Taking a Look at Interesteem

Have you noticed comments made by @interesteem on various posts lately?

I have, but only decided to look more into it after I was mentioned twice by this account. When I checked, I saw one of my posts from a few days ago was recommended both times.

So, what is with these comments?

They show a list of related posts that may interest you, and a percent in the parenthesis, which I believe is how much of a match is the respective post with the one you've just read.

The related posts are discovered through DeepLearning and presented to you, IF the author of the post uses the tag #interesteem on his post.

This is interesting to mention, it means this account doesn't just go around spamming peoples' posts with their comments. Authors ask for these recommendations by using the tag.

On the Recent related articles list we can find the recommended posts which are still in the payout period, thus can still be upvoted.

Older related articles can include posts that have exited the payout period, if relevant ones are discovered by this algorithm.

Apparently it takes one hour for this comment to be made, I found out by reading a reply by the author on interesteem's presentation post.

I believe there is a great need of a "related posts" feature on Steem dApps. Without a good search engine, with a busted trending page, sometimes with a flooded feed (even when you don't follow so many people), this is an important feature to have, to be able to research something deeper, maybe to discover other authors in the process as well.

There are two other developments on the agenda of interesteem right now:

• building their own website
• "Developing eternal writer reward system based on interesteem"

The second one has been, one way or another, desired by the Steem community for a long time. Rewarding posts the pass the test of time, long after the payout period is over, is something that will make people want to write more content that is remarkable and doesn't lose value in time.

I'll wrap up this post with one question that some people may have. Is or will the DeepLearning / selection code be open source?

And for my readers:
Have you seen any interesteem comments yet? What do you think about them?