基于复杂网络理论的舆情检测算法研究

Public Opinion Detection Algorithm based on the Complex Network Theory

目前,已存在的舆情检测算法主要基于网络文本的数据信息,而未考虑网络时间因素对信息元素权重的影响。针对舆情检测算法中网页爬取时间同步问题,提出了基于复杂网络理论的舆情检测算法,构建了复杂网络模型,提出了改进TPSN算法(即TPSN-LS算法),并应用NS2进行了仿真分析。由仿真结果可知,TPSN-LS算法在网络爬取负载、同步精度和同步次数等方面的性能都明显优于TPSN算法,使得舆情检测的结果更加准确。

At present, the public opinion detection algorithms is almost based on data of the web text, and never consid- ered the time factor of network, which impacts the information element weights. Public opinion detection algorithm based on the complex network is proposed, based on the time synchronization problem of web crawling in public opinion detection al gorithm, construct a complex network model, and improve the TPNS algorithm, which is TPNS-LS algorithm. Finally, an- alyze the simulation on NS2. The simulation results show that, the improved TPSN-LS algorithm is much better than TPSN algorithm, on the performance of crawling load on the network, synchronization accuracy and synchronization times and oth- er aspects. The research makes public opinion detection more accurate.