Research On Social Hot Topic Technologies For Internet Public Opinion Analysis

With the continuous development of information technology,the Internet has gradually come into people's life,the Internet public opinion more and more accurately reflect the current social issues,people's attitude,so many government departments and enterprises to public opinion on the network information for monitoring and management becomes very urgent.At the same time,due to the large amount of information on the Internet,and it is widely distributed,these characteristics make it difficult to rely on artificial means to achieve information screening and monitoring of public opinion.Therefore,it is of important social value to find out the current public opinion hot spot information through public opinion monitoring,and to provide theoretical basis for relevant departments and media to deal with emergencies.Firstly,this thesis introduces the key technologies involved in the public opinion analysis of hot topics,Based on the research of web page content extraction and web page layout structure,In order to solve the problem of low accuracy of web page content extraction from different structure of web pages,a content extraction algorithm based on structure similarity page clustering is proposed.The contribution of each "block" to the template is assigned to different weights according to the composition of the front page of the web page.Secondly,the similarity of the corresponding blocks in the two web pages is calculated.The similarity and the weight of each block product as the sum of the two pages of the similarity.This algorithm takes into account the influence of web page structure difference on Web page content extraction.Web page clustering based on computing the similarity between web pages.The results are more accurate for the web page content in the same cluster.Secondly,by comparing the advantages and disadvantages of a variety of text clustering algorithm,combining the research object according to network news,many times with all objects clustering problems need influence the efficiency of the original Single-Pass algorithm randomly selected operation clustering results and algorithm process of cluster centers,the introduction of the initial cluster centers,and continue to join in the new text clustering center update process,to improve the efficiency of the algorithm.Then the attenuation function is introduced to calculate the heat value,according to its characteristics to calculate the heat value of the topic,then,according to the topic of heat value obtained a hot topic in public opinion information.Finally,according to the above research,combined with the design requirements and functional requirements of the platform,completed the design of the system architecture and function modules,implement the topic of network public opinion hot social analysis platform.The platform test results indicate that the platform can quickly and timely access to network information and analysis through digging out the hot topic behind them,the hot topic detection function can basically achieved the desired design goals.