create account

RE: Issues With Using Stake To Filter Content by doodlebear

View this thread on: hive.blogpeakd.comecency.com

Viewing a response to: @greer184/issues-with-using-stake-to-filter-content

· @doodlebear ·
$0.03
Yesterday, I thought about ways of how to improve the feed. I usually ignore the trending and hot categories because as you write, it's about what whales like and about who's paying the most to the bots. So, not really interesting. 

The only good category is the new one, because that's where you don't have the money bias in. Instead there is about 80% BS that you have to filter out first, but then you can find interesting content. (Strangely, I still have about the same share of stuff I don't need in my feed. None of my unfollow purges helped so far...)

A system level approach to fix the content filter problem is out of the question. The people in charge aren't interested in an improvement. That's why I thought into the direction of topic bots that only resteem articles that have for example as their first tag "politics". In this category the bot would resteem  everything, except maybe one of the other tags is one of the notorious spam signals like "dmania" or "colorchallenge".  Also important is the language of the articles that get resteemed. You could name that bot @resteem-en-politics, which resteems politics articles in English. Another one could be named @resteem-en-sports which is doing the same with sports. And one could be named @resteem-de-deutsch, @resteem-en-philosophy, @resteem-es-php etc.

All topics could get their own bot and these bots do nothing else but resteeming a single topic in a single language. What you as a user have to do is to follow that bot and you exclusively get content of your liking into your feed. 

I think that would be the most simple and efficient solution. Maybe there could be combined bots where two specific tags are necessary to thin out the amount, if that is necessary (e.g. @resteem-en-sports+racing).(I don't know how many  articles are produced daily per category, but that should be added as information so you know how many posts to expect.)

Plus of course a blacklist for spammers and scammers.

The programming for that I believe is very simple:
- select a specific tag(s)
- exclude specific tags
- select the correct language
- exclude spammers and scammers
- resteem the selected articles

If you do it right, I guess per language the average user would follow about 20-30 of these bots, of which everyone resteems perhalps about 5 posts per day. This wouldn't be more than the daily output of a big online news outlet.

The biggest problem is how to acquire all these accounts, because you need different phone numbers and IP addresses since the developers try to limit the number of accounts to one per person. So effectively, you'd have to talk to the developers and get their approval - something which could prove to be impossible.

Other biggest problems of Steemit are as I think the reward pool rape by the big bots as I described it in a recent post and the gap between witnesses&developers to the real world and their sheer ignorance, greed and incompetence. This must change and I intend to create an account for which I will collect as many votes for witnesses as possible. With that I will try to pressure them into the direction of more quality, more professionalism, more transparency and more work. Too many witnesses are lazy and just ripping off the system. Either they improve or they won't get a vote. At least theoretically, that should work.^^
👍  
properties (23)
authordoodlebear
permlinkre-greer184-issues-with-using-stake-to-filter-content-20180219t134358984z
categorycontent
json_metadata{"tags":["content"],"users":["resteem-es-php"],"app":"steemit/0.1"}
created2018-02-19 13:44:00
last_update2018-02-19 13:44:00
depth1
children14
last_payout2018-02-26 13:44:00
cashout_time1969-12-31 23:59:59
total_payout_value0.026 HBD
curator_payout_value0.006 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length3,348
author_reputation-1,445,012,251,659
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,794,986
net_rshares6,414,828,082
author_curate_reward""
vote details (1)
@carn · (edited)
$0.15
You can create as many accounts as you want if you fund them with SP. The phone number/IP thing is just for free accounts. You can also create accounts through an API or SteemConnect ([tutorial for SteemConnect](https://steemit.com/news/@timcliff/new-tool-from-busy-org-create-new-steem-blockchain-accounts-with-steemconnect))

