Our FAQs
Updated in September 2024
Readers
- How to get the most out of Towards Data Science?
- How do I unsubscribe from your letters?
- Do I need to pay to read your publication?
Writers
Writing
- Do you publish only in English?
- Should I use my real name on my Medium profile?
- What topics should I write about?
- Can I use an AI tool to help with my writing?
- How can I improve my writing skills?
- What featured image should I choose for my article?
- Can I write a post about my book, event, podcast, conference, etc.?
- Can I write about my company’s products?
- I’ve been asked to include a few sentences about the potential risks or ethical concerns of my topic/dataset. What do you expect?
Submission and Publishing
- Author Terms and Conditions of Use
- Privacy Policy
- Should I submit my new article as a draft or as a published post?
- What happens when I submit my article to TDS?
- Where does my post appear once published?
- Can I send a post I have already published on Medium? ️
- Can I duplicate a post?
- My post has been refused, could you let me know why?
Data Sourcing
- How do I determine if a dataset license is okay?
- Can I use a dataset uploaded to a repository?
- Is data acquired through an API okay?
- What if there isn’t a copyright notice or license?
- What does an email asking for dataset permission include?
Advertising and Promotion
- Can I pay to get published?
- What kind of advertising or sponsored content do you allow?
- Can a company publish in TDS?
Others
Readers
How do I unsubscribe from your letters?
- Go to our publication’s homepage.
- Click the Following button on the publication’s header.
- Uncheck the Receive Letters to your inbox box.
You can also unsubscribe by clicking the Unsubscribe button at the bottom of the Letter you’ve received from TDS, or from your Publications page, accessible via the user icon menu.
Do I need to pay to read your publication?
No, we’ll always have free articles available to read for everyone interested in Data Science. However, as we are part of the Medium ecosystem, our authors can decide to lock their posts to members only. Please click here to learn more.
Writers
Do you publish only in English?
Yes.
Should I use my real name on my Medium profile?
Yes, please include your real name, photo, and bio. We don’t publish posts from an anonymous person. If you want us and your readers to trust you, you have to reveal who you are.
What topics should I write about?
TDS articles cover a wide and diverse terrain — from career advice for data professionals to data-powered approaches to climate change. You can explore some of our most popular topics on our tags page.
Authors approach these (and other) topics from many different angles. Some help beginners with getting started, create hands-on tutorials, or share their latest tips and tricks; others dive deep into complex problems or write about cutting-edge research.
If you’re not sure what to write about, choose something close to home. Walk us through your latest work project. Share some of the insights you’ve gained during your learning journey, or advice based on the highs and lows of your data science career.
Good ideas can often rise to the surface after you read something that resonates with you. For a healthy dose of inspiration, browse our Editors’ Picks for some of the best-crafted articles on TDS, or head over to our Author Spotlights, where we interview prominent TDS contributors about their work and their writing.
Another way to start brainstorming ideas is to think about your audience in more concrete ways. You can get a stronger sense of trending topics among data scientists by checking out the following links:
- Our archive lists the most-read articles on TDS.
- Medium tag pages can be ordered by trending posts.
- On GitHub’s Trending page, you can specify the programming language and date (for example: trending posts on Python since January 2022), or go directly to their data science topic page. You can use your findings to write about new and exciting Python libraries.
- Another powerful tool: Google Trends!
- Did you know Twitter has its very own data science topic page? It can give you good insights into ongoing and emerging conversations.
- Give Quora a try if you’d like to structure a post around an answer to a common question or problem.
Whether you’re a frequent contributor or an aspiring TDS author, we can’t wait to read your next article! If you haven’t published with us before, we’d be especially happy to hear from you.
Can I use an AI tool to help with my writing?
As our guidelines state, we don’t allow posts that contain AI-generated text, and we use both our editorial judgment and third-party software to detect instances of generated writing. We strongly believe in the value of human authors sharing their knowledge and experience in their own words.
Digital writing aids have existed for a very long time, and we have no problem with authors using these tools to check their spelling and grammar or to find the best word choices, to name a couple of common use cases. Many writers have relied on such tools for years, and they’re especially helpful for those working in their non-native language. If that’s the extent to which you’re using tools like ChatGPT, Grammarly, and similar ones, it’s unlikely to be an issue for us.
