Given the increasing number of social media platforms, it comes as no surprise that friendships have become more complicated. Never before has befriending someone from a different state, country, or even continent, become so easy. With this low barrier of entry for friendship, identifying true friends has become drastically more challenging. The average Facebook user, for instance, has 338 friends, an overwhelming number of relationships to maintain. Countless posts and messages between friends on social media platforms renders monitoring the closeness of friendships a near impossible task. As such, there is a need for a platform that quantifies the closeness of friendships and provides actionable insight. In an age of information, knowing your closest friends is indispensable. With this knowledge, you can understand who your best friends are and why you two share a close relationship.

What it does

You first grant FriendTrend access to your Facebook profile so that FriendTrend can analyze your friends list. Through our unique algorithms and logistic regression model, FriendTrend calculates a probability that represents the closeness of the friendship for every friend. The model assigns weights to certain variables such as the frequency of interactions and likes given. FriendTrend also employs sentiment analysis to analyze messages on Facebook messenger to gauge the positivity of the friendship. After completing these analyses, FriendTrend displays a list of your friends by closeness, displays the probability that you want to keep the friend, and shows the complete methodology behind the calculations for the probability.

How we built it

What's next for FriendTrend

We intend on expanding FriendTrend to include Facebook pages. By providing data on closest customers, businesses that have Facebook pages can identify core customer segments and loyal customers. Marketing has rapidly shifted towards the Internet, making social media an increasingly important marketing tool. FriendTrend wishes to continue contributing to the plethora of extant social media analytics tools out there.

Built With

Share this project: