Mobile Phone Quality Detection Method Based On Knowledge Graph Research And Implementation

With the development of science and technology,mobile phones have gradually become popular in our lives,and the quality of mobile phone directly affects people's lives.Mobile phone quality testing is an important part of ensuring product quality.At present,the quality of mobile phones is determined by establishing inspection standards that conform to production,and the product quality is determined by the comparison of functional indicators.The standards of product generally exist in the form of text.The data distribution is scattered,the correlation is small,and it is not easy to directly query.When the product fails to use,the maintenance personnel mainly rely on the maintenance experience to confirm the cause of the failure,and cannot use the existing maintenance knowledge to determine the cause,resulting in the phenomenon of knowledge waste.Therefore,processing the knowledge related to mobile phone quality is conducive to improving the quality inspection efficiency.This thesis introduces the knowledge graph into the process of mobile phone quality detection.By establishing the knowledge graph in the field of mobile phone testing,it provides semantic support for product quality testing,and use the form of graph to intuitively display the connection between various data to optimize the validity and accuracy of information query in the detection process.Finally,the fault cause information of the mobile phone is quickly located by the method of fault cause classification.The specific research contents are as follows:1.Aiming at the problems of large and scattered mobile phone standard data and the lack of utilization of fault detection knowledge,it is proposed to construct a knowledge graph in the field of mobile phone detection for unified data resource management.First,determine the content and measurement indicators of mobile phone quality testing through testing standards,then analyze the knowledge characteristics of mobile phone repair cases,divide the entities of mobile phone quality testing and their relationships,determine the model of mobile phone knowledge base,and finally construct the data graph.Since mobile phone knowledge is mainly stored in unstructured form,this article mainly conducts data processing from three parts: knowledge extraction,knowledge processing and knowledge storage,in which data extraction mainly uses neural networks to realize entity recognition,and in the process of relationship extraction,custom rules are added based on Language Technology Platform(LTP)to realize the triple extraction with predicates as the core,and finally stored it in Neo4 j graph database storage,so as to realize the construction of knowledge graph in the field of mobile phone detection.2.In view of the problem of low fault detection efficiency,the knowledge graph can be used to intuitively determine the detection target index and improve the query detection efficiency.For the analysis of mobile phone fault causes,this paper presents a method of integrating random forest and knowledge graph to classify the reasons for mobile phone failure.7 kinds of mobile phone circuit fault state are summarized mainly from the aspects of mobile phone fault detection knowledge,mobile phone repair cases,and causes of faults,etc.Feature selection and numericalization of the fault repair cases are carried out.The the cause of mobile phone circuit failure is identified by using the classification algorithm of integrating random forest and knowledge graph and the accuracy of the failure cause recognition is up to 12% higher than that of random forest,and the average stability is increased by 4%.3.Design and implementation of mobile phone quality inspection system based on knowledge graph.Mobile phone knowledge graph,mobile phone quality detection,circuit failure case summary and other functions are realized and the construction of mobile phone testing system is completed by needs analysis and architecture design,using Bootstrap and j Query for front-end page design and Spring Boot framework to achieve front-end and back-end data interaction.Then the effectiveness of this method is demonstrated by using cases.