While reading this Rediff.com story about Maruti Madhavrao Phad, a Maharashtra government employee who got injured during the recent terrorist attack on Bombay, I noticed something.

It is not related to the story as such. If you scroll down to the end of the story you’ll see the credits for is particular story. It reads “Image: Maruti [Get Quote] Madhavrao Phad at his home. Text: A Ganesh [Images] Nadar. Photograph: Uttam Ghosh


See the screenshot of the line. Notice the yellow areas? These are smart tags which apparently the engine parsing the code of the pages inserts to “enrich” the user’s browsing experience, by providing additional information related to the content the user is browsing through at the moment.

Note the word “related”? Now see what the yellow highlights in the image show. “Get Quote” for Maruti. Here Maruti is the first name of the hero of the story, not the name of a car-making company. Yet, the Rediff engine treats it as the company’s name, and is offering you stock quotes for it. And for Ganesh, the link is a Rediff search link with the string “Ganesh” – marked “Images”. Which means that it would return images of the Lord Ganesha, and other celebrities called “Ganesh”. Not images of the author of the article in this case, A Ganesh Nadar.

And both these tags are intrusive, they not only break the text they are placed in, they break proper names of people. Intrusive and irreverent. Considering the tone of the article, even more so.

Were they really necessary?

5 thoughts on “Smart Web – How Smart?

  1. I agree with you, but unfortunately the algorithm cannot differentiate between one page and another basis what mode/tone/situation they are written in, unless you put an identifier there…

    While I agree to what you are saying and that it can/should be done, i am wondering if someone would be keen enough to do it !

  2. “i am wondering if someone would be keen enough to do it !”

    that keenness is the distinction between ‘just doing my job’ and ‘we will do it the best way possible’.

  3. Amit, its not that case. You just presented why there is a need for Semantic web (or web 3.0). The machine cannot differentiate between blue lotus (the car) and blue lotus (the flower). That logic is yet to be built in!

Leave a Reply