Weeks of writing and editing have finally paid off, and your article is now published on Towards Data Science. You proudly announce it to your circle of friends and colleagues, showing them your article on our front page and sharing our Twitter and LinkedIn posts. You watch the claps and follows, expecting to see both grow into the thousands, as you’ve seen with some of our most popular Authors. But after an initial burst of interest, things stall. What happened?

Every newly published article gets its spotlight moment on our site and on our social media accounts. But to be truly successful, authors need to play a part in promoting their work. Many authors aren’t sure where to start, so we’re sharing our top tips around promotion. While they won’t necessarily make you a star TDS author out of the gate, they’ll put you on the road to success.

Networking

Even if you have yet to write a single word, it’s important to build your network because it’s tough to spread the word alone. Your work colleagues, friends, and family are a great place to start, but to truly grow your network, you need to reach out to strangers.

Social media accounts. This is the form of Promotion most familiar to authors. It’s a big topic, so we’ll be offering a separate post on it in our Writer’s Workshop series. Here we’ll mention that to increase your followers, you often have to follow others first. Follow other members of the TDS community, as well as leaders in your field.

Educational providers. If you are studying or have graduated from an educational program, connect with them on Twitter. Many actively support the promotional activities of their students and alumni. Some of these institutes will also offer opportunities to promote your work in their newsletters and publications, so they’re worth a deep dive.

User groups, professional associations, and conferences. This is an often overlooked opportunity to establish a network. If you’re a Python expert, for example, seek out Python forums. And don’t forget the software companies – many offer annual conferences and networking events where you’ll find like-minded colleagues.

Website. Some authors create their own websites, and although this can be effective, it takes work – and requires its own promotion because you have to attract people to the site. If you do decide to take this route, keep your site up-to-date, informative, and easy to navigate. Also remember to take advantage of your Medium profile biography to mention and link to your site.

Distribution lists. Gathering contact information from people who have expressed interest in your work is another popular networking tactic. And for good reason! These individuals like your work and are waiting for more, so they’re already supporting your efforts. We make it easy for you to build your distribution list, and we explain that and offer some great advice in a separate section below.

Other authors. It can be uncomfortable to approach someone you don’t know, but remember that many of them, especially other authors, have been in your shoes. They had to build their network from scratch too, and for that reason, you may be surprised to find that it’s easier than you think! Consider reaching out to authors who write similar topics, or simply add author mentions directly to your post. It’s important to use good judgment when involving other authors; chose the author(s) you mention carefully, and don’t go overboard.

Make sure your articles are found

Gone are the days when people eagerly anticipated their print copies of magazines and journals, all of which were read cover-to-cover. Instead, internet searches identify articles of interest, and you want to be in the top results.

Search engines use proprietary algorithms to determine what shows up first, and that makes search engine optimization (SEO) an important aspect of every published article. TDS authors benefit from the SEO boost of an established platform, as we are one of the most popular and most-read data science blogs in the world with +600K Medium followers. But there is more that can be done by authors themselves.

Basically, we want to feed the search engine what it needs to quickly establish that your article is an ideal match for an entered search. The fact that we don’t know exactly how Google uses an estimated 200 factors to rank content makes it important to pay attention to the factors we understand play a role.

Add keywords. Keywords are commonly searched terms, and including relevant ones in your article title and text can bolster visibility. It can be helpful to combine primary keywords such as machine learning or data science, with terms specific to your topic, such as visualization or deployment. If you’re not sure which keywords to include, Google Trends is a good place to start.

Place keywords effectively. Use keywords that relate to your content in your title, subtitle, and opening paragraphs. Ideally, a keyword should appear near the start rather than the end of your title and subtitle. The same strategy applies to your introduction, where you should try to include keywords in the first two or three sentences. The Search Engine Journal shares some good tips about keyword placement.

