Research On The Influence Of Short-video App Recommender System On Users' Flow Experience

With the development of Internet communication technology,all kinds of information has shown an exponential growth,and people have gradually moved from the blind area of information scarcity to the contradiction of information overload.Due to the limited time cost of human beings,the processing efficiency is often greatly reduced when faced with mass information,and the emergence of algorithm recommender system effectively alleviates this problem.Recommender system is a branch field of artificial intelligence data mining,which can recommend information contents of interest to users according to their interests and preferences.Short-video App,a platform that uses recommender system to deliver content,is often described by users as "The more you use it,the more addictive it becomes,which you can't stop." The problem worth exploring is how to scientifically evaluate the intelligence level of the algorithm recommender system from the perspective of user perception,It is not clear how recommender system facilitates a user's flow experience.This study based on the theory of flow,Tic-Tok App as the research object,through interview 30 samples and 405 questionnaire samples of empirical research,by the CIT method,Delphi method,statistical methods such as structural equation model analysis.In this study,it developed an intelligent evaluation model based on the user's perspective,analyzed its connotation and characteristics,and explored its influence mechanism on flow experience.In order to improve the standardization and rigor of the research,this research will use a combination of qualitative and quantitative research methods.The research process is divided into three steps to advance.Study one is exploratory research,based on the actual situation to explore possible influencing factors of the short video App flow experience.It is pointed out that the recommender system is one of the important factors.Study two is to develop an intelligent evaluation model of algorithm recommender system through the standard research steps.Based on the results of two sub-studies,Study three constructs and tests the theoretical model and clarifies the mediation mechanism for the main research problems in this paper.The results show that the AVNS model is composed of four dimensions: accuracy,variety,novelty and serendipity.AVNS model has an ideal reliability and validity.Parameter estimation method found that gender has a certain significant difference in algorithm recommendation intelligent evaluation,while there is no significant difference in AVNS model perception of age,education and monthly consumption.The factors of the four dimensions of the algorithm recommendation intelligence all have a significant positive impact on the perceived quality of UGC,while the diversity and surprise have a significant positive impact on the flow experience of users,and the perceived quality of UGC plays a mediating effect in the relationship between these two dimensions.There are two theoretical contributions to this study.Firstly,through a set of standardized scale development program,it provides a new perspective and tool for short video App and other content distribution platforms to conduct recommender system from the perspective of users.Secondly,it enriches relevant research results of flow theory and explores the path mechanism of algorithm recommendation intelligent to flow experience.In addition,this study has practical value and provides some management enlightenment and suggestions for the recommender system technology application and user retention strategy of short video App.