Asking, Helping, and Thoughtful Behaviour
Lessons from A Recent Kaggle Discussion Grandmaster
Kaggle Communities- Missed Opportunity
When you head over to Kaggle’s hot and new Communities section, you might be a little disappointed. I was. I am, mostly.
Initially, I was excited that Kaggle was very thoughtful and considerate to introduce this feature. But soon, I found that the Communities were abused to stack up upvotes and were filled with spam. Kaggle is not at any fault here.
There was helpful content as well, but, sadly, they were in the minority. Communities were filled with aggregated Notebooks, aggregations of posts, and aggregations of other aggregations… You get the idea.
I did not like this. Moreover, I gasped at the missed opportunity. There are several forums out there- the Data Science Stack Exchange, the Machine Learning Subreddit on Reddit, Cross Validated, and so on.
I missed a unified, organized platform, that functioned not only as a Q&A site but also as a place where you go to post anything Data Science, Machine Learning, and Deep Learning and engage with similar content posted by others. I was happy to see Kaggle launch separate Communities.
I was excited about its future.
But, people who wanted to accumulate votes, and thence Medals to climb the ladder of the Kaggle hierarchy. So it kept being flooded by people wanting to accrue votes.
I felt hugely disappointed.
There’s Hope
While being flooded in the aggregations, there are still some genuinely helpful posts. And there are write-ups by winners of competitions, people genuinely posting content which are helpful in many ways.
Helpful content comes in many forms, but in my opinion, they are mostly from competitions. (I have tried writing posts stemming from my observations, such as this one.) The discussions that are thoughtfully tagged by people writing them are to thank.
Even when someone is not taking part in a competition, or explicitly visiting the Discussion section of the competitions, they get to read valuable posts and related discussions. And above all, such posts are useful and add value to your learning journey.
And people who add value to our lives through their hard work, definitely deserve recognition.
Lessons from a Freshly Minted Grandmaster
Gabriel Preda has recently become a Discussion Grandmaster on Kaggle.
He shared his findings and we can find sound advice from that post. Let’s go over them quickly-
Engage in conversations, and create posts on Datasets and Competitions. I also think it is important to comment on Notebooks as well.
Teach. Share what you already know, in a structured, well-thought way.
What you teach does not always have to be full-length Notebooks, intended to be taken as mini-courses, feel free to share what you know in what Preda calls “pills”-sized chunks as well.
Be genuinely helpful and try to solve other people’s problems.
Praise other people, especially beginners for their hard work.
Don’t be spammy and never use the mention feature in a desperate attempt to gain an upvote.
Gabriel’s advices are genuinely helpful, and I believe, will make Kaggle better for everyone.
Consider giving my website a visit, and follow me on Twitter.
Cheers and peace!