Ever wondered how one creates the tag cloud which we see very often on delicious, technorati, amazon?
Parsing the text and storing it in a database and then highlighting the text based on their frequency. Sounds simple..But what if you were to be given YAHOO’s parsing algo…So, here’s the way to go:

  • Grab newsfeeds that you define. Use caching to make sure that they are not fetched too frequently.
  • Stuff them into a MySQL database. We have to be a little tricky here so we are not storing duplicate articles, even if the same article appears in several feeds. MySQL came to the rescue here.
  • Next, we call on the Content Analysis web service from Yahoo! to extract the keywords from the article titles and descriptions. The keywords that we get back are inserted into the database and associated with the appropriate articles.
  • Everything else is gravy. We pull out the popular keyword tags and use a silly scaling function to assign css according to how often the tag shows up.
  • Clicking a tag’s link will display all the article abstracts for that particular tag as well as display the other tags associated with each article.
  • If you have the time to implement, go ahead and do it for all your emails, files, etc..What I meant was one can easily go ahead and build a desktop tagging tool which when clicked shows a tag cloud for all the files in one’s system. I think this should
    be the immediate plugin needed from google desktop search service.