Flowy

Analytics page using LENS data
Creating a new post (Image classification via GCP-Vision_API)
Feed of relevant posts
Login on the blockcahin

Inspiration

Social graphs can be very beneficial when it comes to optimizing the efficiency of recommendation algorithms. In this project we take this a step further. We try to exploit this representation for improving the efficiency of the underlining mechanism used to utilize the infrastructure more efficiently.

What it does

Flowy provides a dynamic load balancing for improving the efficiency of the social media platform. By identify clusters of users that have similar interest it can rapidly query the database and find content relevant to user's preferences.

How we built it

We used the LENS protocol to deploy Flowy on the blockchain and Google-Vision-API for image classification on users' posts

Challenges we ran into

We couldn't complete the entire project within the given timeframe but a good foundation has been set

Accomplishments that we're proud of

We have simulated our approach we have observed some fair improvement in information retrieval.

What we learned

A fair improvement in performance can be achieved if we manage to run this on a larger scale. This can also be very useful for platforms that use algorithms similar to TickToc where user's short interest span is taken into consideration for a more efficient recommendation.