more thoughts on this
 

I was browsing a blog earlier today and they had a tag
cloud. I could easily see what were the dominant themes
of the blog without having to read a single entry.

A relevance list would be useful, but it's not the
same. Really - if you think about it, you know this is
true. A relevance list as discussed in this thread
presumably shows you the most relevant ITEM TITLES -
which may or may not reveal the most dominant themes.

Let me spell this out to avoid confusion: I could, in
principle, give each item an arbitrary title, like
asdfsdf and jlkjlkjl. And the content could be ordinary
content, like an import from various websites. So there
is no logical connection between an item's title and
its content. The only connections are by intention.
Period.

But a tag cloud wouldn't care. It simply reads the
entire content and reveals the dominant themes, no
matter what I titles I give the items.

And clicking a cloud item would, I hope, reveal all the
Explorer Pane items associated with the tag cloud word.

I would personally find this tremendously helpful -
mainly after import of lots of new content - but in
general as well.

Regards,
Bal

I'd also like to (for now) vote against this suggestion.

Reasoning:
I see the tag cloud as a feature to find trends in the UR data. Which sounds cool, but doesn't do much for the day to day use regarding my use of it as a PIM application.

In other words, I'd much rather have improved functionality for i.e. finding individual items (and a dozen+ other things) than having a feature that enables me to find trends in my data.

Exception to the above against vote: If the effort required to implement this is negligible (which seems unlikely).

"trends in the data" - yes. that's one of the things I'd like. But trends isn't the only thing that this can do.

I'm not going to belabor this more. I've said about all I can say without repeating myself, and I think I've done that plenty (with apologies).

Let's just say that your mileage may vary... :)

Regards,
Bal

I was just about to post this as a suggestion.

I don't see tag cloud as a feature to find trends in UR data, but as a navigational means, and would find this feature very helpful.

Wow - fairly old thread, but since TMF brought it back to the present with a post, I though I would chime in too.

I like the idea of a tag cloud. Not a vital feature, to be certain, but certainly a nice feature to add that actually would not take any serious amount of time and should not have any effect on speed at all. (At least having adding a similar scripts to do this in another application, there was no measurable change in speed at all).

Another way to look at it is like the intersection search capability in Evernote, at least Evernote 2.2; they have stripped any serious search functions from the new version 3). Clicking on two Evernote categories in succession gives you an extremely fast primary search of your notes database with a secondary search of a category within the first's results. And being recursive, you can click additional categories to further stratify the data.

I did notice that some users mentioned earlier in this thread that there are ways to create keyword analyses that do even more than the relatively simple division that a tag cloud would do. It would be valuable - at least to me - if someone could expound on that a bit. I am somewhat illiterate in that area as of yet but would love to read and learn!

Thanks!

Jim

Glad to see this topic independently revived. I'd still love to see this implemented.

- Bal