Finding similar users in social networks by using the depth-k skyline query

Search algorithms designed to seek out similar users in social networking sites are a significant function of recommendation systems. Conventionally, such sub-algorithms consider all the dimensions of user data as a whole. However, as the information in various dimensions is generally independent, the conventional approaches may not be the best way to find similar users. This paper solves this problem by proposing an approach based on depth-k skyline queries that searches for similar users with multiple conditions. This paper also presented an algorithm to accelerate this process, the effectiveness of which was demonstrated in a simulation.