Analysis of media content recommendation in the new media era considering scene clustering algorithm

Abstract With the advent of new media era, we face problem information overload every day, based on which this paper proposes a scenario-conscious clustering algorithm for recommending content. The improved K-means is used to cluster content, an initial center randomly selected, and remaining centers are obtained by executing reduce number iterations avoid neighboring situations. then compared with content-based recommendation techniques, KNN algorithm, collaborative filtering in terms accuracy, recall, MAE value, execution time. significantly better than other algorithms regarding accuracy recall. 0.4 when sequence 30, higher technique, 0.1 KNN, 0.12 content method. In outperforms nearest neighbors selected be above 20. time, longer amount data, more obvious advantage algorithm. Therefore, applicability reliability model proposed verified recall meet design requirements.