If you’re using AI tools for more substantive elements in the writing process, you should know that you’re no longer following our guidelines. For example, if you’ve delegated the tasks of ideating, outlining, structuring, or revising your article to ChatGPT, it means we can no longer draw a clear line between your own thoughts and ideas and the AI’s, so we can’t publish your post.
Here are a few examples to give you a clearer idea of our expectations:
- If you prompt an AI tool to generate a post (or sections of a post, like your introduction, abstract, or conclusion), which you then revise and edit, that goes against our guidelines. The same goes for asking an AI tool to summarize or synthesize content from other sources and using the results in your article.
- Writing a draft, feeding it into ChatGPT to revise it, and then using the polished output in your post is another type of usage we don’t allow.
- If you asked an AI tool to find a synonym you’re struggling to remember, that’s perfectly fine as long as, overall, you limit these types of occasional assists to local issues.
- If an AI-powered tool like Grammarly flagged your sentence and suggested a minor change (like switching tenses or replacing a word for better clarity), we have no issue with that.
- On the other hand, if your post is made up of entire passages polished through an AI tool’s suggestions, it becomes increasingly difficult to tell your own writing from the AI’s. Use these tools only for specific writing issues you’re facing, not as a replacement for your own agency and creativity as an author.
- If your post is about a new AI tool and includes some text it generated, that shouldn’t be an issue — just make sure that it’s clear for readers where the AI’s words begin and end (e.g. by using block quotes). Keep in mind that we’re generally less likely to publish posts that are heavy on AI outputs, so aim to use them judiciously.
How can I improve my writing skills?
The Internet is full of great resources to help improve writing. Yet, we find that many authors lack some basic skills, which prevent them from correctly getting their message across. Here is our selection of writing resources:
- How To Be A Better Writer: 6 Tips From Harvard’s Steven Pinker
- Steven Pinker’s Sense of Style
- Writing for an academic journal: 10 tips. The Guardian
- How to Write a Blog Post (That Actually Gets Read)
- 6 of the Best Pieces of Advice from Successful Writers
Before writing your article, you need to be very clear about what you are trying to achieve. Once you have formed the idea for your post, you should try to answer these questions:
- What is the goal or purpose of your article?
- What key question is your article trying to answer?
- Does the content answer the main question of your article?
- Does your article start strong and finish strong? (Did you create a mystery/issue and then resolve it?)
- Is each concept bared to its essentials, narrated in an orderly sequence and illustrated with analogies?
What featured image should I choose for my article?
A featured image sets the tone for your post, and entices potential readers to click through when they see it on our site, elsewhere on Medium, and on social media. So you want to choose an engaging, high-quality image, while avoiding the pitfalls of visual clickbait. That’s sometimes easier said than done, so here are a few pointers to steer you in the right direction.
Dos:
- If you don’t have a good photo or artwork of your own, we often recommend authors turn to sites like Pexels or Unsplash (the latter can be accessed directly from the Medium editor), where they can find good-quality, copyright-free photos. We know they can sometimes feel a bit safe or generic, but with millions of options, there’s a decent chance you’ll find one that works for your needs.
- Does your post have a strong visual component, like a well-designed chart, a map you generated, or some other data visualization you’re proud of? You can use it as your featured image!
- If you have some design experience of your own, why not create a custom featured image for your post? It can be the starting point for your own visual brand, which can set you apart from other contributors. Looking for inspiration? Check out TDS posts by Leonie Monigatti and Vyacheslav Efimov, who’ve been making their own featured images for a long time, each using their own signature style.
- You can also turn to AI image generators, which let your imagination (and prompt engineering skills) shine. Just make sure you follow our specific guidelines around using these tools — scroll down to item #10 to read all the details.
You’ll find a few more examples of successful featured-image choices below!
Don’ts:
- Never use images you haven’t created yourself or whose license and copyright status you can’t confirm. (In other words: they have to be licensed for commercial use). “It’s available online” does not, in fact, mean you can use it on TDS.
- If you’re using an image library like Unsplash, please avoid overly dramatic and/or stock-like photos, or ones that have been used a million times before. (You know the ones…) You can also modify these photos to make them your own, or to create a sleek composite or collage if you’re so inclined.
- We don’t allow the use of photos that show children or babies, with very few exceptions (like, say, in cases where your post deals specifically with a kid-related topic).
