Modelling on movie box-office prediction based on LFM algorithm

Concerning the limitations that the accuracy of predication is lower in the traditional model of movie box-office prediction and classification, this paper proposes a novel model of movie box-office revenue prediction. The model is based on the relationship of movie box-office and the user behaviors. The algorithm could be summarized as follows. Firstly, this paper uses the Latent Factor Model (LFM) to classify the movies. Secondly, for different categories, this paper constructs a series of linear movie box-office prediction models. And these models use total box-office as the dependent variable, while the independent variables are user reviews and user ratings which are the major manifestation in vertical media. Finally, this paper uses the experimental results to adjust the independent variables in different models. The experimental results demonstrate that the model performs better on prediction than the traditional methods.