FoodNet: All The Reviews. In One Search.

UniHack UI Design

Inspiration

FoodNet was inspired by the need to find all the information of the best restaurant on one site.

Everyone wants the easiest possible path to finding the right restaurant. There’s no shortage of options when finding a restaurant review site and these days and FoodNet aims to help the consumer by aggregating all review data is one place preventing you from having to search multiple individual sites to get a complete view of all information online about the quality of the restaurant and other important information such as whether it has free-wifi or if its close to public transport.

What it does

Aggregates restaurant review data and meta-information in order to provide the user with all the information and filter options on one site. FoodNet aims to earn revenue via referral links to ordering on the restaurants website or delivering via services such as DoorDash.

We use data analytics to aggregate the different types of review systems and data to provide our own Median/Averaged list of the review sites to provide a complete picture of the restaurant without having to open a heap of tabs and social media sites.

How we built it

It is built on various web and cloud computing technologies. Backend: Python, AWS DynamoDB, AWS Lambda, AWS Api Gateway, AWS CloudFormation, AWS CloudTrail. Frontend: MaterialUI, ReactJS, Redux. Frontend Repo: https://github.com/Flumm3ry/FoodNetUi Backend Repo: https://github.com/thed24/FoodNet-API-and-Scraper

Challenges we ran into

Some websites do not provide APIs and scraping was necessary in place of that. We saw a good use of twilio and wanted to use Twilio to be able to call restaurant through it and make an order and we can see that in the full project. Couldn't complete full project and some features due to time issues. We were able to implement some user review scraping for filtering but all the filters generally dont work.

Accomplishments that we're proud of

Filling a useful customer need and creating a product we would legitimately use and could see being used. Being able to produce a website to search for restaurants ordered by rating in a very short time.

What we learned

How to use AWS Api tooling and how complex search engines can be.

What's next for FoodNet

More work into a feature where we provide additional filters gathered from user review data and restaurant information so we can add filters for things such as "serves alcohol", "accessibility-friendly" and use algorithms to discover other information such as distance from public transport. Also being able to find out the latest sentiment positive or negative toward the restaurant using Machine learning to analyse the attitude toward the restaurant from the reviewer when they have left a comment but not a star rating on some websites such as reddit.