Design And Implementation Of Statistical Analysis System For User Interactive Behavior On Mobile Retrieval

In recent years,the mobile Internet has become more and more popular,and the demand for retrieval by mobile devices has also increased.Calculating and analyzing the performance of search is important for evaluating retrieval and an important basis for further optimizing user search experience.How to calculate and analyze search performance and perceive user search satisfaction is very active in the area of PC research,but in the mobile area,there is rarely a specific discussion on mobile devices.Obviously,PCs and mobile devices(such as smart phones)are very differences.In addition,users' search behaviors,especially interactive behaviors,in different areas of mobile search are highly related to the layout of the mobile application UI.It is critical to accurately collect user interactive behaviors and accurately calculate user search performance and reflect the user search experience.Therefore,this article aims to propose a set of user interactive behavior signal schemes on mobile devices in search.Design and implement the analysis system to feedback the search performance and reflect the user experience,so we can improve search performance by the data for next step.For example,according to the user's click behavior on the search results page,it can be detected which search results are more relevant to the user,thereby modifying the ranking of the search results page,and increasing the priority of more relevant result items.This article has done the following work:1)For retrieval,a set of human-computer interaction signal based on mobile devices is introduced.The client collects interaction signals in real time to collect user interaction behaviors.2)Design and implement an interactive signal analysis system,summarize and analyze the interaction log,and finally calculate the retrieval performance and behavior habits of the user in the retrieval process,and give a visual graphical report for display.3)Designed and implemented the Microsoft Teams mobile application,and used millions of users to verify the above solutions and systems.The experimental results show that the search-oriented interactive signal statistical analysis system can effectively reflect the user search behavior habits and search performance,and then provide a reliable basis for subsequent retrieval optimization.