eCensus System

1
e-Census System

• Sep. 2006
• Korea National Statistical Office

2
Agenda

• Why e-Census System?
• Implementation Steps
• Major Supports of this System
• e-Census System Now
• Internet Survey
• Future Works

3
Why e-Census System?

• Changes of Census Environment
• No more cooperative env to Census
• Difficulties in meeting interviewees
• Cost increasing for recruiting enumerators
• Needs of Managements
• Process Management
• Enumerators their works
• Needs of nation-wide management for Census
  Materials

4
Why e-Census System?

• Needs of timeliness
• Results of Census are used as sample frames of
  other surveys
• Statistical Information was needed earlier
• Verification of Appropriate Model through the
  pilot censuses
• OCR based System
• Web based System
• KNSO took advantages of both systems

5
Implementation Steps

• Pilot Censuses
• OCR based test (3-times)
• Web based test (3-times)
• Preparations for Introduction
• Hardware
• Servers, Storage, Backup, Network devices
• Software
• Databases, Middleware, Packages for business,
  System operation software, Development tools

6
Implementation Steps

• Schedules

Selecting census tract
Sampling census tract
Main survey /data processing
Application for Internet survey
Provisional compiling
Employment
Calculating goods and human resource
Integration analysis design
1ST Test
2nd Test
Realization
3rd Test
Detailed design
Realization
Unit Test
Integration analysis/design
Unit Test
Realization
Introduction Installment
Installment test
Introduction installment of system S/W
Introduction Installment of application S/W
Environment composition of system S/W
Pilot Operation (Integration systemTest)
Normal Operation
7
Implementation Steps

• Processes of development

1. Analysis of requirement
3. Design of UI
4. Design of DB
2.Design of architecture
Repetitive,, Gradual development (repeat of two
times)
5. Realization
6. Test
9. Operation of system
8. Training of users
7. Installment of system
8
Major Supports of this System

• Major Supports
• Management of Census Enumerators
• Census proceeding status
• Cyber education for enumerators
• Communication with e-mail cellular phones
• Internet Survey
• Online application to attendance and input

9
Major Supports of this System

• Web based input Checking
• Input and checking in the field(distributed
  process)
• Data is collected in central database
• Post Survey
• Surveys for verification of accuracy
• Tabulation Analysis
• Tabulation using ICT

10
e-Census System Now

• Diagram for entire system

11
e-Census System Now
Gigabit Backbone Network
Fiber Channel

• Diagram for hardware

Web server 1
Web server 2
Mail/Short message transmission server
WAS server 1
WAS server 2
IBM p5 590
IBM p5 590
IBM p5 570
IBM p5 595
IBM p5 595
??? 3
??? 1
??? 4
??? 2
high availability, non-interruption, non-stop
Compiling/analysis server
Backup management/SMS server
DB server 1
DB server 2
IBM p5 595
IBM p5 595
IBM p5 550
IBM p5 570
Oracle RAC Sync
Heartbeat Network
Storage for data storage
Storage for data duplication
Backup equipment
StoregeTek L80
HDS TagmaStoreTM USP
HDS TagmaStoreTM USP
SRDF Link
12
e-Census System Now

• Diagram for network

13
e-Census System Now

• Diagram for security system

14
Internet Survey

• Goals
• Preventing survey omission for the vulnerable who
  are difficult to have interview survey
• Accepting household that avoid interview survey
  for privacy
• Target
• people who want Internet survey for population
  housing census
• Periods
• Application submission Oct 29. Nov 10. (13
  days)
• Input Oct 29. Nov 12. (15 days)

15
Internet Survey

• Effects
• Possible to identify real-time proceedings
• Inducing active participation of applicant by
  automatically sending information notice
• Flow of Internet survey

16
Internet Survey

• Characteristics
• To minimize input errors by applying input
  editing by proceeding step
• To Minimize input work through automatic linkage
  with applicant address and census tract
  information
• To adapt security policies such as application of
  encryption of application input data
• To induce to participate by SMS, e-mail.
• Real-time monitoring of proceedings of Internet
  applicants

17
Internet Survey

• Results
• 44 of participation rate in contrast of the goal
  
• Status of input by survey date / by input time

18
Internet Survey

• Results
• Required time for input Status of drop-out of
  applicant

19
Internet Survey

• Improved
• Non-stop operation of the system
• Identifying real-time status
• Immediate response to correction matters
• Simultaneous noticing and giving directions
• To be improved
• Low participation rate
• Waiting time was needed after application
• Some user PCs occurred errors for low specs
• Renewal automatic grant system of census units

20
Future Works

• We should consider the followings
• How can we improve the rate of attendance to the
  e-Census?
• How can we minimize the errors in input?
• How can we elevate imputation techniques for this
  system ?
• How can we speed up for data processing?
• What should we prepare systematic supports for
  better services to the users of e-Census system?

21
Thank you very muchfor your watching

• Ryu, Gi Hyong,
• Kim, Kyu Young
• nonl_at_nso.go.kr
• kkyoung_at_nso.go.kr
• Korea National Statistical Office