부산대학교 도서관

Social media networking has turned out to be an essential factor for people wherein from sharing relevant documents to exchanging messages; everything is taken place via these social media sites. However, on social media, when a new user joins the group, (s)he must not be given access to all the previous messages. Hence it is necessary to predict the relationship between the users. This primary aim is to predict the link by taking into consideration using three criteria, which includes classifying sentiments from text messages, recognizing faces from pictures and videos posted on social media to find the relationship between the users. Various algorithms and methods were studies for this purpose wherein neural networks can be used for predicting the relationship between the users.