Ask NH: please evaluate my idea, math exercise solver | Hacker News Reader

Ask NH: please evaluate my idea, math exercise solver

It's been a while (some years) that I have had this idea in my mind. Now I'm considering to implement it, but I still have some doubt.

The idea is to build a webapp to solve math exercises aimed to high school (and maybe lower) students having difficulties learning math. The app solves exercises showing passages and explaining the rules used in the solving.

It's not a project where I can release a prototype quickly, because it requires some core functionality to be useful (involving parsing and some AI) as well as some kind of interface capable of printing math on the web.

My doubts are about the market. I know a lot of people have problems understanding math at school, but I don't know if students are a good target when it comes to monetize.

Any feedback on this is really appreciated.

View on HN

@DeusExMachina: Where I come from (India), poor parents go through considerable adversity to send their kids to good schools. Young adults of disadvantaged opportunity (and this is an emerging nation of immense socio-economic disparity) are enhancing their skills - one course at a time - while at work. This is the story of a few billion here, and I can safely say it also holds true for much of South Asia, China(?), Africa and South America. Ours is currently a great world beyond just social security and health insurance - ours is a great world beset with great inequity and suffering - where the only recourse for the underdog is to obtain meritocracy and education. Our world is a story of billions of underdogs. In my lay opinion, your product sounds great, and we are looking at billions of potential users - who will inherit our world and will run it. Every minute. Do not worry about a market. The market is already here.

I would have paid for this when I was in high school. Maybe up to $15 or $20/month. When I found Maple in university and the step-by-step demonstrations of integration and derivatives, it was a godsend. I could pinpoint the exact step I didn't understand and work from there.

Consider targeting specific standardized tests. Parents seem to pay good money for their kids to have an advantage on standardized tests.

As far as prototyping, it could be a matter of taking something like Sage, Maxima, or Eigenmath and adding some hooks to output intermediate steps and text helper messages. They all output LaTeX IIRC, so you just need to hook that up to a latex-to-png program to render. You can target a specific problem like derivatives or simple integrals, and run some adwords for it to see if there is any interest.

Starting with derivatives and integrals was exactly what I was thinking about. What surprises me is the pricing: I was thinking about something lower.

Thank you for the feedback.

Well, I was probably an unusual case. Very interested in math, very bad at learning from teachers, and earning a bit of disposable income. I ended up spending about $50 one term on math learning software that turned out to be total crap, so spending that on something that worked wouldn't have been out of the question.

I'm confused about this idea. Will your app be used as a teaching aid? I think that's hard, because not only it has to solve the problem at hand, but it has to solve it in a pedagogical way. If it just generates the answer without teaching teh student, it will meet with a lot of resistance.

Even if you solve this problem, chances are that many educators will oppose this idea, unless all their students have access. It would be unfair for other students who are not rich or tech savvy enough to use your app.

High school students would be a hard market, I would think, unless you app is viewed as a cool "cheater app". A better approach is to go through learning centers catering to students, e.g. Sylvan Learning Center or direct marketing to bigger and more affluent high schools.

What I'm thinking is exactly something pedagogical, an app that shows you the passages with explanations on why and what rule was applied, with different rules for granularity.

I know it's not an easy app to implement, maybe I'm too naive but I think I have an idea how to do it, applying some knowledge I have in AI as well as math. To me the kind of exercises you do in high school have always been easy to solve, seeming mechanical and algorithmical and suited to computers. During university I worked with theorem provers too, that seem to me much more complicated programs than this, giving me a sense of feasibility. Maybe I'm just mistaken, but I would like to give it a try.

I don't understand why or how educators could oppose the idea, since it would be something that students would use at home to do homework. I'm thinking about an easy interface, so few people will be cut out and I'm thinking about a low price point, so I hope I will not cut out poor people.

My aim is education, but I understand that it could be used for cheating too. But cheaters find ways to cheat anyway (this has always has amused me, they are really creative on this), they are not my target.

Just so you know (competitor analysis?) there already exists something like this: http://www.mathway.com/ is one. I think there are others too (search for 'math solver').

I think it's an interesting idea but I don't know about how good a market it would be. The price would be low, and it is a commodity product, so you would have a hard time separating yourself from your competition.

Also: "I don't understand why or how educators could oppose the idea, since it would be something that students would use at home to do homework." This is exactly why educators might oppose the idea, since students could simply load their homework into your site and be given the full solution to copy out.

Don't get me wrong, I think it's a great thing to do. I just don't know if you can make a business out of it. It might be better to just give the thing away - it would be a great demonstration of your skills as a programmer. You could, for example, create it as a python module (or whatever language you're working with), release it open source, and make a 'homework helper' site as a demonstration of what it can do (while also helping students, and not just the ones with money to spend).

I don't know if you've seen this, but WolframAlpha now shows the steps to a solution in human-readable form. For example: http://www.wolframalpha.com/input/?i=integrate+xe^x+dx Click on "show steps" by the solution.

Reply on news.ycombinator.com