A Study Of Push Notification News Recommendation From The Perspective Of Comparison Between Human Editor And Algorithm ——Wesee As An Example

With the continuous development of new media technology and intelligent algorithms,algorithms and push notifications play an increasingly large role in new media communication.Based on the platform of Tencent Wesee,the article explores the selection preferences of algorithms and human editors in push notifications as well as the division in the whole push notifications delivery process from a comparative perspective.Based on this,the study discusses the types of news that users are actually interested in and whether algorithms and human editors influence each other in terms of news value orientation,in conjunction with the click rate of push notifications news.The study summarizes the similarities and differences in the selection preferences of both human editors and algorithms by conducting content analysis of 697 push notifications news that were actually distributed for 30 days randomly selected from July to September 2021,and comparing them from five perspectives:timeliness,originality,neutrality,serviceability,and comprehensiveness.On this basis,the value of news that is more attractive to users is explored by combining click rates.Secondly,the study interviewed 8 push notifications operators,and established an interview outline based on the role of algorithms in the entire distribution mechanism of push notifications,and combined with the author's 3-month participatory observation of the push team,the study compared the actual roles played by human editors and algorithms in the three aspects of topic acquisition,review and filtering,and news recommendation,and summarized the two processes are positioned differently in the whole process,and then we explore whether they influence each other in terms of news value.The study found the different selection preferences between algorithms and human editors on news content:First,news with strong timeliness is mainly recommended by human editors,and news with less timeliness but more interest is mined and crawled by algorithms for recommendation.Second,the main original,independent content news is handled by human editors.Third,given the stronger awareness of human editors to provide information that serves users,in addition to hazard tips,many news of opinion guidance and methodological guidance are recommended by human editors.Fourth,both generally tend to select domestic social news,but in terms of characters,human editors are mainly responsible for focusing on recommending news about the state and social management class,while the algorithm focuses mostly on the interesting lives of ordinary people.The study also analyzed the popular push notifications that reached a certain click rate separately according to the content analysis framework,and analyzed the types of news that users are actually interested in.The study found that fun,proximity,and prominence were the more attractive news values to users.In addition,the study found that the positioning of algorithms in the downlink mechanism is mainly to liberate and assist human editors to a certain extent,while human editors are still in a central and critical position in the downlink mechanism.Second,the study found that human editors can indeed propose improvements to the algorithmic team in terms of news value,specifically in terms of freshness and authenticity,to achieve the effect of improving news quality.In addition,the news value represented by the algorithm effectively influences the choice of human editors,who also tend to favor the fun stories pursued by the algorithm in the choice of news value,ignoring another element of news value-importance.The study also summarizes the positive and negative effects of the algorithm in the process of push notifications news distribution respectively,and proposes optimizations:First,updating the automatic writing algorithm to improve the readability of the text;Second,raising the quality threshold of the algorithm in crawling the selection to ensure the diversity and comprehensiveness of the news content;Third,strengthening the human-computer cooperation to enhance the serviceability and originality of the news content.Finally,the study offers a reflection that push notification is a powerful complementary illustration of the concept of ambient journalism in reality,and that users who over-rely on this passive way of receiving news will bring challenges to their perception of the real world.Because push notifications are essentially a way to"advertise" potentially popular news,it doesn't help users to fully perceive reality.Research suggests that by pursuing click rates too much,and even using algorithms to enhance such news selection,the mimetic environment presented by push notifications news will only get narrower and more extreme,and it is also a waste of push notifications as a window and opportunity for users to learn more about valuable news.Aggregate content platforms and organizations should make good use of push notifications to show users more topics in interesting ways,broaden their established range of interests,provide them with more perspectives that challenge their own perceptions,and help de-cocoon their information environment.