Slide Mestre

1

• Assuring Data and Information Quality in Sharing
  Process of Population and Health Data (eHealth
  Systems)

Ying SuISITC, Beijing, CHNsuy.rspc_at_istic.ac.cn
Ling YinHospital 301, Chinayinling301_at_126.com
Institute of Scientific and Technical Information
of China (ISTIC) Led by the Ministry of Science
and Technology Funded in October,
1956 Information Quality Lab (IQL) delivering
information quality services focused on
facilitating decision-making processes and on
improving customer satisfaction.
2

• Problems
• Information Quality in Chinese Hospital
• Data Quality in Chinese Information Systems

Key Themes

• Solution
• Framework for assuring IQ in an eHealth context
• to specify their IQ requirements by Semiotics
• introduced Coupling and Explanation models

• Methodology
• Describe information within a process
• Calculate IQ and process performance
• Validate the impact relationships by simulation

• Results
• Reputation, Believability and Trace-ability,
• IQ is critical to patient care
• Quantifiable IQ and PP indicators.

• Further work
• Whats next?

3
Information Quality Problems in Chinese Hospitals

• The phenomenon of "three-long, one-short
• three-long the time of registration, waiting to
  see the doctor and getting the medicine
• one-short getting the treatment

4
Data Quality Problems in Chinese Information
Systems -Clinical Pathways for Acute Coronary
Syndromes in China (CAPCS)

• ????????
• ???????????????

5
CPACS????
?? 4/3, 4/2
75 ?? 50 ???? 25 ????
??? 2/3
?? 4/3, 1/2
??? 3/3, 1/2
?? 2/3, 3/2
?? 3/3, 1/2
?? 4/3
?? 3/3,1/2
?? 3/3
?? 3/3, 3/2
?? 2/3,2/2
?? 2/3
?? 3/3, 4/2
?? 1/3, 4/2
?? 2/3, 2/2
?? 4/3
?? 4/3
6
????????-????
7
IDQ Problems Try to Solve

• How to describe information and related data
  within a process, and how to describe the
  controllable factors among them?
• How to calculate information quality and process
  performance?
• How to build the impact relationship between the
  indicators above and then verify?

8
Objectives of this presentation

• Propose an extensible IQ semiotics containing
  basic domain-independent IQ terms, upon which
  definitions of domain-specific concepts can be
  built.
• IQ descriptions for specific resources need to be
  computed and associated with those resources.
  This can be done by attaching origin information
  to the RDF explanation instances.
• Resources include data and services both of
  these kinds of resource are modeled by concepts
  in the IQ semiotics, so that the semiotics can
  express which kinds of IQ descriptor make sense
  for which kinds of resource. We refer to these
  relationships as couplings, which can be captured
  using an RDF schema

9
An IQ Assurance Framework
10
Basic Semiotics Structure

• In the semiotics, we model IQ concepts by
  introducing Quality Assurances (QA) these are
  decision procedures that are based upon some
  Quality Evidence (QE), which consists either of
  measurable attributes called Quality Indicators,
  or recursively, of functions of those indicators,
  Quality Metrics. Three main sources of indicators
  are common in practice
• Origin metadata, which provides a description of
  the processes that were involved in producing the
  data.
• Quality functions that explicitly measure some
  quality property, these functions are typically
  available from toolkits for data quality
  assessment with reference to specific issues.
• Metadata that is produced as part of the data
  processing.

11
Methodology

• We model the indicator-bearing environment as a
  collection of Data Analysis Tools that may
  incorporate multiple Data Calculation functions,
  and which are applied to some Data Entity.
• Indicators are either parameters to or output of
  these analysis tools. A QA is applied to
  collections of data items, which are individuals
  of the Data Entity class, using the values for
  the indicators associated to those items. The
  practical quality metrics are part of the output
  of a calculation function called QMCalculator,
  used in the IQA Calculator Analysis Tool.
• A quality metric called IQA Calculator Ranking
  associates a score to each data in the set, using
  a function of indicators. This score can be used
  either to classify data as acceptable/non
  acceptable according to a user-defined threshold,
  or to rank the data set. Here we will assume that
  our decision procedure is an grade function
  called QA-Func, that provides a simple binary
  grade of the data set according to the
  credibility score and to a user-defined threshold.