I currently own this account plus two others (@steemitbooks and @shadeslayer) and I've also created one for a friend.
👍  
properties (23)
authorcarn
permlinkre-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180219t181634266z
categorycontent
json_metadata{"tags":["content"],"users":["steemitbooks","shadeslayer"],"links":["https://steemit.com/news/@timcliff/new-tool-from-busy-org-create-new-steem-blockchain-accounts-with-steemconnect"],"app":"steemit/0.1"}
created2018-02-19 18:16:39
last_update2018-02-19 18:17:42
depth2
children7
last_payout2018-02-26 18:16:39
cashout_time1969-12-31 23:59:59
total_payout_value0.118 HBD
curator_payout_value0.036 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length445
author_reputation849,541,941,985
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,852,043
net_rshares27,815,513,473
author_curate_reward""
vote details (1)
@doodlebear ·
thx for the info! The creation of theme-based resteem bots would require to creat dozens of them, which makes it very expensive. The way via different IPs/Tels is the affordable way to do so. 

I didn't know about the Steemconnect way of creating an account, is it for free and can you create multiple accounts with it?
properties (22)
authordoodlebear
permlinkre-carn-re-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180219t184652009z
categorycontent
json_metadata{"tags":["content"],"app":"steemit/0.1"}
created2018-02-19 18:46:54
last_update2018-02-19 18:46:54
depth3
children6
last_payout2018-02-26 18:46:54
cashout_time1969-12-31 23:59:59
total_payout_value0.000 HBD
curator_payout_value0.000 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length319
author_reputation-1,445,012,251,659
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,858,205
net_rshares0
@carn ·
$0.02
The Steemconnect way is just like the way normal accounts are created but more accessible since you don't need to run any code. The only cost is the Steem or SP delegation that the new account needs. You don't ever lose any Steem, it just gets transferred to your new account. You can create unlimited accounts this way (all of the accounts I mentioned were created that way)

Steemit requires phone verification because they have to create accounts the same way (they are "paying" the registration fee with their delegation) and they don't want to pay for your secondary accounts.
👍  
properties (23)
authorcarn
permlinkre-doodlebear-re-carn-re-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180219t195250671z
categorycontent
json_metadata{"tags":["content"],"app":"steemit/0.1"}
created2018-02-19 19:52:57
last_update2018-02-19 19:52:57
depth4
children5
last_payout2018-02-26 19:52:57
cashout_time1969-12-31 23:59:59
total_payout_value0.018 HBD
curator_payout_value0.003 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length581
author_reputation849,541,941,985
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,870,359
net_rshares4,398,929,884
author_curate_reward""
vote details (1)
@carn ·
$0.02
Instead of resteeming posts, the bots would do better to compile a post with links to the articles and the first paragraph. This would help with not generating tons of posts in the followers' feeds. 

Really though, it is better for separate curators to handpick articles then to leave the work up to bots. I'm currently working on getting a Discord server set up for my @steemitbooks account where people can drop links to book related posts and I (or other people I assign as curators) can look through and select the best ones to feature in a curation post. While it requires more work it will lead to better quality feed then if you let a bot do it.
👍  
properties (23)
authorcarn
permlinkre-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180219t200148045z
categorycontent
json_metadata{"tags":["content"],"users":["steemitbooks"],"app":"steemit/0.1"}
created2018-02-19 20:01:48
last_update2018-02-19 20:01:48
depth2
children1
last_payout2018-02-26 20:01:48
cashout_time1969-12-31 23:59:59
total_payout_value0.018 HBD
curator_payout_value0.003 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length653
author_reputation849,541,941,985
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,872,048
net_rshares4,398,929,884
author_curate_reward""
vote details (1)
@doodlebear ·
Yes, currently, the manual selection is probably the better way of spreading content.
properties (22)
authordoodlebear
permlinkre-carn-re-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180219t220440512z
categorycontent
json_metadata{"tags":["content"],"app":"steemit/0.1"}
created2018-02-19 22:04:42
last_update2018-02-19 22:04:42
depth3
children0
last_payout2018-02-26 22:04:42
cashout_time1969-12-31 23:59:59
total_payout_value0.000 HBD
curator_payout_value0.000 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length85
author_reputation-1,445,012,251,659
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,894,265
net_rshares0
@greer184 ·
$0.02
Resteem bots are an interesting idea, but the issue with those is that they can overrun your feed. But they can clean out some of the junk and pollution through cross-tagging. But you could incorporate those with additional filters to create some interesting feeds. Or you could find a way to extract the posts from those bots and build an application on top of those bots. 

