Research On Specific Family Query Method Based On Kinship Network

As a large country in population,the study of population has attracted more and more attention from academia and government.Some people form a family through marriage and parent-child relationships.The family is the basic constituent unit of society.Depending on the relationship within the family and the role of the family member,the family can be divided into different types.The study on the family types is an important part of demographic research.With the advent of the era of population big data,lots of big data analysis technologies provide good technical support for further research and analysis of families.Based on the data of the full population of the H province,the author improves the quality of the data,then extracts the appropriate attributes and reorganize the data.The reorganized data is imported into a graph database to build a kinship network with household information.The kinship network vividly reflects the relationships between spouses,fertility,person and household.Based on the kinship network and the definition of family type in demography,this thesis puts forward a kind of topology graph model for the family type structure,which extends the previous research results of our group.The topology graph model of family type structure can represent more extensive family types.In the thesis,the problem of family type statistics is transformed into the problem of graph pattern matching.Then the data graph and pattern graph are redefined.A novel algorithm for family structure matching is put forward based on the definition of family structure matching to carry out family type query and statistics.The studies above mentioned provide a new idea for the family classification and family type query statistics.The main research work of this thesis includes the following aspects:(1)Proposing the topology graph model of family type structure.This thesis analyzes and studies the family type representation proposed by our group,then puts forward the topology graph model of family type structure.On the basis of expressing conjugal nuclear family and standard nuclear family,the new topology graph model of family type structure can also represent extended nuclear family.The topology graph model of family type structure can cover the whole kinds of nuclear family.On the basis of three basic types of nuclear family,the new model includes some other special family forms at the same time.Based on the new model,this thesis analyzes the specific nuclear family types and summarizes the nature of the family type structure.(2)Proposing a family type structure matching algorithm based on the topology graph model of family type structure.This thesis analyzes the essence of family type statistics,then redefines the data graph and pattern graph.A new matching algorithm is put forward based on the graph pattern matching methods.This thesis uses the provincial and municipal kinship network as data graph,the topology graph model of family type structure as the pattern graph.The thesis proposes a family type structure matching algorithm to solve the problem of specific family type query and statistics by the query of graph database.(3)Construction of a family type query system.Based on the topology graph model of family type structure and the family type structure matching algorithm,a family type query system is implemented.The system is based on the kinship network stored in graph database and the Flask framework.According to different administrative divisions and different family types,the query system provides various interfaces for family type query and statistics.In order to solve the problems of difficulty of data sharing in the family research fields,this thesis defines the data structures of shared data sets and implements the construction method to extract shared data sets.Shared data sets provide the data base for expanding population research.At the same time,the visualization of query results is realized by using visual plug-in.