Steem - Post Data and Automated Accounts

in #stats7 years ago (edited)

A few days ago I looked at the number of Posts and Comments made by unique accounts in the last few months of 2017. The features of this distribution were quite clear, there were a lot of individual accounts posting small numbers of posts per day (on average). This analysis raised a question in my mind.

Is it possible to use the posting activity of accounts to isolate which accounts are bots?



Image Source: pixabay.com

In this post I am going to focus my analysis on Posts (excluding comment) to see what trends emerge. I will follow up in the coming days with a similar analysis for Comments, where I would suspect a lot more bots and automated posting is going on.

Distribution of Posting Frequency

If we slice the data by month and look at the distributions of posting activity by account we get the following boxplot.

  • The pink rectangles show that the majority of accounts posts 10 or less posts per month.
  • The blue shaded region is where some accounts make up to four posts per day.

So how many times did the most active accounts post?

Posts Oct - Dec

This graph shows the distribution of number of posts from the 100 most active individual accounts over this period.

  • Around 75 accounts post more than 1,000 times in this period. These are outliers.
    Clearly there are some automated accounts.

If there is one, what is the number of posts a day that we could use to identify automated accounts?

I would propose anything higher than 4 posts per day (on average) uses some sort of automation. It may be a regular update post, charts or some other automated post.

Using this criteria I next look at the splits of posts per day between automated and manual. Automated Posts have posted more than 120 times in a month. This produces a list of 736 accounts that regularly post more than 4 times a day.

Posts Per Day

This next series of graphs shows the split of number of posts per day based on this criteria to identify manual and automated posts.

Number of Posts (Ex Comments)

  • automated > 4 posts per day (on average)
  • manual everything else

Number of Posts (Ex Comments)

The last graph shows the total post count and the visually the split but we can also plot the individual components separately to identify trends.

  • We can see there has been a significant increase in Automated Posts over the period.

Have you come across any interesting ways to gauge the amount of users on Steemit or Steem?

With this series I am analysing trends in the Steemit Account Activity to see what the most useful metrics are for identifying growth and activity on the platform. I have come across some interesting trends in the data which I hope to analyse regularly and which the community may find useful. There are a few more items I will look at in the coming days related to accounts but please let me know if there is anything in particular you would like to see. Thanks for reading.

Related Posts

I am taking a deep dive into the Accounts of Users in this series of posts. You may also be interest in:



Thank you for reading this. I write on Steemit about Blockchain, Cryptocurrency, Travel and lots of random topics.

Sort:  

While looking at the new page it is easy to find bots at work. I've seen several accounts posting the same image within seconds of each other. I worry what all this fluff is doing to the blockchain. Maybe a tweek of the bandwidth allocation could remove some of this spam.

I am hopeful soon we will see new ways of viewing the posts. Here is an example I have been following for a while.
https://www.insteem.com/

Interesting. But the content fails to load. In the console I see the error: "Unhandled promise rejection
TypeError: NetworkError when attempting to fetch resource."

I think this site is a work in progress. It's a good example of what people are working on. Check back in on it.

Interesting but worrying that you made this estimation that above four posts a day could signify automated posting. It's certainly not the case in my circle of artworks and contests. I think you should remove contest posts from your analysis. Contest entrants can easily make more than four post a day so can contest hosts.

Good point. Here I have looked at averages over a month. I run a contest too and some days there are many posts but not consistently over the month.

I am not saying there is anything wrong with multiple posts, I think it's something we will actually see more of. The blockchain will have many uses but I still think it's useful to isolate automated or frequent posts when looking at user growth.

Yes, I understand :)

I think that a quantitative analysis like this can get us close to identifying the automated posts but that some form of content classification is needed forhigher precision. it can be a simple good-neutral-bad scale to start with,

I agree more detailed analysis will give deeper insights. I am also looking for some high level metrics of growth and activity. It's difficult to find the right balance to derive something that is useful to look at on a regular basis.

Can it be further analyse that what exactly is being posted? Sometimes it is just a meme (that isn't even funny) and yet it garners huge monetary support from the reward pool.

Though it would be difficult to qualitatively analyse posts but I was just thinking out loud.

Interesting. It sure is possible to classify the content in some way. Analysing by tags would give a good overview. I am also working on additional ways to classify content.

Please let me know if I can be of any help. Thanks for replying :)

I think ur post very true and great ....
I upvote this post

Thank you.

That is a very interisting finding. I have seen some comments made by bots but I have no idea how many of them actually

This post very nice..i like steemit.steemit is a great social network..thanks for sharing this post..best of luck.

I am always trying to improve what I post so please let me know what did you like most about it?

interesting stats, will be looking forward to the analysis for comments

Thanks so much. I think the analysis for comments will be very interesting.

Very interesting, keep researching more deeply, so steemit users will be happy with the new features. good work @eroche. regards from @maskur2840, I really appreciate your work.

The analysis is very smart, this is a positive reading for me personally, I can know how the condition statistics in the steem, whether it's about posting and comments, I also do not understand why there should be bots to comment and also do a post automat. are they part of sabotage on this platform? and I think these bots bots are very disturbing to the community, Thank you for sharing sir :)