to book?
• How long is the optimal stay length to receive the best rate?

4. Amazon Top 50 Bestselling Books 2009-2019

Design visualizations that show top authors, best-selling titles, and review ratings for the best-selling books on Amazon.

5. COVID Jobs Impact & Hiring Data

Visualize the impact COVID is having on hiring with this dataset from the Amazon Open Data Registry. It features regularly updated hiring data from 3+ million job organizations.

6. Latest Polls from FiveThirtyEight

If you’re interested in political visualizations, FiveThirtyEight is one of the best data sources. Its updated polling data is great for visualizing averages and polling movements.

7. U.S. International Trade in Goods and Services 1960-Present

Build charts to visualize the United State’s international trade, including top imports, top exports, and annual trade balances.

8. Euro Exchange Rates

This dataset is useful for Matplotlib visualizations. You can create visualizations of exchange rates and currency valuations over time. The dataset features more than 20 years of daily exchange rate data.

9. San Francisco Public Library Usage Data

There are more than 400,000 records in this dataset, featuring daily circulation for the San Francisco library system. You can build visualizations related to new acquisitions, most checked out authors, most checked out titles, etc.

10. Trending YouTube Videos Data

This Kaggle dataset features daily trending video data from YouTube. Trending videos aren’t necessarily the most watched, but are generally the most interacted-with videos. Visualizations include most popular videos of the year or month or most trending videos by artist/creator.

11. World Unemployment Dataset

This dataset features more than 31 years of unemployment for numerous countries around the world. There are a wide range of visualizations you can create, including comparisons of countries, unemployment rates over time, or countries with the lowest unemployment.

12. NYC Subway Entries and Exits Data

This dataset originated on New York State Open Data and features information by station, line, location, etc. You can use this dataset to build visualizations of popular lines or subway maps.

13. 2021 Tokyo Olympics Dataset

This dataset contains information on more than 10,000 athletes in 40+ sports, and it’s a great source of building country medal count visualizations. There’s also coaching data, so you can add medal information by coach.

Say you want to take a big dataset and investigate. As you start to dive into the data, you begin to discover patterns, trends, and anomalies. These datasets are perfect for exploratory data analysis projects because they contain large amounts of mostly clean data.

1. Airbnb Analytics Take-Home

This Airbnb dataset, which is part of a sample data analytics take-home, contains user information for bookings in Brazil.

2. Netflix Original Films & IMDB Scores

A fun dataset to explore, and great for beginners, this features all of the Netflix original movies up to June 1, 2020 and their corresponding IMDb scores.

3. Stripe Analytics Take-Home

This Stripe dataset, which features product usage and marketing data, is perfect for diving into marketing and product analytics to determine how well a product is performing.

4. Superstore Sales Dataset

Featuring 4 years of data from a superstore, this dataset is perfect for analyzing and identifying trends, as well as sales forecasting.

5. Home Depot Product Analysis Take-Home

This sample dataset from a Home Depot data science take-home can be used to produce a gross sales forecast for a new product launch.

6. Marketing Analytics Data

This dataset is made up of mock marketing analytics data used by master’s in business analytics students. A great source for a marketing analytics project.

7. Animal Shelter Analytics Data

This is a great dataset for surfacing actionable insights for animal shelters, including what factors led to successful outcomes for the animals.

8. Why Americans Don’t Vote: Non-Voter Data

Another FiveThirtyEight dataset, this one features survey data from non-voters in the U.S. A few project ideas are identifying key factors that result in non-voting or building a voting likeliness model.

9. Website Crawling Data

A sprawling dataset from Amazon, the Common Crawl corpus features crawling data from billions of websites. Check out the Example Projects page for ideas.

10. European Soccer Dataset

This is a useful dataset for a sports analytics project. Featuring data on more than 20,000 matches, as well as individual stats from 2008 to 2016, this is great for exploratory data analysis projects on line-ups, team stats, wins, and individual player stats.

11. Open Food Data

This large-scale dataset, which was originally developed in 2018, features product information for more than 600,000 food items. Data includes allergens, ingredients, and nutrition facts, and there are a wide range of data analytics projects you can do with it.

12. Social Media Influence Survey Data

This is a useful marketing analytics dataset that features survey data from 2,500+ millennials. The survey asked which social platform has influenced your online shopping the most.

13. Google Analytics Dataset

This dataset features Google Analytics metrics from Austin, TX’s website. This is a great dataset for working in Google Analytics or analyzing website traffic.