Use keywords wisely. Be natural about your keyword usage; many search engines actively watch for too many repeated keywords, known as keyword stuffing – and so do TDS editors. So, use keywords only where it makes sense, and consider related terms.

Don’t forget metadata. In the Story Settings for your post, you’ll find an SEO Settings section where you can view your title and description, which function as search engine meta description tags.

There are a few things to keep in mind when entering these settings:

• Your SEO title can only be changed before publication. Ideally, it will be between 40–50 characters and include keywords. While you can choose a longer title, please keep in mind that search engines often truncate titles to 60 characters. This means that the most important elements and keywords in your title should be at the beginning.
• A good SEO description is a tight, compelling summary of an article that is between 140–156 characters long and includes keywords. By default, the first paragraph of the post is used, so review this carefully to be sure it works. This setting can be changed after publication, but you’ll want to take care of this beforehand to get the most from your initial post promotional period.

Encourage readers to engage with your article

Faced with a long list of internet search results, readers can be picky about what they decide to dig into. So, in addition to a great title, subtitle, and good keywords, it’s important that you consider your opening paragraph and featured image.

Create a meaningful title. Create a title that’s short and designed to catch a reader’s attention. That doesn’t mean clickbait – to meet our guidelines, your title must reflect the content of your article.

Add detail with a subtitle. A good subtitle should expand on your title without repeating it, so that taken together, your title and subtitle reflect the contents of your article.

Use images to your advantage. A carefully selected image speaks volumes about the content. If your image carries a theme that closely reflects your post, it will immediately engage readers interested in the topic.

Open with your best. Polish your opening sentences, and be sure they reflect your writing style. This tells readers that you’re professional in your approach and sets the right expectations.

Make it count. Resist the temptation to fill your opening paragraph with information about yourself. Readers want to know what the article is about by scanning the first paragraph. If they’re viewing your article on a mobile device, that’s all they may see.

If you’d like to learn more about crafting great articles, watch for upcoming columns in our Writer’s Workshop series.

Reader Interest Tags

Take advantage of the reader interests tags on the Story Settings>Story Preview page when you’re publishing your article. These tags categorize articles published on TDS, and serve to help readers find topics of interest. You can add up to five, so a mix of high-level terms and more specific ones makes sense. TDS Editors sometimes edit your tag selection to ensure the most relevant ones are present. We work closely with Medium to ensure that the stories in TDS meet Medium’s editorial and distribution standards, and stories in TDS often see further distribution because of that.

Write more articles

Name recognition is important in promotion. Not only does it impact your search engine optimization (SEO), but readers are more likely to open an article written by someone they know or have heard about. As your writing reputation grows, you’ll find that doors open.

Create more content. Our best advice is to build a body of quality work. For many authors, this means writing more articles, but it might also include YouTube videos, workshops, or conference presentations.

Establish your niche. Many authors grow their reputation and followers by focusing on a specific niche. That may mean writing articles that share your expert experience, but it may also mean tutorials, or articles geared to beginners.

Keep up your momentum. Once you’ve gained followers, it’s important not to lose them. It’s easy to forget about an author, even one whose articles you thoroughly enjoy, if they’re out of sight for too long. It’s a little like a friend you haven’t seen for ten years.

Often authors who have written a great article struggle with what to do next. We offer some suggestions in our FAQ and we’ll be publishing a Writer’s Workshop column on the topic soon. For those looking for current topics and datasets, we are also hard at work on a new column called The Spark.

Distribution lists

We’ve already discussed the importance of networking and its value in building a contact list. But you have to move beyond people you know.

Medium provides an easy-to-use option that’s great for adding readers to your distribution list. Everyone who clicks on the follow button that appears with every story is automatically added to your private list. What’s great about this, is that these new contacts have seen your work and feel that they’d like to see more. It’s a little like finding gold!

If you collect contacts elsewhere, those email addresses can be imported into the same list. And just like that, you have a single central distribution list for sharing your latest story! Full details are available through Medium’s Email subscriptions and Your audience stats pages.

