A few days ago I announced a new Word Cloud report on Hive Analytics. The first draft of this feature stopped a lot of common words, but still got hung up on some meta data in posts that were not properly getting filtered out.
I released a new update today that greatly improved how I filter out html tags but also adds more common words to the filter. You should see a much improved Word Cloud with these improvements. I still plan on making improvements to this feature so it will just get better.
Before
After
You can see this example using @beaker007's account is a lot more focused on content. Things like centerxcenter, centercenter, and classphishy no longer pollute the word cloud.
You can now easily share your word cloud on PeakD snaps by pressing:
You can check out your (or for anyone else) word cloud at https://hiveanalytics.usehive.com/word-cloud.
I am currently working on some improvements for the Curation reports. I will likely consolidate the three reports into one by improving the Curation Leaderboard to have all data.
I also want to improve Word Cloud feature by having multiple choices for how aggressive you want common words filtered out. I want the word cloud to really show what a user is talking about without losing their personality. I do see value however in aggressively removing words that don't offer a lot of value.
I also want to start adding in some of the Hive Engine reports I have planned.
If you have suggestions on words you think should be removed from the Word Cloud, feel free to use the feedback feature or leave a comment.
There are over 250 unique users using Hive Analytics!
Hi @themarkymark I think it would be interesting to cross-reference the data from these two tools:
https://holoz0r.github.io/HiveReportCard/
https://hivestats.io/
What do you think? I know you like data because you said so before.
I have no objection to working with Marky, but I think our tools diverge significantly in functionality.
Mine is focused on single user insights, whereas Marky's is much more robust, better implemented, requires a back end, - and focuses on ranking users, utilising all available data. Mine is not as sophisticated technically speaking.
Mine runs entirely in your browser (and is slower as a result) instead of having a server it talks to every time a query is made.
Edit: I was also going to (Eventually) build a word cloud function, but Marky beat me to it, so I don't see the need to duplicate functionality.
I agree with @holoz0r, two different tools and we need them both.
Have you thought about multi-lingual posts already? I think every posts has to be language filtered before doing the word cloud. Many hive posts have 2 or more languages in a single post.
Not really, the amount of work to make that happen just isn't worth it imo.
It looks like my hyperlinks to various platforms in my post footer boilerplate are included. Is there a good way to filter out anything with
https://*or maybe
to pare down this kind of result?
I am already filtering urls, but I will see why they are still coming through.
When i click this: https://hiveanalytics.usehive.com/word-cloud I only get this word cloud - probably the wrong one :-).
I'm no longer working on it.
hang on - not working on "word cloud" or the entire "Hive Analytics" thing? Both same error - would be a shame this project is not moving forward - did i miss something?
Entire project, it didn't get funding so I moved on to other projects.
Understand, not sure i voted (think I did) but thought it takes a while - sad though, i liked the project and you did not request a lot compared to other DHF proposals
It was purposely unfunded, so I just can't justify the time involved.
i see, thanks for the headsup