Slashdot Mirror


Amazon Tries Its Hand at Tagging

Kailash Nadh writes "Amazon has formed a 'tags team' and has begun using tags on some pages. The idea, apparently, is to slowly experiment with tags and to give users some power over how certain Amazon products - books, for example - are categorized." From the article: "Ultimately, this is interesting because it may well prove to be the most visible example of a company incorporating tags as a way to bring order to information. Outfits like Flickr are big and have tremendous followings, but nothing compared to Amazon's. And if Amazon can make a go of tagging, that may finally be the tipping point that makes the technology something every Tom, Dick and Harry knows about."

1 of 145 comments (clear)

  1. Re:Appeal to a bigger audience by ergo98 · · Score: 5, Informative

    Its nice that the author assumes we know what tags are. It creates an article that only people who know whats going on already understand. Otherwise you go tag? What kind of tag?

    Parent Post Tags: clueless "karma whore" "obvious question" :-)

    Seriously though, tags are user-provided categorization (including multiple "categorizations" given that you can apply multiple tags) of content. e.g. Search on Flickr for all photos that are in the union of the tags Toronto and Girl and you'll get photos that have those two tags. The same concept applies to delicious. This is the so-called folksonomy in action, where us lowly serfs categorization content, rather than "the man" in a taxonomy like Yahoo.

    However the tag thing is going way too far (as are most "Web 2.0" things) - tags are useful in the absence of a superior classification system. For instance we tag photos in Flickr only because the system can't, thus far, determine what the photo is about mechanically. If it could automatically classify photos, then this folksonomy would prove terribly dated, unreliable, and inaccurate. Look at Google - what is better: The META keywords technique of before, or actually contextually placing each page based upon its actual content?