- Puppies, kittens, and bunnies are also best to avoid — they’re cute, but more often than not they’re the very essence of visual clickbait.
- Your image should complement your title and subtitle, not replicate them. Please avoid textual elements and overlays like titles, your name, taglines, and so on. If your image includes some text organically (for example: a label on a chart you’ve created for your post), that’s usually fine. Just keep in mind that you’re looking for a visual distillation of your topic, not designing a slide deck.
- While we’re at it: please avoid including any logos in your featured image, even if it’s for your own company or project.
- It’s tempting sometimes to be very literal — a photo of a snake for your Python tutorial, or an AI-generated multicolored llama for your introduction to Llama-2 — but you really don’t have to be! Use your imagination, and trust readers to respond to your smart visual choices even if they don’t spell out the core message of your article.
Picking a featured image that looks great and achieves its goals — namely, attracting new readers and setting the tone for those who are about to read your post — is more art than science, and might take some trial and error. That’s fine! We’re always here to help, so if you’re unsure what image to go with, you can always ask us for advice.
Likewise, if we think the image you’ve chosen isn’t serving your work as well as it could, we’ll let you know. We understand that authors are sometimes very attached to their visual choices, so we always aim to find common ground and go with an image that you’re happy with and that works within the broader context of our publication.
Here are a few more examples of successful featured-image choices:
Can I write a post about my book, event, podcast, conference, etc.?
It depends! We generally don’t publish event or book announcements, and avoid articles that serve primarily as a vehicle to publicize your latest podcast episode, YouTube video, or other media on another platform. For example, a post consisting of nothing more than a short introduction and a video embed or link to another site isn’t going to be a good fit for TDS.
If your article features one of these elements, we would still consider it for publication if it meets a few conditions:
- It offers enough value to readers — for example, a standalone book excerpt as part of your new-book launch, or a podcast episode embedded within a thoughtful discussion of a relevant data science or ML topic.
- It avoids overly promotional language and framing (e.g. multiple links and CTAs, gushing superlatives, and so on).
Submission and Publishing
Author Terms and Conditions of Use
Should I submit my new article as a draft or as a published post?
We’re glad you asked! We highly recommend submitting your new articles to TDS in draft form, rather than as published posts. Why? The TDS homepage and Medium’s Following feed are both organized in reverse-chronological order by date of first publication. That means the most recently published posts are shown near the top, so any article we receive as a draft will have greater reach and visibility at the moment we add it to our publication.
By contrast, a post you already published on your own — even just a couple of days before its appearance on TDS — will be easily lost amid all the newer posts we publish every day. In short: sharing a draft can increase your article’s potential readership.
Submitting a draft also helps our team share feedback and suggestions for improvement before publishing your article, which means no post-publication edits or corrections. It allows us to work with you on making the best-possible first impression on your audience.
If you’ve already published your post but still wish to share it with TDS, that’s fine — we’re always thrilled to consider your work. For future articles, though, sending us a draft is the way to go.
What happens when I submit my article to TDS?
Thank you so much for taking the time to submit your article to our team! We will review it as soon as we can.
If we believe that your article is excellent and ready to go, this is how you will be able to add your post to our publication. If “Towards Data Science” shows up after you click on “Add to publication” in the dropdown menu at the top of the page, that means we have added you as an author and are waiting for you to submit your article. Once you have submitted your article, it will be reviewed by an editor before a final decision is made.
If we think that your article is interesting but needs to be improved, someone from our team will provide you with feedback directly on your submitted Medium article.
Please note that we only respond to articles that were properly submitted using either our form or via an email that exactly follows the instructions listed here. We don’t respond to pitches or questions already answered in our FAQs or on our Contribute page. We also ignore articles that don’t comply with our rules.
If you haven’t heard from us within the next five working days, please carefully check the article you submitted to our team. See if you can now submit it directly to TDS and look for any private notes from us that you may have missed. You should also make sure to check your spam folder.
If you just can’t reach us, the best thing for you to do is submit your article to another publication. Although we’d love to, we can’t provide customized feedback to everyone because we simply receive too many submissions. You can learn more about our decision here and submit another post in a month.
Where does my post appear once published?