Some authors expand their use of a distribution list to share things like newsletters and tip sheets. They can be helpful, but also challenging. While someone may have signed up initially, perhaps to receive a tip sheet, there’s no guarantee they’ll open anything you send.

If you plan on going this route, take time to research what works.

Add value. If your newsletters or publications are chock full of great tips and advice, then people tend to be more engaged.

Don’t over-advertise. The fastest way to lose your carefully built following is to send out short newsletters that do nothing but promote your latest article. It’s fine to promote, but remember the first rule: add value. Consider how you can provide some additional code or techniques related to your article.

Share information about yourself

Readers connect not only with an article, but with the author themselves. This means that your Medium profile plays a role in promoting your articles.

We find that an author photo and full name helps build trust between authors and readers. But what you say about yourself is your first conversation with readers. It builds credibility and connection.

Introduce yourself. Let your readers know about your work, your specializations, or your interests. It can be a simple list or an elaborate statement. What’s most important is that readers learn a little about you.

Tell us more. Authors often overlook the About tab associated with their profile. Here, there’s loads of space for you to share everything from your professional pursuits to photos.

Improve your craft

This may seem like an odd bit of advice about promotion, but a well-written article shines, both with our readers and our editorial team, and it serves to spotlight an author. Think about your own reaction to a great post. You likely clapped for it, followed the author, and perhaps left a comment. What’s more important is that you may also have shared and talked about it – you promoted the article.

We do that here at TDS as well. We feature and highlight our best articles and authors, extending their reach.

Social media. In addition to the usual Twitter and LinkedIn posts, our best articles are shared on Facebook.

Editor’s Picks. Our editorial team picks their favorite articles for our Editor’s Picks page. In addition to receiving extra time in the spotlight, these top articles are tagged so readers can easily spot them.

Curated Features. Our top articles and authors are highlighted in our Variable newsletter and Monthly Editions.

Author Spotlights. We regularly invite our most engaging and prolific authors to participate in a Q&A that introduces them and their work to our community.

Maybe you wrote a great post and followed most of the advice above but still didn’t get the tens of thousands of views you were hoping for. That’s fine! Success takes time! If you continue to write strong, original content (that your first few readers love), eventually some posts will start breaking through the noise. You will then get more reads and engagement on your content as a wider audience discovers you and your work.

Discovered another good way to promote your articles?Want to know more about something mentioned here?

Share it in the comments! Your question or idea might be part of our next article.

The post The Art of Promotion appeared first on Towards Data Science.

In this article, I will share some tips on how to improve your Data Science thesis. Over the years, I have supervised my share of Data Science thesis projects, ranging from Big Four firms to local SMEs and from multinational banks to software consultancies. The academic program I am active typically involves internships, in which data is utilized to resolve a corporate problem – think designing decision-support dashboards, detecting financial anomalies with machine learning algorithms, or improving real-time parcel routing. Although educational programs, conventions and thesis requirements vary wildly, I hope to offer some common guidelines for any student currently working on a Data Science thesis.

The article offers five guidance points, but may effectively be summarized in a single line:

"Write for your reader, not for yourself."

Data Science is a complex field, and the myriad of algorithms, performance metrics and data structures is hard to fully grasp even for the most seasoned veteran. As such, your job as a writer is to help the reader as much as possible in digesting your research, guiding and clarifying wherever you can. Everyone can make matters more complicated, but to simplify them is the true test of your skill.

I. Intro, content, conclusion

Intro

Always lead with an intro that outlines what the reader can expect. The key is to make this intro specific. Don’t just mention that you will perform a literature review, collect data, model the problem – foreshadow what you will study, what data you collect, what structures or decisions you model. Academic Writing is the opposite of an exciting novel that prolongs the plot— academic readers do not like surprises. A proper introduction provides an framework that aids the reader in structuring the content that follows.

