What inspired us to create SmartCalc was the fact that the practical applications for science and mathematics are ever-increasing in our high-tech world, but today's students still fail to realize just why we need the higher-level classes. In failing to do so, students are having more and more trouble with word problems in the classroom which are meant to demonstrate these real-world applications. We hope that this program will allow students to have a better understanding of how the real-world applications of math affect their everyday lives, and perhaps once again instill a passion for science and mathematics.
Using Python and the Natural Language Toolkit built for it, we created a program that analyzes each sentence and word, and uses Parts of Speech tagging and phrase parsing to allow your input to be interpreted a bit more naturally by the computer, almost as though you were having a conversation with somebody. It then extracts the data it needs, does the necessary computations, and gives you your answer in a context that you can understand.
Natural Language Processing was something brand new to all of us, and we were a little worried once we finished the workshop once we realized just how deep we had gotten ourselves. But we refused to give up. We had taken notes, referred to just about every piece of reference material we could, and by morning, we realized that most of our code was made redundant with just a single line. All of those hours of research and experimentation, and it still seemed like we were at square one. But we pushed through, grabbed some caffeine if needed, and eventually got to a point where we could all be proud of the finished product. Considering just how new the field seemed to us, and how we were given less than 24 hours to learn and create something that seemed like Artificial Intelligence, we were VERY proud that we got as far as we did! We had fun doing it, too, which is always a bonus, and we made some new friends along the way.
What we learned is hard to sum up in just one list, but we can try. We learned that research BEFORE hackathons count, air-conditioning is not always available, time can be just as limited a resource as water, sometimes things don't always go your way, power naps are a smart idea, sugar-filled snacks are not, Ethernet will always be more reliable than Wi-Fi, verify your information no matter how confident you are with your sources, it's hard to start coding but even harder to finish when you're tired, and, of course, an entirely new system to program with.
Soon, for SmartCalc, we plan to make the calculations a little bit more robust, but certainly a lot more feature-rich, perhaps with high-level physics and calculus equations and word problems.