TECH TALK: News Refinery: News Refinery Characteristics (contd)

  1. It should also learn my interests by seeing what stories I click on. By comparing my interests with others, it can (a) refer me to other people like me and their comments, and (b) point out stories which I may otherwise have missed.
  2. It should be able to make summaries of stories automatically so as to save me time.
  3. It should be able to compare stories so as not to give me multiple copies (eg. a story from a news wire like Reuters may actually show up in multiple sites).
  4. Headlines should also be pre-classified into a directory, allowing me to navigate by topic.
  5. I should be able to add a new news source easily – just by giving a URL.
  6. The software and intelligence in the system should adapt to my interests and learn from the collective. It should evolve.
  7. I should be able to place the story in “context” with other such developments.
  8. I want to get a sense of favourites – what are people reading in different parts of the world (like Amazon’s Circles).
  9. I want to be able to chronologically follow certain stories / topics, or even writings by an author.
  10. A “Similar Stories” Search: show me other stories like this.
  11. Repeat specific “Searches” once a week and send me all new stories.
  12. I can pick up articles I want to read, and these can either show up on a single page, or can be emailed to me. On a slow link as I have, it takes 30-45 seconds per link I click on. (My solution to this has been to turn images off for browsing and open articles in multiple windows using the right mouse button.)
  13. A useful service: email me a specific news site page at a fixed time daily. Like I want the Red Herring top page, WSJ Tech news page, top page daily since its hard for me to search for stories on that page once they get “overwritten”.
  14. We also need to think of how this can be extended to multimedia news (images, audio, video) – these types are increasing.
  15. Imagine if every news story published had a unique identifier, like a book’s ISBN number. This would allow one to reference it easily forever. In some ways the URL is this code number. This code could also show up in the print edition at the end of an article, allowing easy access on the web in case I wanted to email it or download it for my future reference (like the physical acts of cutting and filing). This allows for closer linkages between offline and online.

Published by

Rajesh Jain

An Entrepreneur based in Mumbai, India.