In this article im going to discuss why in my opinion the lack of a possibility to show users articles in which they might be interested (old and new posts) is one of the biggest issues in Hives usability.
Having this feature integrated in all hive frontends would have following benefits
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 is "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 it! 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 timespan in which 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 through 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 planned to do about this
I had the goal to develop a recommendation algorithm for hive frontends financed by the Hive DAO with following model:
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.
But now i dont think that i have the time to work this out unless im going to reguest a very high daily pay of about 15HBD/day splitted in:
| In total: | 15 HBD/day |
|---|---|
| for me: | 14 HBD/day |
| for the best proposed change: | 0,5 HBD/day |
| for all voters: | 0,5 HBD/day |
But from the background that i dont have much experience in developing algorithms and/or implementing it in java im not sure if this would be the best deal (although i like my system of giving you some power in the developing process).
What i want from you
Basically i just wanna know if there is somebody developing an algorithm like this.
So that i can support him, because this is a very important topic in my opinion.
What i can do
I could write some posts about how you can classify content and how you can recognize which content could bring value to a specific user out of his behaviour on the Hive Blockchain.
My blog is kind of calmed down the last weeks, because to be honest my motivation for writing has fallen and i have enough other things to do.
Nevertheless i vote every day and think about Hive projects.
Thanks for your attention.
(sry for grammatical errors, english is not my mother language).