Hyper Local Social Platform

Client

Innovative social startup

Challenge

The client wanted a “Hyper Local Social Platform” on which end-users could search and find entertainment and other attractions. The app would collect check-ins, comments, photos and other information from different networks, which would not only provide end-users with information, but also provide venue owners with an understanding of customer behavior that would aid in creating marketing campaigns based in part on the number of visitors at a certain time, gender ratio, complete guest trend analysis and visitor feedback.

Solution

In order to navigate the big volumes of data and turn them into actionable insight, CoreValue’s Data Science team seeded a representative dataset by analyzing alternative data feeds and multiple social platforms. Proximity data was examined to follow processing requirements and minimize the potential for excessive compute costs.

The team used R for generating models and leveraged MongoDB in accordance to data size and scalability requirements. After an in-depth exploratory analysis, the Data Science team identified three key areas for initial modeling.

  • Gender prediction algorithm based on Twitter name
  • Model to identify sentiment of the text from blogs & Twitter feeds
  • Spatial Optimization Algorithm