Research On Information Classification And Algorithmic Annotation Under The Background Of Content Industry Outbreak

At present,mobile Internet has become the mainstream platform of communication.Because the threshold of information communication has been pushed flat,and the restrictions of domestic content production have been broken,leading to the outbreak of content industry.All kinds of producers can create content according to the guidance of the platform,and because of the massive content,the mainstream trend needs to use algorithms for content selection and distribution.If the algorithm becomes the "editing center" in the era of big data,then the product manager of the algorithm is the commander,and the tagging service is the content sorting center,responsible for selecting relevant content and sending it to users according to the needs of "editing".At present,big data information platform has become the main position of information dissemination and the public opinion field.In this context,in addition to focusing on the scale and stickiness of its users,the information platform must comprehensively consider various social public needs,such as social publicity,user diversification,information inclusion,minor protection,etc.,which must be based on an in-depth understanding of content classification and marked information as the basis for distribution.This paper combines the traditional theory of communication and the demand of machine algorithm distribution.Based on the past practice of content grading,it tries to sort out the framework of grading distribution principle and content annotation.The purpose is to let algorithm product managers and engineers understand the overall hierarchical logic of information content first? Why can’t some content be distributed,and some content can’t be distributed without traffic.This kind of understanding is not simply to implement the requirements of the competent departments,but from the perspective of social structure and the social value attribute of the media.From the perspective of the overall framework,we divide the content into three levels.The first level is to understand the values of the media platform from the requirements of the overall social value attributes,and to understand what is the social responsibility content of the media,whether there is traffic or not,which needs to be distributed;the second level is to resolutely not distribute,which will violate the laws and regulations;the third level is not to violate the laws and regulations,but needs to be distributed Based on the user’s interest,distinguish and identify the information to be distributed to users.Based on this content classification framework,this paper gives the suggestion of annotation framework,and thinks that non exclusion should be a basic principle of content annotation in the era of big data,which is an innovative idea different from the previous classification methods.The connotation of cultural creativity is very extensive,and information dissemination must be an important means for the development of cultural industry.In the era of big data artificial intelligence,culture is more needed to guide technology.Information value judgment is one of the core functions of culture.We can take this advantage to standardize artificial intelligence and algorithm.This is also one of the starting points of this paper.