Content

The content will take the bulk of the page volume, and as such it is imperative to clearly structure it. Think in advance about the messages you want to get across. It is often helpful to first write the ‘skeleton’ of the text (e.g., the chapter theme, the section purpose, a bullet point per paragraph). Verify whether the messages have a logical sequence and form a cohesive story. Try to limit yourself to one key message per paragraph. Without a predefined structure and a thought-out message, content quickly collapses into a tangled mess of formulas, data structures and experimental results.

Conclusion

Always end your text (whether it is the thesis, a chapter, or even a single paragraph) with a clear conclusion or summary. Your audience will likely not read, remember or understand every detail you jot down. Of course, the length of the conclusion should be proportionate to what it is concluding. For the thesis itself you are looking at a full chapter, for a paragraph a single closing line suffices. Ending with strong closures is vital for your thesis.

II. Recap, interpret, explain

Truly reading and comprehending a thesis is a tough job; readers need all the help they can get. As a writer, you might be completely immersed in the topic for many weeks, but for your reader the thesis is likely one of many documents to skim through. In fact, few people will ever read your work cover to cover. As such, it is your job as a writer to aid the reader as much as possible. Recap that ω_t you last mentioned 10 pages ago. Interpret what that AUC of 0.7 actually means. Explain why you perform that t-test. Do not assume that the reader will figure out how to put the pieces together – make an active effort to guide your audience through your research.

III. Select solutions appropriate to the problem

The problem should always be leading in selecting the techniques to use. It is tempting to explore the Machine Learning hype of the week, but chances are it simply isn’t the best tool for the job. First (i) study your problem setting, (ii) define suitable research questions, (iii) set the requirements and restrictions, (iv) analyze your data sets, and (v) determine success criteria. Only when all that is done, you can truly make an informed decision on appropriate solution methods. In all fairness, thesis projects typically offer a bit more room for exploration than normally encountered – after all, sometimes companies just want to explore whether something new works or not. Nonetheless, always let the problem drive the solution method, not the other way around.

IV. Start broadly, end broadly

Data Science theses have a tendency to go into great depth on tiny aspects, e.g., finetuning hyperparameters or running scores of experiments. In itself, this is fine. However, if the problem you try to resolve was never clear to begin with (problem statement, context analysis) or your experimental results never link back to the original research motivation (conclusion, recommendations), you miss the chance to make a meaningful impact with your research. The following structures might be helpful to translate your work:

• Hourglass model: start broadly at the level of the corporate/societal problem, gradually zoom in at the technical level, then translate the results back to managerial insights.
• Double Diamond model: Alternate divergent- and convergent thinking, both in the research phase and in the design phase. Deliberately build in times to explore and times to focus.

The bulk of your work will likely be at the content level (data collection- and cleaning, modeling, parameter turning, experimentation). However, do not forget to first sketch the scene, and close with a compelling final act.

V. Define your key metrics

Try to identify the key metrics that capture the success of your research. For a balanced viewpoint, it is often necessary to report multiple metrics (precision, recall, AUC, F1 score, etc.). However, you should avoid providing merely an array of metrics absent a comprehensive interpretation. That 98.3% accuracy for your fraud detection model sounds great, but tells little about its practical usability. Your result table with hundreds of metrics is impressive, but can you capture its key message in one sentence? Are the higher precision yet lower recall an improvement compared to the base model? Vigorously explore your results from various angles, but ensure to distill the key outcomes for your Twitter summary.

Wrapping up

This article discussed five tips to help writing your Data Science thesis. The overarching principle is to always keep your reader in mind, and to go the distance in actively structuring, explaining and interpreting your research for the intended audience. The five tips might be summarized as follows:

• Intro, content, conclusion – Use a consistent structure at all levels of your text (thesis, chapter, section, paragraph), leading with an introductory outline or signal and ending with a conclusion or summary.
• Recap, interpret, explain – A successful thesis guides the reader through your research, providing helpful explanations to support your techniques and results.
• Select solutions appropriate to the problem – Ensure to thoroughly study the problem, context and expectations, before selecting a solution method that fits the nature of the assignment.
• Start broadly, end broadly – Going in-depth is perfectly fine, but don’t forget to (i) clearly outline the problem setting and (ii) translate your main findings into tangible insights.
• Define your key metrics – A single metric rarely suffices to capture the full depth of an analysis, but ultimately it is necessary to boil down your research to some digestible numbers.

