Rede Social Gamer com modelagem do algoritmo K-means para indicação de novas amizades

Abstract

Social networks for the Games domain on the Internet are already a consolidated reality. Thanks to practicality, to mobile access always at hand, and to the fact that we can play anywhere, mobile games are dominating this sector. However, a survey has shown that players have difficulty finding other geographically close players who enjoy this type of sport to connect with. Thus, this work aims to create a social network focused on the gamer audience. For its development, programming languages such as HTML, CSS, Java Script, PHP, and database technologies were used. For the friendship recommendation, the k-means clustering method was used in conjunction with the Elbow Method to find out what would be the ideal value of K for the application. Experiments carried out with synthetic data showed the viability of the proposal.