https://steemitimages.com/DQmazRMnJ4wrVvpnVKSz9cv3JnfHX32MDr59dVXbjfVFan8/%E1%84%89%E1%85%B3%E1%84%8F%E1%85%B3%E1%84%85%E1%85%B5%E1%86%AB%E1%84%89%E1%85%A3%E1%86%BA%202017-08-28%20%E1%84%8B%E1%85%A9%E1%84%92%E1%85%AE%2012.44.47.png

This article utilizes publicly available data to capture suspicious users in Steemit, where writers and subscribers are rewarded through posting an article and/or voting an article.

I believe that the biggest advantage of the Block-Chain technology is the disclosure of all transaction information. Steemit is built on the basis of black-chain technology in which all of the user's actions, including writing / commenting and boarding history are all publicly available. But everyone are not able to access those information easily. The purpose of this article is to provide the aforementioned information in a graphical representation so anyone can easily view/understand.

                  For now, I will not disclose the actual user name of the user. However, you should be aware that the block-chain technology is open to the public and any of your actions will be stored permenantly and can be easily monitored.

Period of Data Source (2017-08-24 ~ 2017-08-31)

-Total Amount of Data: 145393

-Total Number of Steem Users: 17407



Self-Voting (SV)

https://i.imgur.com/FYS9RvO.png
The above graph shows that the total number of Steemit Users who commited Self-Voting(SV) is 17407, where the descriptive statistics including average, maximum and standard deviation showed 8.35, 1164 and 21.59 respectively during 2017-08-24 ~ 2017-08-31.

Top 10 Steemit Users who have commited most self-voting during 2017-08-24 ~ 2017-08-31:

https://i.imgur.com/wdKEdmU.png
Self-Voting Top 10 user's 7 day voting pattern

https://i.imgur.com/0tcjD88.jpg
The above network describes the voting relationship between SV users and others where the SV users are colored with red during 2017-08-24 ~ 2017-08-31.

https://i.imgur.com/D6VXovl.png
Self-Voting Ranked between 1~5

https://i.imgur.com/L2k2G9p.png
Self-Voting Ranked between 6 ~ 10

https://i.imgur.com/x3GGP93.png
The above graph shows the top 10 HSV users behaviour including the total number of voting (gray), the number of SV (red), and the number of NSV (blue) during (2017-08-24 ~ 2017-08-31)

High-Self-Voter (H-SV) Steemit Users' Suspicious Scores

The Suspicious score of H-SV Steemit users is extracted with the following features:

-VP -- User's voting power

-voting_cnt -- Total number of voting during the period

-sv_voting_cnt -- Total number of self voting during the period

-nsv_voting_cnt -- Total number of voting placed for other users articles/comments during the period

-author -- Number of other users

The indicators extracted from the above features are as follows:

-sv_avg_vp -- Average voting power for self voting (=SV 총 VP합/sv_voting_cnt)

-nsv_avg_vp --  Average voting power for other users (=NSV 총 VP합/nsv_voting_cnt)

-sv_ratio -- User's self-voting rate (=sv_voting_cnt/voting_cnt * 100)

-VPA -- Average number of votes of other users (=nsv_voting_cnt/author)

The virtual indicators for compensation measurement are (USD):

-Self Rewards(SR - Self Rewards) -- The total amount of self compensatio when the user's voting reward is assumed to be $1 (=(1 * sv_avg_vp) * sv_voting_cnt)

-Not Self Rewards(NSR) -- The total amount of other users' compensation when the user's voting reward is assumed to be $1 (= (1 * nsv_avg_vp) * nsv_voting_cnt).

-Reward Per User (RPU) - Average compensation amount of other users when the user's voting reward is assumed to be $1 (= NSR / author).

https://i.imgur.com/Sls1NDk.png
Voting patterns of H-SV users during 2017-08-24 ~ 2017-08-31 including the number of voting (voting_cnt), the number of self-visits (sv_voting_cnt), the number of other users (nsv_voting_cnt), and the number of user visits (hsv_voting_cnt), Author number (authors), self-voting rate (sv_ratio), and average number of other users (VPA).

https://i.imgur.com/b1mRtwJ.png
Voting patterns of H-SV users with a relatively high voting rate, SV_Ratio 9 ~ 10 - Red

https://i.imgur.com/nnC10MW.png
The distribution of rewards for H-SV users (red) and others (blue)  distributions when the reward of individual voting is assumed to be $1.

https://i.imgur.com/Aw6HODL.png