Endorsement Recommendation Using Instagram Follower Profiling

In the fierce competition of product sale, in recent years product owners have used the services of Artists and Celebrities to co-promote the products they sell called endorsements. One way the celebrities promotes the products they endorse is through various online social media like Facebook, Twitter, or Instagram. However, there are many cases where product owners merely choose any famous celebrities to promote their products based on the fame of the artist without a proper analysis. We think that it is important to consider the chance of sales success viewed from a match between promoted items and the tastes or likeness of the artist's fans. In this paper, we propose a method to automatically recommend what types of product to endorse for an Instagram user based on his/her followers' profiling. We use Instagram to perform a user profiling to determine what the follower users like based on annotation tags of the images they upload. By using a clustering method to group the followers, we were able to provide an endorsement recommendation to a user that matched with the followers' preferences.