A New Method based on Artificial Intelligence: to Improve the New Suggestion System by Analyzing User Interaction Patterns

The purpose of this research is to provide a method based on neural networks for suggesting news through user interaction patterns. To address this important challenge, a graph-based news recommender system was presented, which recommends the best news for the user according to the global representation and combining it with the user's local information. The method of this research is focused on enhancing the representation of historical news through the use of a global news graph and improving the representation of candidate news content through a global entity graph. First, the representation of news text and news entities is learned from a local perspective. Then, world-aware historical news coder and world-aware entity news coder are used. Finally, a concise user coder and a news recommendation component are used. In this research, transformer networks were used for content-based news placements, as well as neuro graphic networks that provide communication reasoning. In addition, considering world news, we tried to suggest news that were hidden from the view of previous models.