Drama plot development law and its prediction model based on time series analysis

Abstract Most dramatic works narrate the story in chronological order, from the beginning to the end, according to the sequence of events, so the time series analysis can be used to explore the law of dramatic plot development. In this paper, to improve the problems in time series forecasting, a combination model ARIMA-LSTM of ARIMA model and DA-LSTM model is constructed to realize the prediction of drama plot development, and the prediction performance of the model is evaluated experimentally. The adopted time series data passed the smoothness test and the residual white noise test, and finally the ARIMA model was ranked as ARIMA(3,1,4) model, which predicts the overall trend of the value and the real value basically consistent, but there still exists a certain amount of prediction error. At the same time, the LSTM model also has a certain error. After combining the two models, the predicted values are basically consistent with the real values. The prediction performance of the combined ARIMA-LSTM model is significantly better than that of the RNN, classical LSTM, DA-LSTM and ARIMA models, and the MAE, RMSE and MAPE are all smaller than the other obvious ones, which are 0.000601, 0.000852 and 0.000504, respectively, which verifies that The superiority and practicality of the model in this paper in the prediction of drama plot development pattern.