Characterizing And Predicting The Cross-APPs Behavior Of Mobile Search

In the current mobile Internet environment,mobile search has become the main search method for mobile users in China.With the rapid development of mobile applications(APPs),a variety of mobile applications have become the main channel for Smartphone user interaction.In recent years,the user’s mobile search behavior has gradually attracted the attention of the academic community.However,there are few researches on the association between mobile application usage and mobile search behavior.Especially in the increasingly common environment of users’ cross application(cross-APPs)interaction behavior,the research on the user’s cross-APPs behavior in the process of mobile search has academic value and practical significance.In view of this,this paper focuse s on the user’s cross-APPs behavior of mobile search in daily life scenarios,explores and identifies the characteristics of cross-APPs behavior of mobile search,and studies how to effectively predict the APPs that users may use in mobile search.Firstly,through the research of the basic theories,this paper comb ed the relevant literature and theoretical basis,and clarifie d the relevant concepts involved in the research of this paper.Then,based on the relevant theoretical framework of information seeking behavior and information context,this paper proposed a theoretical model of cross-APPs behavior of mobile search.Based on the theoretical model,this paper determine d the experimental process and the diverse data collection methods.In order to study the user’s cross-APPs behavior of mobile search in the real scene,and analyze the in-depth reasons of user’s behavior,this paper collected 30 participants and carried out a fifteen-day user experiment.The experiment includes pre experiment,formal exper iment and experimental interview.It used a variety of data collection methods such as mobile log data,structured log,questionnaire survey and user interview.Through the theoretical model and research design,this paper focuse d on:(i)Mobile search cross-APPs behavior’s path and pattern.This paper first studied the basic characteristics of cross-APPs behavior of mobile search from the dimensions of APP types,time and topic distribution of mobile search queries,and basic characteristics of mobile search sessions.Then,this paper analyzed the cross-APPs path from the perspective of APP transition probability between different APPs;and then summarizes the results from the perspective of behavior types in mobile search.Furthermore,this paper studied t he relationship between different cross-APPs behavior patterns and cross-APPs path.(ii)From the perspective of multi-dimensional context,this paper studied the deep intention of cross-APPs behavior of mobile search.This paper mainly analyzed the user’s intention of cross-APPs behavior from the perspective of search topic,search task,search motivation,environmental factors and follow-up behavior,and studied the relationship between different context information and follow-up behavior of mobile search.(iii)Based on the previous research and analysis,this study defined the process of cross-APPs behavior prediction of mobile search,constructed the model of cross-APPs behavior prediction of mobile search by extracting multi-dimensional behavior characteristic variables.Besides,this paper also studied the and studies the prediction effect of different dimension data sets in cross-APPs behavior prediction of mobile search.(iv)This paper revised the theoretical model of cross-APPs behavior of mobile search and discussed its practical value.Based on the research conclusions,this paper modified the theoretical model proposed in this paper.In addition,this paper discussed the application and optimization of mobile search cross-APPs prediction algorithm for the existing mobile APPs recommendation design.At last,this paper put forward the design idea of mobile search cross-APP behavior prediction system.This paper not only put forward the theoretical model,but also conducted empirical research.This research can enrich the existing information behavior and information retrieval theory,and help to improve the user mobile search experience and improve the quality of mobile search service.