Friday, October 24, 2003

Making research easier

This new Amazon feature where you can search the entire text of a huge number of books they have online is extremely cool, in my opinion. One thing that annoys me is that I cannot grep printed books as I read them. If I remember something that happened on an earlier page and want to go back to it, I can't always find it. This is annoying. Now though, I can search for it on Amazon, find out the page number and quickly find it in the book. This is great, and gets rid of one of the major advantages electronic documents have over printed books, which is that they are easy to search.

Of course, many of my books are British editions with slightly different page numbers. This leads to an interesting search tree approach, where you go to approximately where you think what you are looking for will be, type in a few words of the British edition from that page of the book, find out the American page number from Amazon and thus the approximate offset between the two editions that far through the book, then type in the phrase you are looking for, find where it is in the American edition, subtract the offset, and bingo, you have found it in the British edition. Not perfect, but workable. Still, some better system of indexing than page numbers needs to be found or invented.

I have lots more to say on the new Amazon feature, but for now I am going drinking with some bloggers.

Update: Virginia Postrel describes the new features as "a great advance for civilization". Steven Johnson (who mainly talks about the same use I was discussing - using the tool to search within books you own already) calls it "The best search idea since Google". I shall at some point be commenting on it in more detail, but I won't have time at least until the weekend, and I shall be flying to Australia between now and then. Ultimately, I want a single database containing all human knowledge that can be searched from essentially any computer terminal in the world. (In many cases I will be happy to then pay to use the knowledge, but I want the serach itself to be free). In the long run, I don't think this is out of the question. Just think how much closer to this we are now than we were ten years ago. This development is a major step forward towards that ultimate goal.

No comments:

Blog Archive