The post Five Tips For Writing A Great Data Science Thesis appeared first on Towards Data Science.

In the final year of my undergraduate degree, I distinctly remember my supervisor saying to me something along the lines of:

"Although I’m being hypocritical, I think one of the best things that you could do would be to start a blog. Unfortunately, I’ve never done it myself, but I wish that I had"

I responded with:

"I’d love to, but I don’t think I have anything to write about".

After all, there are people out there that know far more than I do, and who is interested in listening to me anyway?

This attitude stayed with me and prevented me from doing so until many years later, and I have heard the same, or similar statements, from my colleagues. Recently, on more than one occasion, this has been followed by a conversation on how exactly I got started, and how do I find things to write about.

The purpose of this article is to detail the thought process, and practices that I use, to overcome this mindset, in the hope that it may resonate with others in a similar position. If you have ever read anything I’ve written in the past, this is completely different. It is entirely subjective, and you may not agree with all or any of it. Hopefully it works for you, it probably won’t for everyone.

Finding something to write about

Let’s start with the biggest blocker, finding a suitable topic to write about. Despite working on challenging problems on a daily basis, from university and throughout my career, I struggled with this for a long time. One of the biggest excuses I used, and even still do to an extent, is that it needs to be something ground-breaking and original. In my current role – in which I work on engagements with many different customers, working on one project at a time before moving on to the next – even in these cases, we often aren’t allowed to write about such revelations publicly due to customers’ confidentiality concerns and competitive advantage!

For as long as I can remember, I have set specific time aside to research and learn about subjects that I’m interested in – maths, machine learning, programming languages, even reading history books. Often, I would document my explorations and thoughts on these areas and, if I felt like it may be useful to me again, I would keep it as a reference. On several occasions, when friends or colleagues have commented that they were interested in exploring one of these topics, I have shared these notes with them.

An eye-opening moment for me was, after sharing a guide that I use to run customer envisioning sessions, when a colleague remarked to me that I should publish this somewhere. At this point, I realised that – despite trying and failing to sit down and write blog posts from scratch on previous occasions – in these cases, I had done most of the work, except for the final publishing. As I had been meaning to start writing for a while, I set myself a challenge. Every time I spend a substantial amount of time working on, or learning, something new, I’m going to turn it into a blog post. Honestly, most of these don’t make it to publication – perhaps I’m still a harsh critic – but occasionally, it works out well.

To me, a great blog post is something that saves me time. This doesn’t have to be anything original, it could be a summarisation of a paper, or book, that I’ve been meaning to read. It could have been a presentation of something I’ve read about before but presented or explained in a slightly different manner so that it just ‘clicks’; a great example of this is Jay Alammar’s Illustrated transformer. Alternatively, it could be a well explained tutorial that walks through a problem, so that I don’t have to invest hours in solving it myself.

When you get that feeling that you have truly accomplished something after struggling for a period of time – whether that is finding a new feature in a programming language that you have used for years that blows your mind, finally understanding a research paper after pouring over it for hours, or you have finally completed a task using a library that has limited documentation after thoroughly dissecting the source code – that could be a great candidate for a blog post. You would be saving someone else the time that you have had to spend to acquire that knowledge yourself. I would recommend asking yourself: is this something that would be useful to other people, that would take them time to learn otherwise? If you have just followed an existing tutorial, the answer may be no. However, if you have spent hours reading lots of verbose documentation and aggregated this into a short demonstration, absolutely!

