Font Size: a A A

Research On The Maintenance Characteristics Of Mobile Games

Posted on:2023-09-02Degree:MasterType:Thesis
Country:ChinaCandidate:J HeFull Text:PDF
GTID:2557306839461484Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
With the arrival of the information age,the mobile Application market has become one of the most active and promising markets in the world,with a large number of Applications leading to fierce competition.In order to be competitive and provide a better user experience,Application providers constantly update and maintain their Applications.In practice,however,such an important process often lacks the support of accurate information,useful guidance or specific concepts.Therefore,it is of great significance to have a deeper understanding of the maintenance process of mobile Applications in order to further develop maintenance guidelines.This article focuses on the question of how the maintenance characteristics of mobile Applications(games)have evolved over time,and the answers to this question will also provide some reference for mobile game maintenance.Before answering the research questions,we first introduce concepts and processes related to software maintenance and mobile Application maintenance as background knowledge,and then use data mining methods to solve the research questions in this paper.This paper crawls the historical version information of i OS game software.Firstly,it determines the classification type based on the domestic and foreign software maintenance types.Generally,the version description is closely related to the characteristics of different maintenance types.For example,"Bug fix" corresponds to corrective maintenance,"We added some new functionality." It is a perfect maintenance.Then,given the amount of work involved in manually classifying all version descriptions,it is more efficient to classify them by computer.The text is converted to word bag form,words are stopped,t F-IDF is used for feature extraction,and finally machine learning model prediction is used instead of manual classification.Finally,the trend curve of each Application is described with the release time as X-axis and the total number of maintenance type features as Y-axis.The Application is divided into different stages according to the development time,and the DTW method is used for hierarchical clustering of the curves of different stages.Comparing with different dimension and scale curve(i.e.maintenance type trend along with the change of time and maintain the quantitative character),and explores how mobile Applications(game)maintenance features evolved over time,eventually will maintain the characteristic model is summarized as four stages,each stage corresponding to different types of maintenance and maintenance quantity characteristics curve,This paper abstracts a common practice of mobile game maintenance,and plays a guiding role in the maintenance work of software suppliers in the future.
Keywords/Search Tags:game application, Machine learning, Text classification, Hierarchical clustering
PDF Full Text Request
Related items