The main inspiration for this idea comes from the fact that there can be better and cheaper alternatives for ads by suppliers in the market so that they don't get lost in the big place that the internet is and moreover, many big social media companies have ventured into the retail market so how can Twitter do that but in a more innovative way.
It tracks tweets with ! and a product name attached to it searches for the location in the tweet. Then a notification is sent to the supplier of that particular product in that area. To ensure that the Twitter customers get the best quality products, suppliers have to affiliate themselves with Twitter and register themselves on our app called Tweetmarket. Twitter can then do a background check on them. Moreover, a sentiment analysis feature is included in the web app so that the suppliers can know what are people talking about their product. Suppliers get in contact with the customer and sell the product.
The backend used for the project is Flask and the front-end is built using React. For the sentiment analysis part, Twitter API and a machine learning model has been used.
We faced challenges regarding deployment, integration of the ML model with the web app. Moreover, five of us are from three different time-zones, so holding meetings and discussing everything and involving everyone in them was also a bit tricky but everyone worked hard and made adjustments and we could complete our project.
We worked hard and figured out how to make a connection so that we could use React as our front-end and Flask in our back-end and I think that was one of the biggest accomplishments for the team.
I personally learned a lot of new things while working on this project, and the same goes for the whole team. Everyone stepped out of their comfort zone, made adjustments and we worked as a team and had a lot of fun hacking.
We plan to take it to a bigger scale and work more on security features regarding our project. Next to that, we want to add more options to the home feed for the business to contact the tweeter. We also want to add a mood dashboard so the business is able to see what the sentiment of their product is at the moment at their location.