Restaurant Advisor

Find a restaurant according to the current time

Original Design: Imagine your first class has just got over and you are super hungry! You want to find the perfect place to have your breakfast, lunch, or dinner. You are trying very hard to think about a place, and you are also searching it on google. But than you realize you also have a class after 30 minutes or after an hour on the other side of the campus or anywhere. Now you need to find a restaurant that isn't just near to your current location, but it should be at a perfect place from where you can get to you your next class without being late. This food scheduler would help you find a restaurant that matches your preference and fits perfectly in between your class schedule.

What we got so far: A very early prototype with python that can take a user time and location based on ip address and the machine's time. Unfortunately, we didn't have the opportunity to generate data that depends on all the variables, so we focus on using the time.

Challenges: Originally we thought our idea is fairly simple and we can get most of the things done with libraries (the prediction can use a ml library, time is usually built in, and we can use geolocation from google maps and a schedule from google calendar). As most of us are mostly experienced with Java, we went ahead and tries Java. However, Java poses quite a bit of challenge as their libraries are a bit hard to use. Furthermore, using google API prove to be quite challenging for us as we are unexperienced with it. We spent a large chunk of our time trying to work with Java and was unsuccessful. At some point we decided to use python as some of the libraries are easier to use, and we found a way to geo locate using IP address instead of google map. However this still pose the problem of using a calendar, which we decided to drop in the end. Another challenge we face is getting data. To solve this we tried to make some randomized data that is biased so that there's a pattern in the data. This also prove quite challenging as we have to have different bias for different variables. As it is hard to keep track of all the variables, we decided just to keep track of the day of the week and hour of the day, and randomized all the other input.

What we learn: Although it is unfortunate that we weren't be able to achieve our expectation for this Hackathon, we believed that the experience of developing and prototyping a project rapidly is valuable to us. We also learned to better estimate our abilities and hopefully we can develop a much more feasible idea for our next Hackathon (which hopefully also produce a much more satisfying result)