ResearchBuzz

One day I’ll write about how frustrating I find searching Amazon. It’ll be a bitter, gloomy post. Ingmar Bergman will turn it into a film. The climax will occur when I discover that Amazon has discontinued e-mail alerts for new products and I throw hummus all over my kitchen. Watch for it. In the meantime, I do have some good new for those of you who want to monitor Amazon for new products — RSS feeds for Amazon tags.

As announced on the Amazon Web Services blog, you can now build an RSS feed for tags. The blog post gives details about how the feeds are set up, but here’s a brief example:

http://www.amazon.com/rss/tag/giraffe/new?length=100

Giraffe is the tag for which you’re searching. /new indicates that the search is for items newly tagged, and ?length is for indicating how many items you wish in the RSS feed. You can also add an associate tag to the feed, see the blog post for details.

It’s interesting to see what ends up tagged; it’s not items that have the tag in the title (necessarily). The RSS feed above generates items like the Paper Creations: Wild Animal Origami Book & Gift Set , which makes sense and Animals of the World 2.0 (CD-ROM) . There’s even a giraffe shaped like a rubber duck, which you would not, by its description (”Stretch - Rubber Duck by Rubba Ducks”) immediately recognize as giraffe-oriented.

I will be making use of this, though it’ll take some experimenting to find the best tags.

This post came from ResearchBuzz, a site with news and information about online data collections. Visit us at ResearchBuzz.com .

Like tags? Want to see what all the fuss is about? Check out TagBulb at http://www.tagbulb.com/, which offers tag search in a head-pounding number of categories.

Yes, head-pounding. No less than fourteen categories of tag search — yikes. I did a sample search for engineering to run through all the searches. TagBulb has all the categories in a line but the search results start with the second category (videos) instead of the first (images). This is annoying. Anyway, the kinds of results you get depends on the category — images and videos give you thumbnails, products give you product shots, blog search gives you plain results, etc. For each category the sources are listed across the top of the results. Click on a source to get those results — they’re not lumped together.

Across the very top of the search results, you can get some related tags for your search, a link to a tag cloud, recent tags, or popular tags. Both the recent tags and popular tags are absolutely not safe for work. If you want to see all the tags — TagBulb calls this a tag universe, and it is too big to be a tag cloud, check out http://www.tagbulb.com/TagUniverse.php . Again, not safe for work.

I would use this engine for more of an overview of what’s out there, as it isn’t comprehensive — it lists only two blog search engines, for example, while Zuula lists four (and of course Zuula isn’t comprehensive either.) You’ll also want to be aware that sometimes you’re not told how things are being sorted; Google Blog Search simply sorts the items by relevance, you don’t get the option to search by date. On the other hand I was made aware of several search resources I didn’t know about using this site, I like the thumbnail layouts, and the results were very quick.

I don’t think this is being done intentionally, but more and more I’m coming across tools — mashups, tag sites, or uses of Wikipedia data — that are increasingly making searchable subject indexes irrelevant. Makes you wonder what’s going to happen to Yahoo’s Directory, the Open Directory Project, etc etc. The latest example of this kind of site is Delexa, a site that mixes del.icio.us tags with Alexa data to provide information on the top 50,000 US Web sites (as ranked by Alexa.) Check it out at http://www.delexa.org/ . It’s in beta.

You can search this site by tag (supplied by Del) or search by domain name. The first thing I did was search by tag for Wii. I got a list of search results that showed the Delexa rank (how relevant it is to your search results), the US rank (which I think is set by Alexa) and the tag rank (which I think — there’s not any documentation I can find — counts the number times the resource has been tagged for the keyword for which you’re searching). You can reorder the list by any of these three ranks. At the end of these three ranks is the site URL.

There’s a link to take you directly to the site, but if you click on the URL you’ll get a page of information, including a page snapshot, tag listings, and links to a whole bunch of data, including Alexa and Del pages, a link to pagerank prediction, links to a wiki about domains (About Us), and site age. (That’s the result you get if you choose to do a search by domain instead of a search by tag. Remember this site ranks only the top 50,000 site — you’re not going to find every domain here.)

There are a few problems with the site. It doesn’t seem that you can do multiple word searches, so pick the single search term carefully. There’s no result count that I can see. On the other hand it’s interesting to how high the rankings are for the various words you might choose, it was nice to use Del tags to search a predefined set of sites, and I saw nothing in the line of spam results. Worth a look.

(Do you think they could create another list of tagged customer complaint sites and call it Dislexa?)