Transcript 01 - School of Computing

COMP3740 CR32:
Knowledge Management
and Adaptive Systems
Introduction
By Eric Atwell, School of Computing,
University of Leeds
S1: Eric Atwell
Office: 6.06a
[email protected]
S2: Vania Dimitrova
Office: 9.10p
[email protected]
http://www.comp.leeds.ac.uk/eric http://www.comp.leeds.ac.uk/vania
http://www.comp.leeds.ac.uk/nlp http://www.comp.leeds.ac.uk/agc/krgroup.html
Semester 1 Topics in KM
• Knowledge in Knowledge Management
– the nature of knowledge, definitions and different types
– Knowledge used in Knowledge Based Systems, KM systems
• Knowledge and Information Retrieval / Extraction
–
–
–
–
–
Analysis of unstructured data in text/WWW
IR: finding documents which match keywords / concepts
IE: extracting terms, items, facts (DB-fields) from documents
Matching user requirements, advanced/intelligent matching
Mining WWW as source of data and knowledge
• Knowledge Discovery
–
–
–
–
Collating data in data warehouse; transforming and cleaning
Cross-industry standard process for data mining (CRISP-DM)
OLAP, knowledge visualisation, machine learning, data mining
Analysis of WWW-sourced data
Lectures and Assessment
• CR32 Lectures: *NORMALLY*…
– Friday 10-11 LT09, 11-12 LT08
• Watch for EMAIL announcements of changes:
– No lectures SoC Half Term; possible illness; additional
Guest Research Seminars; etc…
• Assessment
– Coursework exercises during term-time
– Jan mock exam, May Exam: 3 hrs, open book
– Students are expected to perform to an acceptable level in
coursework in order to pass related Exam questions.
Commercial Breaks
• A break from the lecture / (?lecturer…?)
• up to 5 minutes (MAX) for:
– Interesting KM application or tool
– E.g. KM company, product, research project, website,
users/market, news story, …
– E.g. RELEVANT YouTube video …
• Email [email protected] to volunteer
• Peer appreciation! Does not “count” in Grade
Learning Resources
• Technical problems with Blackboard VLE
• Web site: http://www.comp.leeds.ac.uk/eric/cr32
• /home/www/eric/cr32
– Powerpoint slides
– Selected paper(s) to read
• Electronic copies of slides available from web-site
after the lecture
• Students as learning resources: talk to each other!
• … and talk to me; [email protected]
Three Questions you can answer:
• What is knowledge?
• What does it mean to manage / discover knowledge?
• How can information technology help?
What is Knowledge?
• A range of definitions:
• CED: Collins English Dictionary
• LDOCE Longman Dictionary of Contemporary English
- “simpler” definitions for English language learners
(and Language Engineering, eg IR, IE, text analytics)
• Online definitions: Google, Wikipedia
• Knowledge in Knowledge Management
• Knowledge in Knowledge Based Systems
Collins English Dictionary:
1. The facts, feelings or experiences known by a person
or group of people.
2. The state of knowing.
3. Awareness, consciousness, or familiarity gained by
experience or learning.
4. Erudition or informed learning.
5. Specific information about a subject.
6. Sexual intercourse (carnal knowledge).
Which types of knowledge can a computer handle?
LDOCE: Longman Dictionary of
Contemporary English
1. the facts, skills, and understanding that you have
gained through learning or experience
2. knowing that something has happened or is true
4. information that you have about a particular
situation, event etc.
See also general knowledge, common knowledge,
working knowledge
Which of these can a computer handle?
LDOCE as a text database
- We have a version of LDOCE in a database
- actually, a text-file, with mark-up showing records, fields
structure http://www.comp.leeds.ac.uk/eric/ldoce
- 1978 LISP Markup predates XML, HTML etc:
((headword)
(<field-no> <data>)
(<filed-no><data>) …
)…
- This illustrates some challenges of DATA MINING:
UNDERSTANDING, CLEANSING, MODELLING…
- (see CRISP-DM methodology for Data-Mining)
Knowledge in
Knowledge Management
Three meanings:
• the state of knowing or to be acquainted or familiar with
(“know about”)
• the capacity for action (“know how”)
• codified, captured and accumulated facts, methods,
principles and techniques.
Based on: F Nickols. 2000.
The Knowledge in Knowledge Management. KM Yearbook.
Knowledge in
Knowledge Management
Three meanings:
• the state of knowing or to be acquainted or familiar with
– an IT System does not “know about” its data
• the capacity for action
– an IT System does not “know how to do” anything
• codified, captured and accumulated facts, methods,
principles and techniques. – maybe IT could store this?
• This is a Computing module … focus on IT for KM
Knowledge Based Systems
Atwell, E S (editor) Knowledge at Work in
Universities - Proceedings of the second annual
conference of the Higher Education Funding
Council's Knowledge Based Systems Initiative,
146pp University of Leeds Press. 1993.
No definition of KBS, except by examples…
KB: facts and logical rules for inferring new facts
Eg: if (sun=yes) and (humidity=low) then play=yes
…but also Info Retrieval, language/speech/image
Explicit Knowledge
• Knowledge that has been articulated
–
–
–
–
–
–
product specifications
scientific formulae
computer programs
patents
documented best practice
Handbooks
• Could be stored in a KB (if we can solve
problems of data capture / transformation, …)
Tacit Knowledge
• Knowledge that cannot be articulated (eg in a KB)
is called tacit knowledge
–
–
–
–
how to ride a bicycle
how we recognise a face
How to understand an English sentence / document
how to create a work of art
• You could say AI is trying to recast Tacit knowledge
as Explicit knowledge – eg rules to process English
sentences
Also … Implicit Knowledge
• Knowledge that could be articulated but hasn’t (yet)
… is called implicit knowledge.
– knowledge engineers and systems analysts are trained to
identify implicit knowledge and to help experts articulate
their knowledge.
• Could be stored in a KB (if we can solve problems
of data capture / transformation, …)
Also … Cultural Knowledge
• The knowledge an organisation has about itself
and its environment (“meta-level knowledge”?)
• Shared beliefs, norms and values.
• A framework within which organisational
members construct reality
• Required to
– understand and use facts, rules and heuristics
– To make inductions in the same way as others in order
to enable concerted action
Terminology and Ontology
• Ontology: the “concepts” in a discipline,
• and meaning-relations between these concepts
• “concepts” roughly equates to terminology –
specialist words and phrases in a discipline
• Terminology and Ontology encode knowledge in the
discipline
From Data to Knowledge
KNOWLEDGE
high
INFORMATION
Order/
Structure
DATA
low
SIGNALS
low
Human Agency
high
Different perspectives of data?
• Data mining: finding patterns (knowledge) in data
• Knowledge discovery: finding knowledge (in data)
• Database: stores data
• Information management: making use of data
• Knowledge Management: finding and making use of
patterns in data, “taming the data”
Conclusions
• KM is not just about Database technologies!
• KM is about acquiring data, cleansing data,
extracting useful structure or knowledge from data
• BUT not just data in databases…
• … most data/information on WWW is unstructured
TEXT, though HTML/XML markup may help (a bit)
• … so we need to extract, clean, reformat data from
WWW into DB-like format (fields/attributes,
records, tables) for data-mining.
Questions to think about…
• Can you distinguish knowledge from information?
• Think up examples of explicit, tacit, implicit and
cultural knowledge that exist in the School of
Computing.
• Do we manage knowledge by managing information,
or is there more to it than that?
• Can knowledge be ‘created’ or only acquired?
• Think of some applications of KM.
http://www.youtube.com/watch?v=uz0KXaflY2w
• Does this really “explain knowledge management”?