We are proud to present an AI and machine learning system, that started to work with Steem blockchain. The system would provide a personalized statistics and information to the users calculated with some mathematical methods with AI elements. For example as one of the future tasks for the system we envision the discovery of the regularities in users behaviour and provision of a personified content feed based on user’s personal preferences as well as the preferences of his friends.
Aigents technology is being developed by the group of scientists at the Novosibirsk State University lead by Anton Kolonin (Сandidate of Science, expert of the Neuronet division of the National Technical Initiative of Russian Federation, inventor of algorithms and creator of the computing systems for computation of geophysical and financial data, image recognition and data retrieval from text, video and audio data).
We already adapted Aigents to work with Steem blockchain and obtained some interesting data, that we would like to share with you in the some beginning posts. In the future, with the feedback we’d like to gather, we’ll launch a friendly web interface, that would allow any user to get some interesting information about themselves, or anyone else.
As the first “text subject” we’ve selected @dantheman (Daniel Larimer, Steemit’s architect).

picture by @konti

---

My interests

Based on the analysis of the post contents and comments to them, as well as upvotes of posts and comments (using the clusterization principle from the AI terminology)

Subjects	Number of publications
based, blockchain, free, individuals, once, under, upon	30
network, rate, resources	23

---

Similar to me

Based on the contents of @dantheman posts and other Steem users.

Rank,%	Friend	Overlapping	My approval	Approval	Comments
100	@lavater	227	2	5	6
94	@liondani	226	4	41	21
92	@kyriacos	372	13	28	11
91	@mrosenquist	330	4	13	14
91	@lukestokes	346	8	15	11
90	@anotherjoe	262	3	31	9
88	@son-of-satire	313	3	3	4
88	@fyrstikken	405	8	39	11
88	@dragosroua	390	4	30	12
88	@ats-david	364	3	19	15
87	@tibonova	176	2	19	10
85	@knozaki2015	199	1	5	5
84	@donkeypong	146	2	10	5

---

Best friends

Based on the social interaction (upvotes and comments)

Rank,%	Friend	My approval	Approval	Comments
100	@krnel	82	42	28
50	@dwinblood	31	69	24
13	@kevinwong	19	34	5
12	@timcliff	11	47	17
10	@stellabelle	10	29	28
10	@klye	9	43	21
9	@kyriacos	13	28	11
8	@thecryptofiend	10	25	21
7	@beanz	7	43	17
7	@fyrstikken	8	39	11
5	@the-alien	7	36	5

---

Followers

Users that upvote user’s content most frequently (a bit incorrect because of the bot activity, which we are not yet detecting and accounting for, but will start filtering in the future).

Rank,%	Friend	My approval	Approval	Comments
100	@johnathanhenry	0	180	0
57	@jasen.g1311	0	102	0
50	@reported	0	90	0
46	@jaredandanissa	0	82	0
43	@cnmtz	0	78	0
41	@southbaybits	0	74	0
41	@steemleak	0	73	0
38	@christoph3	0	69	0
36	@cmtzco	0	64	0
34	@krystle	0	62	0
27	@bitcoinparadise	0	39	9
23	@guest123	0	41	0
23	@debmund	0	42	0

---

Figures of authority

Authors that produce content that is most interesting to the user (they upvote and comment it), but that is not reciprocal.

Rank,%	Friend	My approval	Approval	Comments
100	@steemitblog	14	0	0
69	@dollarvigilante	29	2	0
50	@conspi-theorist	7	0	0
50	@lukewearechange	7	0	0
43	@churdtzu	6	0	0
39	@larkenrose	11	0	1
38	@ozchartart	37	5	1
36	@jimlimberdavis	5	0	0
29	@lifeisawesome	4	0	0
29	@jaredhowe	8	0	1
21	@timthevagabond	3	0	0
21	@busy.org	3	0	0
21	@danilo-cuellar	3	0	0
21	@throw-away911	3	0	0

---

Users favorite words

Based on analysis of all actions by @dantheman - posts, upvotes, and comments.

Rank,%	Friend	My approval	Approval	Comments	Posts	Mentions
100	consensus	11	6414	352	11	204
96	people	25	14012	792	25	205
76	steem	18	10489	590	18	160
59	everyone	20	11205	660	20	125
54	means	24	13278	707	24	115
52	reputation	8	4599	313	8	102
50	blockchain	14	7934	437	14	104
46	value	20	11338	665	20	98
45	transactions	9	5169	219	9	90
44	property	12	6388	314	12	91
43	money	21	11600	651	21	91
42	income	11	5820	292	11	86
42	market	19	10411	558	19	88

---

Users words by period

This analysis shows the themes, that the author wrote most about in any given period.

---

Please write which users of Steem you’d like to have information about in the further posts?
Which data and connections between it you’d like to see?

---

Regular statistics would be published by the @aigents, our first priority is to create a monetization stream for A. Kolonin, so the development would continue. @akolonin was invited by @dark.sun. Service is being developed under his leadership. Thanks to @akolonin for participation and belief in our community ;-) Thanks to @lehard for help in preparing this material and participation in use case development.

Translation by @xanoxt.