I find that the best time to complete a blog post is as close to this feeling of accomplishment as possible; it can be very difficult to find the motivation to come back to something later after the initial excitement has faded.

Who am I writing for?

One of the main things that I’ve heard from very successful bloggers is to thoroughly know, and research, your audience; that you should always keep them in mind and focus all your efforts to satisfying them. Overall, I think this is great advice. However, I would argue that, for someone just getting started, you should ignore this completely.

One of the main mental blockers that I faced was the persistent thought: who wants to listen to me? I don’t consider myself an authority on any particular topic and, as my interests are quite diverse, I don’t feel like I am strongly aligned to any particular subset of the data science community. In short, I had no idea who my intended audience should be, and this prevented me from ever getting started.

What enabled me to overcome this blocker was writing blog posts with one audience in mind – myself. More specifically, myself a couple of weeks ago, before I had invested the time studying whatever I’m writing about, and myself in a couple of weeks from now, when enough time has past for me to forget the finer details.

Now, when I start to work on a task that I feel may be a good candidate to write about, I ask myself the following question: if there was a blog post in front of me now on this subject, what would I want to know? This thought process helps me to define my objectives and focus on the areas that I should explore.

After finishing my exploration, I ask myself another question: If I need to do the same task again at some point in the future, which details will I need to remember? This helps me to ensure that I have included all the reasoning behind why I made certain decisions, and outline any context needed.

If you stop worrying about who you are writing for and focus on publishing content that you find useful yourself, you never know – your audience may find you.

What if no one reads my posts?

Another concern that I had was that how would I feel if I invested time into creating content, but then no one read it; wouldn’t that mean that the time I spent was wasted?

When working with others, especially with junior team members, we can often find ourselves in a situation where we are asked to explain something that we have experience with to someone else; this is a natural part of teamwork and collaboration. If you provide answers which convince the asker that you are knowledgeable about that particular area, they may refer other people to talk to you about the same topic. As this happens more and more, you may find that you become the go-to person for a particular subject. Imagine that, after being asked about a subject a couple of times, you documented your explanation; the next time someone asks you, you can provide them with that document – there is your first reader.

If there was an existing blog post, or tutorial, already available that provided everything you needed to know about solving a particular problem, it is unlikely that you would have to invest a significant amount of time or effort to arrive at a solution. However, I usually find that there are many resources, each providing some of the information required, which must be aggregated for the task at hand.

I believe that if reading about how I solved the problem can prevent a single person from taking the time to perform the same steps that I did to arrive at a solution, or contributing towards them solving a different problem, then it was worth publishing; it is likely that if two people are both experiencing the same problem, there will be others too!

Personally, I very rarely focus on views or readership stats, or even much promotion for my posts outside of an obligatory status update on LinkedIn – for those in my network who may be interested – and recommending them to colleagues who are facing similar challenges.

Overall, my advice would be not to dwell too much on who is going to read your content. By formalising your thoughts, you force yourself to be much more rigorous about the subject matter that you are learning. Additionally, in the unlikely event that you have encountered a completely unique problem that is not useful to anyone else, a well written, published solution will act as a good reference should you ever need it in the future!

General Writing advice

Whilst no two writing experiences are exactly the same, here are some general pointers on some areas that I’ve found to be useful.