I foresee that some curation organizations will develop in the future such that they simply resteem content, but spend copious amounts of time looking for specific posts, hire qualified writers, etc.
👍  
properties (23)
authorgreer184
permlinkre-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180219t194038130z
categorycontent
json_metadata{"tags":["content"],"app":"steemit/0.1"}
created2018-02-19 19:40:39
last_update2018-02-19 19:40:39
depth2
children2
last_payout2018-02-26 19:40:39
cashout_time1969-12-31 23:59:59
total_payout_value0.018 HBD
curator_payout_value0.003 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length575
author_reputation12,340,680,341,190
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,868,170
net_rshares4,540,830,848
author_curate_reward""
vote details (1)
@doodlebear · (edited)
Yes, the problem of getting overran is always there. I would try to get that under control by looking at the average number of posts and if it goes up it implies a manual rewiew. Another rule could be to only allow users with a certain reputation and only articles with a minimum amount of words. Or as you say it: Add other filters.

If such a resteem bot grows popular, it could even charge money if someone wants his content to be added, but without guarantee of a resteem.

*edit*

Something unrelated: Can you give me a subdomain of your herokuapp.com URL (plus some space and a PHP access if possible)? I would give you 5 SBD for it and I want to start building a site for my little witness quality control project.

*edit2*

I just thought that thanks to the public blockchain it would also be possible to automatically resteem content that has been upvoted or  resteemed by certain talented users who are good curators. You basically take a free-ride on their activity.
properties (22)
authordoodlebear
permlinkre-greer184-re-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180219t220317577z
categorycontent
json_metadata{"tags":["content"],"app":"steemit/0.1"}
created2018-02-19 22:03:18
last_update2018-02-19 22:21:15
depth3
children1
last_payout2018-02-26 22:03:18
cashout_time1969-12-31 23:59:59
total_payout_value0.000 HBD
curator_payout_value0.000 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length977
author_reputation-1,445,012,251,659
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id38,894,018
net_rshares0
@greer184 ·
So Heroku manages and owns subdomains of herokuapp.com. They allow you to deploy small applications on their servers for free and then charge you if you want to scale up. Unfortunately, I don't own the domain or server space.
properties (22)
authorgreer184
permlinkre-doodlebear-re-greer184-re-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180220t170632807z
categorycontent
json_metadata{"tags":["content"],"app":"steemit/0.1"}
created2018-02-20 17:06:33
last_update2018-02-20 17:06:33
depth4
children0
last_payout2018-02-27 17:06:33
cashout_time1969-12-31 23:59:59
total_payout_value0.000 HBD
curator_payout_value0.000 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length225
author_reputation12,340,680,341,190
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id39,118,562
net_rshares0
@nokodemion ·
$0.24
You are talking proxy voting? Hmmm, interesting ...
👍  
properties (23)
authornokodemion
permlinkre-doodlebear-re-greer184-issues-with-using-stake-to-filter-content-20180220t233912293z
categorycontent
json_metadata{"tags":["content"],"app":"steemit/0.1"}
created2018-02-20 23:39:12
last_update2018-02-20 23:39:12
depth2
children0
last_payout2018-02-27 23:39:12
cashout_time1969-12-31 23:59:59
total_payout_value0.229 HBD
curator_payout_value0.013 HBD
pending_payout_value0.000 HBD
promoted0.000 HBD
body_length51
author_reputation6,059,124,243,903
root_title"Issues With Using Stake To Filter Content"
beneficiaries[]
max_accepted_payout1,000,000.000 HBD
percent_hbd10,000
post_id39,194,367
net_rshares44,536,738,550
author_curate_reward""
vote details (1)