Ask HN: Why hasn't anyone done this sort of hack for (intro) textbooks?

(originally submitted for the $9.99 textbook thread, but it's already dead)

I've been tinkering around with the textbook/publishing problem in my head for a while (nothing concrete yet, sorry!) and am thinking of applying to one of the upcoming YC rounds. Might try to do this in the meantime as a pure hack:

Students hate being ripped off by frequent 'new' editions of textbooks, which basically just reorder the text slightly and shuffle the problems. (I'm aware quite a few people download illegal textbook copies, but I find reading on a computer screen difficult.) Why hasn't anyone hacked this system? You could just OCR the textbook editions, have the computer identify similar/identical bits, and create a database of dependencies. Then write a web app to query it.

Someone has the 7th edition and their class uses the 9th? Every time they get an assignment, they can go to the site and type in: "I need to read Chapter 8.2-8.6 and do the end of the chapter problems 7-30, odd. The class uses ____ 9th edition. I have ___ 7th edition."

Then the site shoots back, "OK, you should read pages blank through blank. Below is a list of the 9th edition problems matched to their 7th edition counterparts. It also tells you if there's anything different or new."

I'm sure the textbook industry would hate this and try to do anything to shut it down, so it's probably not a great startup idea. They've already taken rear guard action with clickers and web homework.

3 comments

[ 2.3 ms ] story [ 15.8 ms ] thread

jasonlbaptiste 16y ago ↗

Other idea is to build a network around everyone using a particular book. The popular books out there have thousands of students using them at the same time. There's probably a lot of value of connecting them all together. Notes, answers, questions, etc.

Fixnum 16y ago ↗

I haven't really investigated, but I suspect that in the introductory math/phys/chem courses you'll find that variable names/values also change in various ways. So in keeping track of what's "different or new", you end up not needing just dependencies but, realistically (in terms of reader convenience, but also due to difficulty of parsing OCR'd math problems to determine which are 'isomorphic'), all the problems from the latest edition. Which might not quite be 'fair use.'

I am not sure about other courses like bio or psych.

Good luck solving the textbook problem, though ...

realitygrill 16y ago ↗

I hadn't thought about the math OCR (nor 'fair use', obviously), since intro books tend to use very simple math. But your point with variable names seems especially difficult.. thanks!

For more verbose subjects like bio, or psych, or econ something like the software they use to check for plagiarism should work beautifully.