eben nicht ausgearbeitet und nur ein post um zu schauen ob sich überhaupt wer dafür interessiert.
Would you support (my) research on a suggestion algorithm for Hive?Why we need a recommendation page:Some of you might miss a recommendation page on Hive frontends like Youtube, Medium, Instagram and other social media offers.In my opinion such a recommendation page is a basic element of a good social media plattform which is still missing on Hive with great effects on the value which Hive add to its users/consumers.At this moment its extremely different to find content you are interested in and which brings you value.A good recommendation page would have several positive effects in my opinion, like:
* Making Hive more Consumer friendly; it would make much more fun to spend time on Hive, because you could better spend your time consuming instead of searching through communitys and beeing stressed by content you are not interested in.
* It would bring Hive Content quality to a new level, because good content would be more appreciated, especially old Content because it gets visible again. Today there is nearly no chance to enjoy great old content, because its not visible on Hive frontends through communitys,... except somebody crosspost/rehived it or you directly are scrolling to an blog... but come on this has no practical relevance, so actual old content "absorbed" in the depth of the Hive Blockchain. But its still there, we only have to make it visible again to people who are interested in it and would benefit by seeing! This would make it more worthwhile to create good content because more people will see it even weeks/months/years after its creation. (i think in compound with this suggestion page it could make sense to extend the time content can get curated).
* Making Hive suitable for more Users, because the more users Hive gets, the more confusing it will get to find interesting stuff with todays oppurtunities of following, scanning through communities and tags. I think most of you can relate to me when im saying that i go more through my recommendation pages on youtube, instagram,... then through my following list or scanning trhough tags/directly searching stuff in the search bar. This recommendation page is important to keep control over beeing user friendly and not "messy".
So in my opinion a recommendation page is really needed on Hive frontends.
What i would try to do:
I would like to do my best on resarching it for a suitable way to measure how we value content and scanning in which topics/kind of content a person is interested and out of this building an algorithm which is teachable and recommending you content you are interested in and you enjoy consuming.
But i commit, that im kind of insecure about how i would be able to manage this great project.
If im going to request money (HBD) from the Hive DAO i will do this developing process very transparent, because the DAO money is from all of us and in my opinion im kind of in duty to work community directed when receiving funds from the Hive DAO. So i thought a bit about how i could make my progress transparent and involve you guys in the developing of an recommendation algorithm for Hive Frontends.
Following is an Idea i got and which is to my knowledge much more community directed and transparent then other DAO funds requesting proposals:
Community developing
My goal for developing (and perhaps implementing) an teachable "AI" suggestion algorithm would be to work in some defined time periods and then showing you my work. I know that there are a lot of people in this community which have far more knowledge and experience in developing then me, so i will try to publish my work completely transparent and then getting feedback from you guys, which i will reward with some HBD which i get from the DAO.
(im going to publish not how much weight i gave every paramter in my evaluation, but i publish the basic model, which is enough for you to come with suggestions for improvement (and important for transparation). On demand i will give you more details, if you want to know more/develop on yourself.)
So basically my Idea is a "DAO in a DAO".
I will publish my work, collect feedback, proposed changes and then doing some kind of poll on which you all can vote for which proposed change is the best.I thought about how i should measure vote weight and was coming to the result of non stake based voting weight but "one person, one vote".To make sure nobody votes his one proposed change with bots, i would use an "reputation filter", for example only votes by accounts with rep >50 are valid.
Im completely aware of the laziness of people about voting 😄 so i planned to give a little monetary voting incentive.
I would suggest following payout from the DAO funds:
In total:5 HBD/day
for me:4 HBD/day
for the best proposed change:0,5 HBD/day
for all voters:0,5 HBD/day
This is my first coding/developing algorithm project and this project is something that needs a decent amount of time to be ready for implementation.
But i think, even if i am not going to be able to finish the project, the fact, that im going to share my working results with you is worth something, because this is which could help other developers which have the same goal, to avoid mistakes, or to use some of my algorithms to evaluate content and social connections.
But because its kind of unsafe for me, how my progress will be going, im going to make the first proposal periode only going 2 month long.So if we see, that this doesnt bring value we cancel it or shift it dramatically.
(i will develope this algorithm exclusively for Hive).
Whats going to happen with the HBD?
To prove you, that i believe in the future of Hive i will sell not more then 60% of the so earned HBD.
The other minimum of 40% will be used to power HP up, to beeing able to support you guys more, beeing able to give stronger vote support to proposed changes articles, comments and feedback.Also i will use it to give some delegations to newbies, to help Hive grow.(and spending some HBD to good feedback/proposed changes, that are not voted first place by the community (by you)).
I hope that this sounds good to you so far.
If you are my opinion, that Hive needs an "suggested articles" site i hope you will support my proposal.
If you dont like anything about it, please tell me, there is everytime something which could be improved and im very open to proposed changes as i planned to give them such a great value in developing this project.This is the official proposal in its basic form.
I will also try to support new Users even more as i telled you about giving delegations away, stronger vote support,...
Perhaps its good suited to tell something about my Blog and my goals for this plattform:
Im a student from germany, IT is a part of my study but its also about economics and law.
Im in crypto since January 2018 and active on Hive since may 2018.So over 2 years on Steem in which i was permanent active with writing posts (which i was doing in low frequency because my goal was to deliver high quality output) writing comments and voting.
My mother language is germany so if you found grammatical mistakes in this post please forgive me (and write your critic to @marcus0alameda who helped me delivering a post with no grammatic issues:) die Verweisung auf dich kannst du auch löschen wenn du willst:)), im trying to work on this to beeing able to better interact with the whole Hive community which is mostly able to understand english. This i the reason i write this post in english not in german.
I really have a lot of ideas to build this but to be fair i repeat that im not able to promise you that i will develop this algorithm and its obviously that there are a lot of people out there with more experience and knowledge, which is my reason for requesting a relatively low daily pay for me (just look at some other proposals).
But i think its a fair price. But this is something we can talk about 😉
How are my plans to start this project?
At this moment i dont have the time to work multiple hours a day on this so im definitly not starting this project in the next weeks.Im going to collect your feedback and ideas on my ideas i presented you in this post and then working out my proposal entry post which will be more detailed and having an time frame.current status is, that at a 2 month proposal should have an post after one month in which im going through my work and requesting feeback, suggested changes by you.
I will collect them and publish them for example 3 days later which should be a good suited time frame for you to think about it and then you can vote which proposed community change sounds best to you during one or two days.
The most votet change/feedback gets the from me during the 1 month collected 0,5 HBD/day. The voters get the 0,5 HBD/day*number of days the month had divided by the number of voters.
Then in the "last month" im iplementing the won proposed change/feedback and of course developing new ideas and after this month publishing them again and then we can decide if this process was valuable and if it makes sense to make another proposal perhaps for a longer timeframe like 4 months.
I have to admit, that i really like this concept ( i dnont know if anyone else is doing the same, it crossed my mind without external influences.)
The update post after one month will be very well and i promise to give my best writing it and making it easy and fast to understand with selfmade illustrations etc. So no bullshit which nobody is going to be able understanding.
What i need to start this proposal
Is simply your support. If like 3 people answer me to this post im going out with okay nobody likes this idea or nobody cares about it and then im not going to push this project any further.
So i would be very thankful if cou could write a simply yes or yeah i like this this idea, if you would support my proposal.
If you think yeah basically i like it but parameter xy is crap (for example the daily pay is too high in your opinion) please write me and perhaps we can discuss things and come to a solution we both are happy with.
So please give me your feedback and critic.
Thank you.