Research On The Risk Of Algorithmic News Inaccuracy And Its Governance

The combination of artificial intelligence and news has prompted structural reorganization and content innovation in the media industry,and algorithm news is one of its products.Under the coupling effect of data and technology,algorithmic news fills up the shortcomings of traditional news in all-round collection,efficient writing,and accurate push,and liberates reporters from complicated primary labor to engage in more creative work,It also greatly reduces the production cost of news,and algorithm news undoubtedly brings huge opportunities to the development of journalism.However,the inaccuracy of news has become an unavoidable ethical problem under the deep integration of artificial intelligence.Truth is the life of news.If the truth is lost,news will be impossible to talk about.The artificial intelligence technology with the aura of "objectivity" and "science" has made the algorithm dressed in the "scientific myth",but the research has found that there may be the risk of inaccurate news in every step of algorithmic news production,but the inaccurate news under the banner of "technology neutrality" becomes more concealed.Firstly,the limitations of sensors and other monitoring equipment in the link of clue discovery and the misjudgment of algorithms on "news" in the process of monitoring lead to the risk of false truth;Second,in the process of material collection,it is difficult for a wide range of data sources to verify one by one,and deep falsification technology produces deep false data that is difficult to distinguish by human beings,which leads to the risk of falsification.Thirdly,in the process of editing and production,the logic of algorithm program is misplaced and unsupervised machine learning is unexplained,which leads to the absence of checks and the immature natural language processing technology,which leads to the risk of inaccuracy.The causes can be analyzed from the objective and subjective aspects.The first is the objective level,that is,from the perspective of technology to explore the limitations of the current algorithm program on the impact of news reality,including the existence of the algorithm can not be directly empirical and the rigidity of the algorithm reporting mode.The second is the subjective level,that is,from the perspective of the subjectivity of production,to explore the inauthenticity of news caused by the ambiguity of the current subject boundary,including the possibility of artificial manipulation hidden behind the seemingly neutral algorithm and the loopholes caused by the weakening and ambiguity of the subject of responsibility.To explore its optimization paths,the authenticity of algorithm news can be constructed from two levels of data and algorithm.At the data level,the first is the data can be detected by establishing multi-dimensional vectors,and the authenticity can be quantified.Although there are still many shortcomings,it provides an idea of the possibility of technical fraud;secondly,it is necessary to improve the infrastructure of data resources Build and establish a database supervision system;finally,blockchain technology can be used to ensure the authenticity of data collection and the accuracy of reports,and to resist the intervention of external forces.The second is the algorithm level.The first is to enhance the transparency of the algorithm.Opening the black box and placing the algorithm logic under the sun is the basic prerequisite for improving the authenticity of algorithm news;secondly,it is necessary to explore the "artificial intelligence + news" interdisciplinary talent training model to cultivate oriented training In-depth learning trainers;finally,reverse engineering can be supplemented with accountability mechanisms to strengthen the internal and external supervision of the algorithm and enhance the authenticity of algorithm news.