Design and Analysis of Multi-Source Network Public Opinion Clustering Strategy in Cloud Environment

摘要

The Aiming at public opinion mass multi-source network public opinion information extraction accuracy needs to be improved, designs a clustering strategy which can cluster multi-source network public opinion accurately based on cloud environment. This strategy firstly establishes multi-source public opinion information architecture and labels the public opinion information according to its source in order to identify and distinguish the source of information. Secondly constructs the multi-source network public opinion clustering structure, and on the basis of thinking all kinds of sources of network public opinion to choose the appropriate clustering method for each one. Finally, according to the construction of the multi-source network public opinion clustering architecture, with the use of cloud computing in parallel computing environment, we can cluster the multi-source network public opinion fast and accurately. The experimental results show that the clustering strategy can cluster multi-source network public opinion with high accuracy, and accelerate the speed of access to public opinion, can provide a method for real-time monitoring and analysis capabilities of network public opinion promotion.