一种基于支持向量机的论坛帖子热度回归预测算法

A BBS Post Heat Regression Prediction Algorithm Based on Support Vector Machine

对论坛帖子的热度进行预测成为了当前舆情监控的一项重点研究方向,本文提出一种新的论坛帖子热度回归预测算法。在进行数据处理时,首先将帖子的影响要素进行了量化处理,研究舆情热度发展的预测。本文在进行研究时,首先做的是对论坛帖子的热度的分布特征进行研究;然后量化处理影响帖子热度的影响因子,利用以回归为基础的向量机来生成舆论的发展走向;最后,对本文的预测方法进行测试,测试结果表明本文提出的帖子热度预测算法具有良好的准确性。

In the process of data processing,this paper first quantifies the influence factors of the posts and studies the prediction of the development of public opinion heat.In this paper,the distribution characteristics of BBS posts are studied firstly.Then it quantifies the influence factors that affect the popularity of the post,and USES the vector machine based on regression to generate the development trend of public opinion.Finally,the prediction method of this paper is tested,and the test results show that the proposed method has good accuracy.