Credit to my brother Hive account@manipulable, the following is a repost:

Tags are currently a bit crap

Anyone can add a tag to any article without moderation or validation.
There is no assessment of how closely a post relates to a specific tag
This means that the tag search is very poor, only showing highly popular tags in the dropdown, because tags are unreliable
Upvotes are vague
-- What is an upvote? Seriously… that's not curation. You don't walk into a library and see the books ordered best to worst from front to back based upon whatever people happened to read that day. Things must be organised based upon the content therein so that you can efficiently browse.

What LinkedIn does

The challenge for LinkedIn was that anyone could add any skills to their profile, meaning they were unreliable. So LinkedIn allowed other users to upvote those tags, thereby developing an intersubjective community verification system around people's skills, which was very successful.

Proposal: add +1 to tags on Steemit.com

Anyone can add as many tags as they like to any post, including the author
On reading the content of the post, other users +1 the tags
They will upvote the tags that they feel best reflect the content
-- Optional: adding a tag that gets a lot of upvotes can receive a small reward
-- Optional: upvoting existing tags that gain a lot of subsequent upvotes can receive a small reward
Search will then be useful, a user can search for a tag and reliably find articles related to that search
Search algorithms can then be designed around these new data points
Searches can be based upon multiple tags
Search results for tags can then be sorted like in the existing standard feed based upon Reddit
Search can be predictive
Search can offer you related tags to include to widen your search
Search can exclude specific tags

Final thought

The tag system should be pretty straightforward to set up
The search will be complex but doable, for example
- One article might have 10 upvotes on a specific tag, but 200 upvotes overall
- Another article may have 50 upvotes on a specific tag but 100 upvotes overall
  --How do you determine which search results to show