posted
Have you seen this yet? Rather silly, but fun, especially if you type in a bestselling title like "The Davinci Code" which rates poorly. So much for science.
posted
Well, to be fair, there are a couple of firm constraints on this tool.
First, the data-set it looked at was best-selling works of the last fifty years (more or less) and a control group of works written by the same authors. So it only predicts how attractive the title will be on a work by a bestselling author of the last half century.
Therefore, titles like "The Bible" and "The Quran" are evaluated as if some new book were published under those names. The Davinci Code is an understandable exception, the book made it entirely on hype and contraversy, it had nothing else going for it (other than the side-splitting but sadly unintentional hilarity, of course).
More importantly, it only takes the works of known best-selling authors into consideration. So if you aren't already a best selling author, then this tool is of little use to you.
Also, the titlescorer has only been tested (indeed, can only be tested) against the exact same data set used to generate it in the first place. You need to plug the title of a best selling book into it to check the validity. Even at that, it only claims 70% accuracy.
I think that it would be more fun to find out what the rules are by which it scores a title, but I don't see a link to that information. Perhaps google....
posted
Dude, did you actually read The Davinci Code? (Saying this with the awe in good vs. Awesome as opposed to injured shock).
Posts: 366 | Registered: Sep 2006
| IP: Logged |
posted
Well, I just typed in "Plag", claimed it was a figurative adjective, and got a 69.9% chance. Maybe I should change the title of my book. . . .
Posts: 932 | Registered: Jul 2001
| IP: Logged |