If your post is recent, it will appear on our homepage below “Latest”. Depending on your tags, your post will also appear on our different pages.
Before submitting your post make sure you read Medium’s Curation Guidelines. Even our best authors tend to make basic mistakes that disqualify their article from being distributed in Medium’s topics (Data Science, Machine Learning, Artificial Intelligence …), and from being featured on Towards Data Science.
If you want our team to mention you on Twitter, please ensure that your medium profile contains your Twitter account. It can take a few days to get your post featured on our social media.
Finally, your post will be searchable in our publication, appears in our archive, and can be recommended to our readers at the bottom of other articles written by your peers.
Can I send a post I have already published on Medium?
Yes, as long as your article is relevant to our readers, we are interested in adding it to TDS. Please do not duplicate your article on Medium! They do not allow posting duplicate content on their platform.
Can I duplicate a post?
Cross-posting content you have published elsewhere (not on Medium) is allowed provided it follows the Medium guidelines and that your post wasn’t already published on Medium.
TDS didn’t accept my post. Can you tell me why?
While we’d like to be able to leave feedback on all of the incredible submissions we receive, we are a small team, and we aren’t able to leave suggestions on all of the articles we review. However, there are some common reasons why we decline a new post.
We’re a Medium publication and we use the curation guidelines for every article we publish. This means that we look for articles with high-quality writing and content, articles that offer solid value to readers, and articles that are free of typos and grammatical errors. It also means that we decline articles that include disqualifying elements like clickbait, sponsored content, marketing/ads, press releases, or plagiarism. We recommend that you take a look at the guidelines!
We also have our own guidelines that you can find here. Our guidelines include topics like how your code should be displayed and how your images must be sourced and cited. Along with our guidelines, you’ll find advice and tips for getting published and ensuring that you are getting your message across to our audience.
If you read through the guidelines above carefully, it’s likely that you’ll find the reason(s) your article was rejected. However, here are some of the most common reasons we decline submissions:
- Format: it matters! If your code isn’t correctly formatted, your images are blurry and randomly sized, you have long walls of text or you’ve given each sentence its own paragraph, for example, we probably won’t accept your submission. Please take the time to make your article clear, helpful, and a pleasure to read.
For help with Grammar, you might want to try a free tool like grammarly.com. Hemingway can be another great resource for writing assistance. If you aren’t sure how to format mathematical equations on Medium, you might enjoy this article or the embed.fun tool. If you want examples of what editors at Towards Data Science are looking for, take a look at our editor’s picks!
- Images: if you use an image in your article, you need to verify that you have the right to use it and correctly cite its source. Please double-check your images and make sure you have the right to use every image you include in your submission, including ones that you’ve altered. You can learn more about this here (#10).
- Datasets: always ensure that you have the right to collect, analyze, and present the data you’re using. In your article, please add a citation to the dataset and state its license. Note that publishing in TDS requires a commercial license or permission from the dataset’s owner (yes, this includes web scraping too). You can learn more about our data policy here (#11).
- Clickbait and listicles: we aren’t interested in publishing clickbait, listicles, or basic content and we’ll probably decline your article if you’re giving widely available information without an original point of view. We’re interested in what makes your article different from other articles on your topic. Do your research and make sure that you have something original to say.
- Topic choice: while we make our editorial decisions on a case-by-case basis, we’re very unlikely to publish posts that focus on financial or investment advice, medical advice, and betting and gambling. We also don’t publish posts that promote questionable or unethical uses of data and AI. Thanks for keeping this in mind when you select projects and datasets to write on.
- Marketing content and sponsored posts: marketing content, advertising, and sponsored posts aren’t allowed under Medium’s curation guidelines. Since we use Medium’s curation guidelines for everything we publish, we often decline content that appears to be marketing.
- Company accounts and anonymous accounts: we currently only publish articles written by individuals. We don’t accept articles that are written by a company account or an anonymous account.
- Technical tutorials with minimal explanation: we love that you want to share your techniques with our readers. However, please keep your audience in mind! If you’re teaching readers how to build or fix something, it’s likely they don’t already know how to do it. Please make sure you provide sufficiently clear information in your own words.
- A personal project using common techniques: if you’ve built an amazing project, we want to hear about it! However, if you’re using basic techniques on a popular dataset or you’re simply going over steps from a course or tutorial without adding anything new, please wait to submit your article. Keep working on your project and, when you have something unique to write about, we’d love to share it with our readers.
