Research And Application Of Visualization Of People Relationship Based On Spark Platform And Call Records Analysis

With the rapid advancement of the information age,mobile phones,as a necessary product,have become more and more diversified in functions and effects.Today's mobile phones have various functions such as calling,sending messages,surfing the Internet,navigation,and office,as well as recording call information,interpersonal relationships,lifestyle habits,etc.At the same time,mobile phones have also been widely used by criminals in the process of crimes.Most criminals have used mobile phones to varying degrees in their premeditated and criminal actions.Therefore,the analysis of the call records in the dialogue has become an indispensable and important means in the daily handling of cases.Analyzing the call records information to obtain the call characteristics of the owner and the character relationship between the relevant person and the owner has become a very important means for the police in the investigation of the case and the investigation of the suspect.This paper studies the Spark platform as the basis of the analysis platform.The data in call records is used as the data basis.Through the preprocessing,storage,query,and analysis of the dialogue data,relevant call feature information is generated,and a decision tree classification algorithm is used to obtain characters.Relationships and visualizations have very effective guidance on the direction of investigation.The main work completed in this dissertation includes the following aspects:1.Based on the Spark platform research,a classification analysis dialogue data based on small-world theory and decision tree is used to conduct data mining and generate character relationships.Through the Spark SQL query technology,statistics on the call information of the owner's records data,master the basic information of the owner,and obtain the call information,call time,call duration,location and other important call feature information.Using the Spark MLlib of Spark's big data platform,combined with the acquired call feature information,the decision tree is modeled,and a classification model of the decision tree is constructed.The classification of the character relationships in the dialogue can be used to obtain the required character relationships and assist the user.Quickly understand the relationship between the owner's social circle,excavator owner and related numbers.Using the Spark MLlib of Spark's big data platform,combined with the acquired call feature information,the decision tree is modeled,and a classification model of the decision tree is constructed.The classification of the character relationships in the dialogue can be used to obtain the required character relationships and assist the user.Quickly understand the relationship between the owner's social circle,excavator owner and related numbers.2.Design a method for obtaining mutual friends and visualization based on Spark Graphx graph calculation.By using the calculations of the Spark Graphx graph calculation framework and the connected graph subgraph algorithm,all the owner's buddy lists and multidimensional buddy lists are obtained.Based on the six-degree separation theory,the algorithm of single source shortest path in Spark Graphx is improved,and the connected path between two points is obtained to obtain the mutual friend algorithm.In order to determine whether there is a mutual contact between different owners or if there are the same friends.The visual display process of the relevant call features and personal relationship diagrams,through the Echarts visualization library,visualized results such as line charts,bar charts,pie charts,and character diagrams as needed.3.A character relationship analysis and visualization system based on the Spark platform and records was designed and implemented,and the user's call feature statistics and related person relationship analysis were performed.The dialog box records the feature information obtained by the query,and visually displays the statistical results of the feature information such as a line graph,a bar graph,and a pie chart.Through the characteristic information of these calls,a decision tree analysis is constructed to obtain the relationship between the characters,and the networked relationships are visually displayed using data visualization means;for the close contacts,the relationship between the owner and the other party is established for the person relationship.People relationship network.Get friends and mutual friends through the Spark platform and analyze suspicious contacts.