Tags and Tag clouds
Tags which in other words known as keywords provide a good means of capturing key terms within a content source (which in most cases is text based). Now there is no limits to one's imagination as to what all you can do with tags and tag clouds to derive inferences from an unstructured data source. One such imagination gone wild can be found at http://chir.ag/phernalia/preztags. This tracks the State of the Union addresses of US presidents to generate tag clouds. A casual researcher can have a quick glance to see how the government's policy has changed over the years. Now this can easily be extended to within domains having large unstructured data sources such as Insurance, Banks. Text miners can draw inferences on the type of claims filed by customers, the most common reason for claims etc. I shall write more on the advantages and disadvantages of this approach in the subsequent blog entries.
Comments
Thanks, Ganesh.
It is an interesting application of Tag Clouds. As you said, it has potential applications in a range of areas - business, law, education, publishing, library, electronic documents, and more.
From a technical point of view, it would be interesting to study issues such as computational requirements to generate a tag cloud, scalability, refinement of the tag cloud generation algorithm to respond to dynamically changing information (for example, articles in magazine over the last one year or six months, where information may be updated say every hour).
Posted by: Prof San Murugesan | February 27, 2007 09:30 PM
When you talk of the web, imagination has no limit.
Posted by: AjiNIMC | March 2, 2007 05:51 PM