- Unrelated Content: We are a publication about data science. If you’re writing about topics that aren’t directly related to data science, machine learning, programming, data visualization, and artificial intelligence, it’s unlikely that we’ll accept your submission.
In the majority of cases, we don’t decline articles for a single reason; our decision is usually based on a combination of several factors. If the only blockers we see are small, fixable issues, we let authors know and aim to resolve these together so we can proceed with publishing the article in question. When we decide not to accept a post, it’s often because the level of revision necessary requires too much heavy lifting and isn’t sustainable for authors and for TDS editors.
It sometimes takes authors several tries before we accept their work for publication on TDS; if you decide to give it another shot but you’re not sure how to move forward in a productive way, here are a few pointers:
- Take another close look at our guidelines. Yes, we know it’s a bit of a long slog, but this page is as detailed as it is for a reason: to help you put your best foot forward when you submit your next article.
- If you’re struggling to come up with an interesting topic after we declined your previous post, our FAQ offers quite a few ideas to get your brainstorming process going.
- Browse our Editors’ Picks to get a stronger sense of the posts we consider outstanding. There’s no need to emulate them or to try to create a close variation of an existing article; on the contrary — that’s unlikely to work. Instead, try to get a feel for the kinds of formats, topics, and writing styles that tend to do well on TDS.
Can I write about my company’s products?
We love publishing posts from data science professionals with industry experience. If your post revolves around a product or project you’ve worked on, you’re welcome to submit it to TDS as long as its primary purpose is to educate and inform, not to sell or promote.
Please note that we’re less likely to publish product-centered posts from authors who have recently written on similar topics, or whose posts focus primarily on their product/company.
A few pointers:
- Disclose your affiliation with the company or project in question, and make it clear whether the latter is open source or not.
- Avoid using your company’s logo, especially in the featured image — this can make your post feel too much like an ad and drive readers away.
- Describing the benefits of your product is fine if it’s relevant for the topic at hand, but think of your readers as peers, not as potential customers. Avoid marketing terms (“limited-time offer,” “special promotion”) and gushing superlatives (“easiest,” “fastest,” “best-in-class”).
- Whether you’re talking about your everyday work or a new feature release, make sure readers can learn something new even if they never use your product. For example, you could discuss a difficult technical challenge your team overcame, new workflows you’ve created to build your product, or how your work leverages recent advances in AI or machine learning.
- We typically decline posts that replicate product documentation or support pages, as well as those whose purpose is to walk readers through setting up and using the product in question.
I’ve been asked to include a few sentences about the potential risks or ethical concerns of my topic/dataset. What do you expect?
If we’ve asked you to reflect on the ethical or safety stakes of your project, it’s likely that you’ve written about topics such as facial recognition, tracking, or detecting identifying personal markers (age, ethnicity, and gender, among others), or you’re using a dataset that might raise ethical concerns.
As the ethical and risk issues are inherent to the topic and the dataset, there’s no one-size-fits-all solution. So, we’ve prepared some questions you might answer to create a discussion appropriate for your article. We’ve also included some published articles where the author has done a good job of discussing risks or ethical concerns.
Datasets
- Is it important to your article? Why/how?
- Why did you choose this particular dataset?
- Have you taken care to protect personal details in output shown in the article and any shared GitHub repositories?
- Have you looked into potential biases within the data you’re using? What did you do to address them?
Topic
- Have you reflected on potential risks or ethical concerns surrounding your project?
- How does your methodology relate to the risk or ethical concern?
- Have you looked at existing work around ethical concerns related to your topic? Have you incorporated these perspectives into your article?
- Have you taken steps to address the risks and/or ethical concerns?
- Is there future work associated with your topic that might mitigate the risk?
Published articles with ethics/risk discussion
Check your biases by Veronica Villa (ethical/bias risk)
Train a Model to Detect Breast MRI tumors with Deep Learning and PyTorch by Nick Konz (medical risk)
Data Sourcing
How do I determine if a dataset license is okay?
It would be great if all dataset licenses were easy to find, had the same language, and were easy to read. Sadly, this isn’t the case. Often it takes patience and perseverance to determine whether you have the right to use a dataset.