14. Uber Pickup Data for NYC

This dataset features more than 20 million metrics on Uber pickups in NYC in 2014 and 2015. This is great for an exploratory data analysis or analytics project, and you can gather insights into popular pickup locations, common trip routes, and the locations with the longest pickups.

15. Marketing Analytics Dataset

This dataset is a great source for a campaign budget optimization project or for diving into an exploratory data analysis for marketing analytics projects.

There are plenty of large datasets great for sentiment analysis and natural language processing (NLP) projects. Data like movie reviews, tweets, Reddit comments, and more are all great for these types of projects.

1. Nielsen IQ NLP Take-Home

This take-home provides a dataset of human vs bot texts and asks you to build a classification model to correctly label the data.

2. Reddit Vaccine Myths Dataset

An interesting dataset for performing sentiment or text analysis, this features thousands of posts from the popular subreddit Vaccine Myths.

3. Wikibooks Dataset

There’s more than 270,000 book chapters in 12 languages in this dataset. It’s perfect for performing a wide range of NLP tasks like text parsing, text generation, or semantic analysis.

4. Spam Clickbait Headlines Catalog

Featuring 3+ million headlines from the now-defunct tabloid The Examiner, this is a great place to start an NLP news analysis Python project.

5. TripAdvisor Hotel Reviews

Explore thousands of hotel reviews from TripAdvisor and build semantic prediction or top clustering models.

6. A Million News Headlines

Another helpful medium source, this features headlines from nearly 20 years. It’s a great dataset for performing latent semantic analysis or latent dirichlet allocation tasks.

7. Disneyland Reviews Data

With more than 40,000 reviews from three Disneyland locations, this is a great data source for performing sentiment analysis.

8. Amazon Product Review Dataset

This dataset, which is a classic that was produced in 2009, features star ratings for numerous Amazon products.

9. Rotten Tomatoes Sentiment Dataset

The Stanford Sentiment Treebank contains more than 10,000 Rotten Tomatoes files and provides sentiment annotations on a 25-point scale.

10. Twitter Airline Sentiment Data

This dataset features thousands of airline reviews on Twitter from February 2015. The data has already been classified as positive, negative, or neutral, and in some instances, includes a reason for the negative tweet.

11. IMDB Sentiment Dataset

Featuring 25,000 movie reviews, you can use this dataset for a binary classification project or to analyze movie review ratings by title.

12. Enron Email Dataset

This is a classic NLP dataset that’s been studied and written about numerous times that’s great for text classification and analysis projects.

Whether you’re looking into an image recognition project or speech recognition project, these are great image and audio datasets that will help you practice your deep learning skills.

1. VoxCeleb Speech Corpus

The VoxCeleb large-scale dataset features audio-visual data from 7,000 speakers. It’s a great dataset for performing emotional recognition, speaker recognition, or talking face synthesis.

2. Face Mask Detection Database

There are about 900 images in this dataset of people wearing facemasks. You can use this to build models to detect if someone is wearing a mask, not wearing a mask, or wearing a mask improperly.

3. Unsplash Open Library

This rich visual-text dataset is loaded with helpful information. Use the photos for object detection. A bonus: there are millions of keywords and metadata you can use for exploratory data analysis projects as well.

4. CheXperts: Chest X-Rays from Stanford AIMI

This dataset from Stanford features 200,000+ chest radiographs. Build a model to detect pathologies and see how well your model performs against radiologists.

5. Pokemon Images and Types

There are thousands of images of Pokemon characters in this dataset. Use the data to build a prediction model to determine the type of Pokemon based on the image.

6. ImageNet Image Database

A classic image dataset from Stanford, you’ll find more than 14 million images here. This is one of the best datasets for performing object recognition tasks.

7. Stanford Dogs Dataset

Featuring more than 20,000 photos of dogs, this is a useful dataset for building classification models or a dog breed image classifier project.

8. Cityscapes Dataset

Featuring more than 5,000 images with fine annotations, as well as 20,000 images with coarse annotations, this is one of the best datasets for understanding urban street scenes at the pixel level.

9. Yale Face Database

This is a smaller dataset, featuring 165 images of 11 subjects. For each subject, there are images with various expressions and configurations, for example, “sleepy” or “without glasses.”

10. Fashion MNIST Dataset

Similar to the MNIST handwritten text dataset, this image set includes a training set of 60,000 images of articles of clothing along with a test set of 10,000 images. There are 10 classes for the dataset, as well as a label like “bag” or “trouser.” This is useful for testing machine learning models.

11. Indoor Scene Recognition

Image recognition indoors is more difficult, and this dataset, which features 15,000+ images of indoor scenes, is useful for building indoor recognition models.