基于BP神经网络的网络小说排行预测

近年来随着"IP"热潮兴起,网络文学市场发展迅速,逐渐成为文化娱乐行业投资热点.本文将机器学习方法引入到小说排行预测方面,通过网络爬虫获取网络小说信息并提取了影响排行的特征,提出了基于BP神经网络模型进行小说排行预测.针对训练数据的不均衡,本文采用ROC和AUC作为预测评价指标;实验结果表明,基于BP神经网络的网络小说排行预测的准确率较高,相比传统的文学定性分析方法,机器学习预测方法可解释性和应用性更高.

Prediction of online novel rankings based on BP neural network

In recent years, with the rise of the "IP" boom, the market of online literature is developing rapidly, has gradually become a popular type of entertainment investment industry. This paper introduces the machine learning method to the prediction of the novel rankings, obtains the network novel information through the web crawler, extracts the characteristics of the influence rankings, constructs the BP neural network model to predict the range of the novel rankings. In view of the non equilibrium of the prediction results, the ROC curves and AUC value used as the classification performance metrics, the accuracy is more accurate. Compared with the traditional literature qualitative analysis method, machine learning method is more predictable, interpretable and applicable.