• When I am preparing to spend a significant amount of time investigating a problem, I approach it with a view that – if I feel that my learnings are useful – I will write a blog post about this. Even approaching a problem with this mindset encourages me to make more of an effort to keep my exploration organised.
• If I’m working in Python, my tool of choice for exploratory work is a Jupyter notebook; this is an exploratory environment where I can experiment, as well as documenting my thought process. However, my main focus is to logically document the experimentation process, and the steps that are taken; a mix of text and code screenshots in a word document would serve the same purpose.
• Keep track of your sources from the very beginning. This includes any books or blogs you read, as well as any code you borrow and modify. I like to take the URL and paste this into my notebook; this makes it easy to properly attribute things later and ensure that I’ve referenced everything that I used.
• I always start with an objective, and explicitly write this down at the top of the notebook. This can be quite broad, e.g., "How to run a PyTorch distributed training job using Azure Machine Learning".
• When working with notebooks – even outside of Blogging – I always find it helpful to imagine that I’m writing a tutorial for someone else to read; keeping notes on what I’m doing, and why, between code cells. When things are likely to change, these may be very short – not even full sentences – and expanded upon later.
• There is nothing worse than a messy notebook, and I find a good indicator of this is if you find yourself jumping around to execute cells out of order. In these cases, I tend to duplicate cells, so that it all runs in order, and then refactor any repeated sections into functions or classes.
• Once I feel like I’ve completed my objective, I find it helpful to proceed sequentially through the notebook and imagine that I’m showing my work to someone sitting next to me. Every time I feel like I’d have to explain something, I write this down. This is also a good opportunity to do a final clean-up of the code and notebook structure.
• Doubt can start to creep in the longer that you take to publish. Once you feel that you have finished, proofread it a couple of times, port to your platform of choice, and publish! Any small errors that you’ve missed can be corrected later.

Final Thoughts

I am still relatively new to this, but I genuinely believe that Writing blog posts has been incredibly helpful to my overall learning process and has given me resources to direct people to when they ask me to share the learnings of areas that I’ve been exploring.

Although it still takes me by surprise, on the rare occasion that someone approaches me – either in person or through message – to let me know that something I’ve written helps them out, it feels incredibly rewarding!

If writing blog posts is something that you have been thinking about for a while, or if you are on the fence about whether it is right for you, I would thoroughly recommend that you suspend all of your doubts and try it at least once; after all, what’s the worst that can happen?

Chris Hughes is on LinkedIn.

References

• Setting up Machine Learning projects for success | by Chris Hughes | Data Science at Microsoft | Medium
• The Illustrated Transformer – Jay Alammar – Visualizing machine learning one concept at a time. (jalammar.github.io)
• Effortless distributed training for PyTorch models with Azure Machine Learning and PyTorch-accelerated | by Chris Hughes | Dec, 2021 | Towards Data Science

The post How I started writing Data Science blog posts: overcoming fear and procrastination appeared first on Towards Data Science.

I appeared on the FastBook Reading Sessions organised by Weights & Biases to discuss about the benefits of writing in Data Science. I wrote this piece to summarize what I covered there. This article was originally published on their forums but I’m sharing a somewhat edited version here as well. Primarily it discusses why writing matters in data science and how it can be used as a tool to leverage your portfolio.

The best way to learn any concept, especially in data science, is by writing about it. It helps you understand the topic in detail, and your work might, in turn, help others. But this is easier said than done. Even though many people want to write, it takes them months and sometimes years to move past the initial challenges of self-doubts. Rachel Thomas‘s blog post on Why you (yes, you) should blog very aptly touches upon this issue. Infact it covers all the vital points that you should keep in mind when starting to writing. My professional journey started with writing, and in this post, I’ll like to share how you can use your writing skills to create a compelling Portfolio.

Why is writing useful?

Writing, especially in Data Science, is an important skill set. It gives you both voice and visibility. According to me, the benefits of writing can be summed up as follows:

• Retention: Writing helps you to retain a new concept.
• Research: Writing helps to develop a research mindset.
• Reputation: Writing helps to build a reputation in the community. People cite your articles, mention you in their talks, etc.
• Revenue: Writing can also be a self-sufficient career on its own. With the creator economy booming, writing can lead to potential job offers.

How to begin your Writing journey

Half the battle is won when you decide to write an article. However, the second part is to decide on things like what and where to write, topics, length, etc. It would be best if you created your own writing path. Start with the most comfortable concepts; This will give you the required confidence to start. Gradually diversify your writing portfolio. Start touching on new things and try writing about them. Writing is an iterative process. The more you write, the more you learn, and the better you write.

