At present, there are some resistible illegal operations aiming at creating false public opinions in internet public opinions on emergent event, which seriously disrupted the normal Internet order. However, the tradit...At present, there are some resistible illegal operations aiming at creating false public opinions in internet public opinions on emergent event, which seriously disrupted the normal Internet order. However, the traditional research method of internet public opinion pre-waming mainly relies on manual analysis, which is too inefficient to adapt to the analysis of massive internet public opinion information. According to the above analysis, this paper puts forward an internet public opinion pre-warning mechanism on emergent event based on multi-relational data clustering algorithm, discusses the specific pre-waming from the aspects of the state and dissemination of internet public opinions and the historical data, and automatically classifies the internet public opinions through multi-relational data clustering algorithm. And the results show that such method can be used to effectively study the internet public opinion pre-waming on emergent event, with the accuracy rate of as high as 95%.展开更多
文摘At present, there are some resistible illegal operations aiming at creating false public opinions in internet public opinions on emergent event, which seriously disrupted the normal Internet order. However, the traditional research method of internet public opinion pre-waming mainly relies on manual analysis, which is too inefficient to adapt to the analysis of massive internet public opinion information. According to the above analysis, this paper puts forward an internet public opinion pre-warning mechanism on emergent event based on multi-relational data clustering algorithm, discusses the specific pre-waming from the aspects of the state and dissemination of internet public opinions and the historical data, and automatically classifies the internet public opinions through multi-relational data clustering algorithm. And the results show that such method can be used to effectively study the internet public opinion pre-waming on emergent event, with the accuracy rate of as high as 95%.