12
Classes and Relationships Introduced

• Summary of the classes and relationships
  introduced above, using informal notation for the
  sake of readability user-defined axioms.
• Quality-Assurance is based on Quality-Evidence
• Quality-Indicator is-a Quality-Evidence
• Quality-Metric is-a Quality-Evidence
• Quality-Metric is based on Quality-Indicator
• Quality-Evidence is output of Data-test-function
  
• Data-analysis-tool is based on Data-test-function
  

13
Overview of the IQA coupling model
14
Structure of Explanation Model
15
eQualityHealth Program NSFC-MOSTGoal and
Service Oriented Approach to Assure Data and
Information Quality in eHealth Systems
16
eQualityHealth

• eQualityHealth is a metadata platform for quality
  assessment
• eQualityHealth allows the definition of
  high-level quality goals and the specialization
  of typical measurement services according to
  quality goals

17
eQualityHealth Architecture
personalization
binding
references
Information Systems Meta-Model
General Quality Meta-Model
Personalized Quality Model (PQM)
Quality Service 1
Service Description

QFoundation
PQM
Service Description
QManagement

Store
Get
Search
Quality Requirements
Service Registry (UDDI)
QMediator
Quality Service n
Delegate
18
eQualityHealth Catalog

• eQualityHealth provides an extensible catalog of
  quality metrics, which presents general quality
  concepts and behaviors
• It also provides a catalog for the services that
  implement the quality metrics

19
Quality Catalog
Quality Metrics
Quality Dimensions
Quality Factors
20
Adding Quality Dimension Consistency
21
Adding Quality Dimension Consistency
22
Adding Quality Factor Consistency
23
Adding Quality Factor Consistency
24
Adding Quality Metric Consistency Ratio
25
Adding Quality Metric Consistency Ratio
26
Web Services in eQualityHealth

• Any quality service can be used in eQualityHealth
• Relevant quality methods not published as web
  services can be
• Methods embedded in quality tools
• Code libraries containing quality methods

Web Service
Web Service
Web Service
Adapter
Library
Quality Tool
API
Core
27
Services Catalog
28
Hospital operating room simulation model
Results
Locations Entities (Documents, people, or phone
calls should be modeled as entities.) Resources (a
person, equipment, device used for transporting
entities, performing operations, performing
maintenance on locations) Path Networks Processing
Arrivals Shifts Breaks Cost
29
Assumption of impact relationship of IQ to PP
The hypotheses of the effect relationship of
information quality to process performance
Results

• Takes Reputation as an example

30
Changzhou Case
EHR
Information portal
Health Call center
Wireless, Medical Devices, Database, Internet
Health Service Organization
30
31
Next StepsBlueprint of Human-centered eHealth
FurtherWork
32
The 6th International Conference on Cooperation
and Promotion of Information Resources in Science
and Technology (COINFO11)International
Workshop on Information Data Qualityhttp//coin
fo.istic.ac.cn/coinfo11/November 11-13, 2011,
Hang zhou, Paradise in ChinaThanks
33
Thanks for your Listening

• Dr. Ying Su
• Institute of Scientific and Technical Information
  of China
• Associate Professor (suy.rspc_at_istic.ac.cn )
• Director-in-Charge, IQL (Information Quality Lab)
• Post-Doctor, SEM (School of Economics and
  Management)
• Tsinghua University suy4_at_sem.tsinghua.edu.cn
• Co-Chair of International Conference on
  Information Quality(ICIQ), 2010
• Visiting Professor, UNIVERSITY OF ARKANSAS AT
  LITTLE ROCK (UALR)
• Invited by Professor John Talburt
• Advisor for the Master of Science in Information
  Quality program
• Director, UALR Laboratory for Advanced Research
  in Entity Resolution and Information Quality
  (ERIQ)
• Smart eHealth Program between Provinces, CHINA
  and ARKANSAS, US
• Email jrtalburt_at_ualr.edu Phone [phone]