Things you can write about

• How to use Data Science libraries
• Research paper summaries
• Courses reviews
• Interview people in the community
• Clarify concepts
• Summarize tech videos/ webinars
• Convert your Kaggle notebooks to blog posts, etc

These are just a handful but can act as a great starting point.

Showcase your work

The amount of hard work that goes into writing an article is the same whether you write it for one person or one thousand. As such, make sure to share them on other platforms too, like Linkedin, Twitter, and other community groups, of which you are part. Please don’t ask them to like or share per se. This is because there is a fine line between sharing your work and spamming. If people like it and find it interesting, trust me, they would want to share it themselves.

Another way to showcase your work is to use them in meetups, presentations, and conferences. Content creation is challenging, but once it is done, it can be reused in multiple ways.

Create a public portfolio

You can write on open blogging platforms or create your website. This is entirely up to you. But make sure to start building on a good portfolio right from the start. A Github page, a Kaggle profile, a Stack Overflow, etc., can support your resume.

Contribute to the community

Community is the backbone of Data Science. Get involved with the local Meetup community. Talk at conferences – from local to regional and even national. You could even mentor others who are new to this field. Answer and help others in forums. Try contributing to the documentation of open source libraries.

This is all good but will anyone read my article?

This is by far the most common question that I come across. Self-doubt before even starting is pretty common not only in data science but in almost all fields. Some common doubts that people have are:

There are already tons of articles on the topic. What difference would my article make?

Let’s say you want to write about a library. You do a quick search and find that there are ten more good articles on the topic. So would you drop the idea? No, definitely not. Try giving different treatment to your article by using an interesting dataset and focusing on a lesser-known aspect of the library. For instance, I wrote an article about Lux – A Python library, where I showcased Simpson’s effect using the library. The library creators loved it and included it as part of the documentation.

Here is another example. I recently wrote an article Interpretable or Accurate? Why Not Both?, where I explained the concept of Explanable Boosting Machines – models designed to have accuracy comparable to Boosted Trees while being highly intelligible and explainable. EBM is an open-source library by Microsoft. I got a personal message from the team for writing the article.

These are just a few examples highlighting the immense value that writing could bring not only to your resume but in terms of networking within the community too.

Resources to get started

Now that you have decided to take the plunge, here are some open-source tools that will help you to get started.

• Blogsite: Fastpages – An easy-to-use blogging platform, with support for Jupyter notebooks, Word docs, and Markdown. Another option could be Medium.
• Images: Unsplash, Pixabay, and Pexels are some sites that provide free images and photos that you can download and use for any project.
• Illustrations: unDraw for open-source illustrations that can be customized.
• Custom Images: Canva for designing and publishing almost anything.
• Proofreading: Grammarly’s freemium version for spell checking and catching the most basic grammar mistakes. Small SEO Tools, on the other hand, provides free plagiarism check in addition to checking grammar.

Feedback is essential but don’t get disheartened.

Leadership expert Ken Blanchard aptly coined the phrase, "Feedback is the breakfast of champions.". Getting critical feedback is imperative to improve, and the same applies to writing also. However, giving feedback is important but denouncing someone’s work( especially publically) is totally wrong. Sadly, many a time, you will encounter such situations. Some people leave very distasteful remarks on your articles, without giving any reason.

However, my advice for you would be not to let negativity overcome your passion. I resonate with Suzana’s tweet above. Creating anything is hard and requires a lot of patience, perseverance, and persistence.

Conclusion

The most important thing, in my opinion, is to find the topic you are interested in that creates questions you want to answer. We tend to procrastinate a lot over things like writing. We believe our first blog has to be the best. Like I mentioned above, writing is an iterative process, and every article that we write makes us better writers. So get out there, get your hands dirty and have fun writing.

The post Building a compelling Data Science Portfolio with writing appeared first on Towards Data Science.