Here are some key points:
- Make sure that you’re actually on the site of the original source of the data, not someone’s copy.
- Some sites have the license information in plain view on the main page or included in information about the data. Other times, you’ll have to dig a little deeper and search the FAQ, Terms and Conditions, and Support pages. If all else fails, it’s best to contact the owner for permission. Please remember to forward a copy of the permission to us at publication@towardsdatascience.com.
- Keep an eye out for the word “commercial” in the allowed usage described in the license. Towards Data Science is a commercial publication, and commercial use must be permitted.
- Although not standard, many sites have adopted Creative Commons licensing, which makes interpreting the license simpler. These are a few license codes that allow commercial use:
Careful! There are two similar codes that do not allow for commercial use, which means these datasets can’t be used in TDS articles:
- CC BY-NC — Allows non-commercial use
- CC BY-NC-SA 4 — Allows non-commercial use with attribution
5. Licensing that applies to program source code, for example, MIT and Apache, doesn’t also automatically cover, or imply a right to use a related dataset.
6. A license that limits use to “research and educational purposes” presents a problem. Although you may have done research in your article, or are presenting material for educational purposes, you are publishing in a commercial publication. Generally research and education applies only to work published in academic journals, college/university courses, and conference proceedings.
7. Books, screenplays, magazines, and newspaper articles are covered by copyright. So, those Harry Potter scripts you’d like to use? They’re off-limits unless you obtain permission. However, older publications may be in the public domain and free to use for your article. Copyright generally lasts the life of the author plus 50 years (Canada) to 70 years (European Union, United States). There are more rules around copyright, and it’s best to consult with the Copyright Act of the country where the work was first published to determine if copyright has expired and you are free to use the document. When in doubt, ask for permission!
Can I use a dataset uploaded to a repository?
There’s no easy answer to this one because anyone may upload data to a repository, and each dataset has its own license. It’s important that you read the supplied information carefully to determine where the data came from and the license that applies to it, before using it in your article.
Many times datasets are created from scraped data and then given a public domain license (by the party who scraped it) in the mistaken belief that the repository owner permitted making the data/program available. Remember that only the original data owner can grant permission.
If you are participating in a data science competition or hackathon and plan to share your progress or solution on TDS, please make sure that their rules allow you to share the data. Some datasets are intended to be used only for the competition or hackathon, and cannot be shared in an article published with TDS.
Is data acquired through an API okay?
If you plan to use an API to create your own dataset, please be absolutely certain that you have the right to use the API (and any data acquired via the API) before submitting your article to TDS. Check the API or the website’s terms of use to see if they allow commercial use, and read any restrictions carefully.
What if there isn’t a copyright notice or license?
First, it is important to note that the lack of a notice or license doesn’t mean that it’s free to use. Sometimes, the data is covered under an umbrella organization, for example, a city’s utility department might be covered under the city’s general data sharing policy.
Other times, the owner has simply neglected to post a notice. In these cases, we suggest that you consider contacting the owner to ask for permission, rather than assume the data is fine to use.
What does an email asking for dataset permission include?
An email sent to a dataset owner asking for permission doesn’t have to be lengthy. It should include information about your intended article, yourself, and Towards Data Science. Here’s a sample you might find helpful:
I am writing an article about [topic] to be published in Towards Data Science, a publication available on Medium. I found your dataset here [link] and thought it was well aligned with my work. I am contacting you to ask for permission to use this dataset for my article. You can learn more about me here [link to profile, LinkedIn, website, etc.], and about Towards Data Science here.
Advertising and Promotion
Can I pay to get published?
No, we don’t allow sponsored posts.
What kind of advertising or sponsored content do you allow?
We don’t allow any advertising or sponsored posts.
Can a company publish in TDS?
Yes. But, we prefer to publish posts that are submitted from a Medium profile that belongs to a real person. Also, we don’t allow advertising and don’t propose writing services.
Others
Can I translate a post to another language?
You will need to contact the author of the post directly. Articles belong to authors so we can’t approve their translation for them.
Can I mention Towards Data Science in my Linkedin profile?
Yes, you can mention that you are an independent contributing author to the TDS publication. Also, it’s good practice to add the link to your contributions to the TDS publication by including the link to your Medium profile.
How to contact you?
You can contact us here.
