Research And Implementation Of Mobile Intelligent Terminal Application Information Analysis Technology

Nowdays,mobile phone and other mobile terminal devices have been widely popular owning to the fast development of mobile Internet.They have become a main tool for people to release information,entertainment,social networking,shopping,digital office and education in life.At the same time,with the continuous innovation of mobile APP technology,data carrying capacity of mobile APP continues to improve,APP contains more and more information.The traditional method of artificial APP data acquisition cannot meet the demand of data analysis grow,although the industry has adopted APP crawler method for mobile APP data acquisition,but because the data protection technology is increasingly strengthened,the applicable scope of these methods gradually reduced.This thesis presents a new method for APP data acquisition,through the implementation of APP automatic operation,screen snapshot ordered taking and snapshot content recognition with the help of OCR technology,so as to achieve the purpose of APP data acquisition.This method can provide a useful supplement to traditional web crawler methods.In this thesis,based on the automatic operation technology of Android APP and OCR character recognition technology,the automatic extraction of APP text content is realized.The main work of the thesis includes three Parts.First,the design and implementation of automatic operation system of Android APP,by APP preprocessing,APP interface response judgment,APP behavior simulation,interface redundancy judgment,screen snapshot,boundary judgment,title information acquisition and otherways to achieve the automatic acquisition of APP screen snapshots.According to the difference of APP interface data display,the system use widget analysis technology and image text detection and localization technology to realize title information collection.Besides,the method of image contrast and network analysis is used to realize the judgment of interface and network response.Second,based on the image preprocessing method,text area detection technology,character recognition technology,using the Tesseract-OCR engine to achieve the screen snapshot text information automatic recognition subsystem.In addition,through the design of the screen snapshot capture strategy and the snapshot content reorganization strategy,the snapshot recognition results are recombined,and the original APP information content is spliced and restored.Third,based on the multi host parallel processing technology,with the help of the Redis message queue,the thesis design and realize the system with parallel analysis ability that is called APP Information Analysis System,referred to as AIAS.The experimental results show that AIAS has a good effect,and it can realize the automatic and efficient data acquisition of various types of APP.The results also reflect the relationship between the contents of the original APP.The research of this thesis provides powerful support for APP data collection.