Make your own blog and tweet clouds
Wouldn’t you know it, if I just would have waited a wee bit longer some smart people would have saved me hours of time. For those of you following I started cut-and-pasting some people’s blogs (male thought leaders and female thought leaders) into Many Eyes in order to create more accurate tag clouds of what they were talking about.
From there, JP Rangaswami asked me to create some Tweetclouds for him, so I did. But wouldn’t you know it, a couple of interesting website have sprung up so that will do all the work for you.
TagCrowd
My man Ken Burbary (@kenburbary) tipped me off to TagCrowd. You can either simply type the URL of any website in or upload a file and it will create a cloud for you. I tried it on a number of site and while it did create a nice looking cloud, obviously the entire site isn’t crawled. The clouds I looked at were clearly limited to the one URL I pointed to. Here’s the one that got spat out when I pointed to this blog.
TweetClouds 
Seemingly more accurate is Tweetclouds, which pulls the text from your Twitter archives and creates a cloud for you. I liked that there was an option to omit the directed comments to others so that the analysis was limited to only original statements. I was sorta surprised that “thanks” and “love” were words I used a lot. On the downside, I’ve tried getting to Tweetclouds’ website for about a week and it seems to always be down. So, as with anything, your mileage may vary.
This stuff should be built into blog software
There are two things missing from blog software:
- Auto generating clouds like this stuff.
- Auto blog rolls which show who you really frequent.
These sorts of clouds (and ultimately blog rolls) are so much more truthful and valuable. I hope there’s even more innovation in this category.
Things people have said about this post
Very cool. We actually just discovered TagCrowd over here at FG2 and my boss’s most recent blog post makes use of it. Also, on the auto-generated blog roll, it seems like there are some people starting to get close to that. Take MyBlogLog for instance, I just recently noticed that any sites I frequent that use MyBlogLog will automatically get added to my Communities page. Now if I could grab a feed of that list, I’d be set!
I agree that this kind of visualization gives us a new way to think about our words. It’s powerful to me. Not sure how I’ll adjust related to it, but it’s interesting.
Agree with Chris that this kind of visualisation is powerful. It should be possible to share this cloud, and have it updated in real time, on Netvibes etc. It would give an indication as to the content of somebodys tweets, and perhaps give an indication whether they are worth following or not.
Sam - like Brad, I just discovered the TweetCloud yesterday. Obviously, it’s a very cool concept and builds on the even more powerful concept of tag clouds in general. The reason I like clouds is that they provide for a fourth way to navigate/digest content (search, left/top nav and center-page content being the others.) Given the fact that I’m at about 3,500 tweets on Twitter, having a way to go back and get a sense of what I talk about most is quite helpful. What would be a neat experiment would be to screenshort our TweetClouds now and then take another snapshot in 6 months to a year to see how they differ.
Happy Weekend!
Aaron (@astrout)
I really like to have a blog roll cloud. It would so useful and truthful of what is really going on.
I cobbled together something in Movable Type that gives me a list of websites I have referenced most recently. This is a “recent blogroll” of sorts, but it doesn’t feel very slick. In MT, it uses MTCollect (http://www.staggernation.com/mtplugins/Collect/) to list recent links and then sort them by frequency. I think there are some examples that do this.
Totally agree this functionality be built into blogging software. And, it should be built into Web analytics software, too. Can you imagine your tag cloud of content versus your readers’ tag cloud of interest? Oh my. That would be incredibly valuable way to visualize what you like to say versus what people like to read.
On a side note, I also wanted to mention that a personal favorite of mine, TweetStats, has recently added a tweet cloud tab, as well.
Tag clouds give an interesting perspective on what we’re talking about. Weighting of words with size and bold gives an even richer feel to what is seen.
I gave Tag Crowds a try on my blog and it ran fairly quickly, though only choosing 50 ‘tags’. Running Tweetclouds was slower but didn’t effectively grab too many more tags. See my Tweetclouds here - http://tinyurl.com/5j6z97
While I change what I write on Twitter or the blog based on the tag cloud? Not immediately.
I did realize I @ message significantly more than straight messages. That might be something I look at. On the plus side, Tag Crowds shows I tag a lot of blog posts ’social networking’. Perhaps rethinking the variety of tags might be in order.
Thanks for pointer to Tag Crowds.
[…] Sam Lawrence looks at this as well. Earlier, he created my one and two-word tag cloud based on this blog. Again some interesting combinations: […]
[…] tags, Tweetclouds Sam Lawrence over on GoBigAlways shared the coolest thing in a recent blog post. He told us about TagCrowd. Where TweetClouds creates a cloud of Twitter words, TagCrowd makes a […]
[…] Lawrence over on GoBigAlways shared the coolest thing in a recent blog post. He tells about TagCrowd. Where TweetClouds creates a cloud of Twitter words, TagCrowd makes a […]
Say what you will about Livejournal, but they’ve given me a tag cloud for YEARS now.
It’s too bad it’s all going downhill fast since they sold, despite their all star (never consulted) board.
Tag clouds: Visually charting our written content…
The value of tag clouds is that they give us an excellent visual representation of the words, phrases, or tags that we’re using most often. You’ll often see them on blog sidebars, or on folksonomy-built sites such as Flickr and Delicious.
M…
These are nice applications, but don’t use the concept of a tag in the right way, IMO. A tag should be chosen by a human to be of maximum use and to aggregate effectively into a folksonomy. It should denote a descriptive summary word, ie a classification, not just a word that happens to appear (even frequently) in the text.
[…] with our good friend Twitter. Bryan Person has a good summary (he was writing based on a post from Sam Lawrence) and his own Twitter Tag Cloud, which you can make at TweetClouds. Here is […]
I recently created a YouCloud web application to generate tag clouds based on the tags a YouTube user has used in all their videos:
http://www.williamsportwebdeveloper.com/YouCloud.aspx
I could also create tag clouds based on any information made available by XML or JSON data feeds.