Information on Informatics - University of Houston–Downtown

Download Report

Transcript Information on Informatics - University of Houston–Downtown

Information on
Informatics
鍾青萍 P. Pete Chong
長庚大學
資管系客座教授
休士頓城中大學
馬太爾講座教授 (資管)
Informatics
Information Science (library science)
The sciences concerned with gathering and
manipulating and storing and retrieving
and classifying recorded information
Informatics
The application of computers and statistics
to the management of information. For
example, in the Human Genome Project,
informatics has permitted the
development and use of methods to
search databases quickly, analyze DNA
sequence information, and predict
protein sequence and structure from
DNA sequence data.
Informatics
信息学: 研究计算机和统计学技术在信息处理中的应
用的学科。在基因组计划中,信息学的内容包括
快速搜索数据库方法的开发、DNA序列信息分析
方法的开发和从DNA序列数据中预测蛋白质序列
和结构方法的开发。
資訊學: 研究電腦和統計學技術在信息處理中的應用
的學科。在基因組計劃中,資訊學的內容包括快
速搜索資料庫方法的開發、DNA序列資訊分析方
法的開發和從DNA序列資料中預測蛋白質序列和
結構方法的開發。
Informatics, the Bottom Line





Purpose (output): to predict protein
sequence and structure
Requirements: analyze for patterns
Knowledge needed: statistics
Tools needed: computer & software
Input: DNA sequence data
What data to collect? Why these attributes?
Domain Knowledge needed.
What is IS?



Management: allocation of resources
Information: data analysis
Systems: (1) Information Technology,
(2) Organization Restructuring due to
IT implementation
IS may help create efficiency, but it is
not efficient by itself without
planning
Where can IS help?
Value Chain
 Input Logistic
 Processing
 Output Logistic
 Marketing and Sales
 Services
“Value” in Informatics
Important: Accuracy and Validity
 Accuracy: reduce the cost of
measurement and verification
(including not having to verify)
 Validity: reduce the cost of testing so
the sample size my increase, or the
test may be refined
The Need of Operational
Definition



Without quantifiable attributes, IS
can only, at best, serve as a storage
service
The determination of attributes and
their values are assessed require
domain knowledge
Reengineering: converting
qualitative attributes to quantitative
ones
Quality in Healthcare
What is “quality”?
 What is the goal of healthcare?
 A list of measurement: immunization
ratio, life expectancy, days off work,
etc.
Really?

Quality in Healthcare
Education
Domain Knowledge: what to do and
how to do

Communication: sender -> medium
-> receiver; coding -> transmit ->
decoding
Wait…how do you define quality here?

Quality in Health Education


全民皆兵
全民皆醫?
Which part of “what to do” and “how to
do” will generate the highest benefit
and avoid disasters?
After we figure this part out, then what
devices can we use to implement the
system?
Quality in Health Education




Which part of such education should
be memorized? Kept as Reference?
Just-in-time Learning in Education
JIT Learning in Health Education?
The effect of mobile device (including
3G phones and wireless PDAs) to
“referencing”
Quality in Informatics



E(x) = Σ x*P(x)
To be able to “predict” we need to
know what attributes to observe and
have a way of measuring their
changes
We need to know the probability (i.e.,
the distribution) of each potential
values for the attributes
Researches in Informatics





Are data really normally distributed?
Bradford’s Law, Lotka’s Law, Zipf’s
Law, the Pareto Principle (the 80/20
Rule)
Skew Distribution?
Junk DNA shows Zipf’s Distribution
What does it mean?
Researches in Informatics



Can we classify items according to
their distribution?
Libraries with the same usage
pattern could be managed by the
same methodology regardless of
contents?
Illnesses with the same occurring
pattern may be handled by the same
protocol regardless of their nature?
Questions and [no] Answer?





What is Informatics?
What is the purpose of Informatics?
How do we know these “purposes”
are correct? Are there better ones?
What is healthcare (assuming that is
what bio-informatics are for)?
What is the purpose of healthcare? …