中国农业科学 ›› 2022, Vol. 55 ›› Issue (21): 4158-4174.doi: 10.3864/j.issn.0578-1752.2022.21.006

• 耕作栽培·生理生化·农业信息技术 • 上一篇    下一篇

基于裂变模式的农业农村信息快速调查及分析平台开发与应用

张杰1(),王川1,董晓霞1,祝文琪1,岳慧丽1,2,刘升平1,2(),周清波1,2()   

  1. 1中国农业科学院农业信息研究所,北京 100081
    2农业农村部区块链农业应用重点实验室,北京 100081
  • 收稿日期:2022-01-03 接受日期:2022-04-01 出版日期:2022-11-01 发布日期:2022-11-09
  • 通讯作者: 刘升平,周清波
  • 作者简介:张杰,E-mail:zhangjie10@caas.cn
  • 基金资助:
    国家现代农业产业技术体系(CARS-44-KXJ24);中央级院所基本科研业务费(JBYW-AII-2021-11);中国农业科学院创新工程项目(CAAS-ASTIP-2016-AII)

Development and Application of Rapid Investigation and Analysis Platform for Agricultural and Rural Information Based on Fission Model

ZHANG Jie1(),WANG Chuan1,DONG XiaoXia1,ZHU WenQi1,YUE HuiLi1,2,LIU ShengPing1,2(),ZHOU QingBo1,2()   

  1. 1Agricultural Information Institute, Chinese Academy of Agricultural Sciences, Beijing 100081
    2Key Laboratory of Agricultural Blockchain Application, Ministry of Agriculture and Rural Affairs, Beijing 100081
  • Received:2022-01-03 Accepted:2022-04-01 Online:2022-11-01 Published:2022-11-09
  • Contact: ShengPing LIU,QingBo ZHOU

摘要:

【目的】为了有效解决传统农业农村基础信息采集过程中存在的采集周期长、样本量小、投放不精准、数据质量低、审核难度大等难题,实现突发事件背景下海量数据快速获取及可视化分析,本研究设计、研发了一套集问卷精准投放、数据快速采集、结果自动分析为一体的农业农村信息快速调查平台,为各级农业管理部门快速决策提供便捷工具。【方法】(1)总体设计:借助5层架构技术和中台理念设计平台,采用新媒体、GIS、社会网络分析(SNA)等技术,构建问卷精准投放模型、多级裂变模式、快速审核和自动分析策略,利用点度中心度、中介中心度、空间聚类等方法对调查问卷传播网络的裂变关系、扩散能力以及社群特征进行分析,深度挖掘、多维度呈现平台数据价值。(2)系统实现:采用前后端分离模式研制了2个子系统,分别为面向调研用户的速调采集系统、服务于信息调查者的速调管理及分析系统。(3)平台应用:利用VMware vSphere集成容器和集群服务器部署系统,借助Nginx负载均衡方式提高系统响应速度。【结果】通过半年持续运行,系统普适性强,已提供8大领域12次大规模农业农村信息调查服务,共采集问卷76 953份;农业农村经济信息调查案例中5 h内采集了12 243份问卷,裂变比例为88.23%,其中20.24%的人形成扩散并裂变,社群效果明显,裂变程度高、速度快、范围广;平台综合功能强,覆盖了问卷设计、投放地控制、裂变分享、智能审核、酬劳发放、统计分析、用户沉淀等功能,可为农业农村信息快速调查提供高效的一站式服务。【结论】基于裂变模式的农业农村信息快速调查与分析平台,具有数据采集快、信息获取真、样本收集多、功能性强的特点,实现了突发事件背景下农业农村基础信息紧急获取及分析结果可视化呈现,形成了一种基于海量数据、区别于官方调查的民调模式,可为各级农业管理部门开展行业信息的精准、高效调查,实现科学决策提供有力的技术支撑,对其他行业信息调查具有借鉴意义。

关键词: 农业农村信息, 快速调查, 裂变模式, 社会网络分析, 精准投放模型, 系统设计

Abstract:

【Objective】In order to effectively solve the problems existing in the process of collecting basic information of traditional agriculture and rural areas, such as long collection period, small sample size, inaccurate delivery, low data quality and difficulty in auditing, and to achieve rapid acquisition and visual analysis of massive data under the background of emergencies, a rapid survey platform for agricultural and rural information was designed and developed, which integrated accurate questionnaire delivery, rapid data collection and automatic analysis of results, so as to provide a convenient tool for agricultural management departments at all levels to make efficient decisions. 【Method】 (1) Overall design: The platform was designed based on the 5-layer architecture technology and the concept of “Mid-Plane”, and the new media, GIS and social network analysis (SNA) were used to build a questionnaire accurate delivery model, multi-level fission model, rapid review, and automatic analysis strategies. In order to deeply mine and present the value of platform data in multiple dimensions, this research used methods such as point centrality, betweenness centrality and spatial clustering to analyze the fission relationship, diffusion capacity, and community characteristics of the questionnaire dissemination network. (2) System realization: two subsystems were developed by adopting the front-end and back-end separation mode, namely, the rapid investigation collection system for survey users, and the rapid investigation management and analysis system for the information investigators. (3) Application: VMware vSphere integrated container and cluster server were used to deploy system and Nginx load balancing to improve system response speed. 【Result】 After six months of continuous operation, the system had strong universality. It has provided 12 large-scale agricultural and rural information survey services in 8 fields and collected a total of 76 953 questionnaires; In the agricultural and rural economic information survey cases 12 243 questionnaires have been collected within 5 hours, with a fission ratio of 88.23%, of which 20.24% formed diffusion and fission. The community effect was obvious, with high degree of fission, fast speed and wide range. The platform had strong comprehensive functions, covering the functions of questionnaire design, placement control, fission sharing, intelligent judgment of papers, reward distribution, statistical analysis, user classification, etc., which could provide one-stop service for rapid investigation of agricultural and rural information.【Conclusion】The rapid investigation and analysis platform of agricultural and rural information based on fission mode had the characteristics of fast data collection, true information acquisition, more sample collection, and strong functionality. It realized the function of obtaining basic information urgently and visualizing investigation results quickly under the background of emergencies, and formed a poll mode based on huge data and different from official investigation. It could provide strong technical support for efficient decision-making of agricultural management departments, and had reference significance for the information investigation of other industries.

Key words: agricultural and rural information, quick investigation, fission mode, social network analysis, precise placement model, system design