Pros And Cons Analysis And Future Expectations Of Today's Headline Algorithm Recommendation

The news and communication industry has a high dependence on science and technology,and every development of science and technology will bring new changes to the news and communication industry.Nowadays,algorithmic recommendation technology has become the main content distribution method in the field of network communication,bringing news dissemination into the "fast lane".Today's Headline has completed the efficient supply of information by means of algorithmic recommendation technology,and effectively guaranteed the rights and interests of people to obtain information quickly.However,Today's Headline still has deficiencies in content quality,network public opinion and other aspects.It is the focus of algorithm recommendation improvement to recognize and solve these problems.This paper takes Today's Headline as the research object and analyzes the advantages and disadvantages of its algorithm recommendation by using comparative research method,questionnaire survey method and interview method.The paper points out that the advantage of algorithm recommendation lies in the selection,filtering and distribution of information,which effectively alleviates the problem of information overload,and the dynamic user portrait also promotes the accurate matching of algorithm.At the same time,the aggregation information platform based on algorithm recommendation is quietly becoming the main force of the domestic Internet ecology.For the platform,the algorithm model of Today's Headline has a low cost and high reproducibility,which can be constantly upgraded and applied to other models.Compared with rival apps,the news stream ads of headlines will be timely adjusted in terms of distribution frequency and multi-touch orientation to bring better user experience.Although Today's Headline's algorithmic recommendation technology has been greatly improved since its inception,it still has some problems.First,the algorithm tends to recommend low-quality content that is easier to attract people's attention,and the importance of information does not match its distribution resources.In addition,in the reporting of covid-19 outbreak public health emergencies,the algorithm recommends the transfer of rights to human editing.Secondly,the algorithm can not effectively govern the fake news with high immediacy.Algorithmic weighting creates an information gap between the subjects of power.The homogenized content promotes "individual polarization" and "group polarization" and hinders the rationalformation of public opinion.Finally,the subversion of the communication mode by algorithm weakens the communication effect of mainstream values.Based on the questionnaire survey data as well as to other platforms practice experience of learning,the article puts forward some Suggestions and prospects of optimization: technically,can be set up by algorithm more reference indexes,injection delay meet the recommendation system logic methods such as optimizing algorithm,by increasing the semantic tolerance to expanding the scope of the content of the push,you can also set a reading remind,create a bot to rectify technology.In the intelligent world where everything is media,more enterprises should be encouraged to disclose their algorithms and promote the progress of the industry with the power of interpretation and accountability of algorithms.At the same time,build a cross-domain algorithm governance platform,and finally make efforts to realize the resource-flexible intelligent networking society.