The Ontology Project

1
The Ontology Project

• Boots Cassel
• Active members of the committee
• Gordon Davies
• Rich LeBlanc
• Andrew McGettrick
• Heikki Topi

2
Current Activities

• Some of you know me, but many do not so, here
  is a summary of current interests.
• Member of ACM Education Board
• Ontology Project
• Curriculum review
• CC2001 revision
• MS level program recommendations
• Repository of Educational Materials
• CITIDEL
• Etc.

3
Overview

• Introduce the Computing Ontology Project
• Explore possible uses of the ontology
• Describe current and planned future curriculum
  recommendation developments

4
The Computing Ontology Project

• The Project
• An Ontology of all of the Computing and
  Information related subject areas
• Who
• Sponsored by the US NSF, ACM, IEEE-CS with
  participation and interest from a number of
  people from this side of the ocean.

5
A Problem?

• Our discipline is young, and growing fast
• Experts are independently defining and exploring
  various portions of the discipline
• These portions may overlap.
• There may be portions that are missed, at least
  for a time.

6
An opportunity?

• Because it is still relatively early in the life
  of the computing disciplines
• It may still be possible to coordinate the
  development and dissemination of knowledge about
  the various parts
• It may still be possible to keep the experts
  aware of each others work
• Avoid unnecessary duplication
• Be aware of missing pieces

7
What do we need?

• A complete representation of the entirety of the
  discipline
• In all its component parts
• In all its manifestations
• This representation
• Must be dynamic, responding to changes as they
  occur
• Must be easily accessible to all who can use it.

8
What do we have to work with?

• The ACM Computing Classification System
• A very important listing of the topics in the
  computing discipline
• First system produced in 1964
• Total 191 entries, including repeated general
  and miscellaneous
• Current system dated 1998, updated in 2006
• Lists topics, including subtopic relationships
  and a few isRelatedTo connections

9
More to work with

• The curriculum documents
• Produced by various societies
• ACM, IEEE-CS, AIS, IFIP
• Australian Computer Society
• German Accreditation for Informatics Programs
• British Computer Society
• Etc.
• Topics listed from the perspective of educational
  needs
• Relationships include
• Core educational requirements
• Topics that support learning outcomes
• Topic dependencies

10
Too much of a good thing?

• The lists from the various organizations are not
  consistent, nor are they complementary
• Overlap
• Contradict
• In some cases, even between different
  publications of the same groups!
• What is a conscientious curriculum developer
  supposed to do?

11
An Example

• An area that appears in most curriculum documents
  is database
• Heikki Topi drew a comparison between the topics
  related to databases in the ACM/IEEE-CS CC2001
  and the ACM/AIS IS recommendation
• Size of figure represents degree of emphasis

12
CS IM12
CS IM13
Hypertext and Hypermedia
Multimedia Informationand Systems
CS IM8
Distributed Databases
CS IM3
IS 1.6.2
CS IM7
CS IM4
Data Modeling
IS 3.3.2
Relational Databases
Transaction Processing
IS 1.6.6
IS 1.6.4
Data Models
Data Modeling
Application Interface
Integrity
CS IM5
IS 1.6.12
Database Query Languages
IS 1.6.5
Data Dictionary
Data DefinitionLanguages
IS 1.6.1
CS IM2
IS 1.6.9
DBMS features, functions, architecture
DBMS Products
IS 1.6.10
Database Systems
Database machines and servers
13

• And that does not include all levels of detail
• Goal an interactive structure for representation
  and exploration of the unified body of knowledge
  of all of the computing and information related
  disciplines

14
Desired outcomes

• Keep computing related disciplines as a single
  voice to maintain strength
• Maintain a current classification scheme for
  computing and information related work
• Address the challenges of updated curriculum
  recommendations
• Support curriculum development for creative new
  types of programs of study
• Ease the path toward accreditation for
  non-standard programs of study

15
Not just for curriculum

• This is not just for curriculum development and
  comparison
• Research work is tagged with appropriate topic
  areas to facilitate groupings and support
  searching
• If well done, this project could lead to more
  effective categorization of related research

16
What are we doing?

• Careful review of existing classification systems
• Perspectives of various sub cultures within
  computing
• Computer engineering, computer science,
  information systems, information technology,
  software engineering -- should we add Web
  Science?
• Consider both educational and research
  classification requirements

17
What is the difference

• between a taxonomy and an ontology?
• An ontology gives more than a list of terms with
  subtopic relationships
• An ontology also encodes relationships that allow
  inferences to be drawn, new knowledge created
  from what is encoded.

18
ACM CCS Top Level

• General Literature
• Hardware
• Computer Systems Organization
• Software
• Data
• Theory of Computation
• Mathematics of Computing
• Information Systems
• Computing Methodologies
• Computer Applications
• Computing Milieux

19
ACS Core Body of Knowledge

• Computer Organisation and Architecture
• Conceptual Modeling
• Database Management
• Data Communications and Networks
• Discrete Mathematics
• Ethics/Social Implications/ Professional Practice
• Interpersonal Communications
• Program Design and Implementation
• Project management and quality assurance
• Security
• Software Engineering and Methodologies
• Systems analysis and design
• Systems software

20
German Accreditation

• Automata and formal languages
• Algorithms and data structures
• Databases
• Operations systems
• Communication systems (particularly networks)
• Computer architecture
• Programming engineering
• Software engineering (particularly modeling) and
  project management
• Projects with high software engineering content

21
And more

• Of course, there are others.
• The point is that these are all related, but they
  express similar ideas differently and have some
  substantive differences
• That is fine, but it would be nice to see how
  they are similar and how they differ.
• Compare to a common underlying base.

22
Simplification

• Can we get a less complicated list of topics that
  defines what the computing and information
  disciplines are about?

23
Basic elements

• What are we about?
• Hardware
• Software
• Information
• Human related aspects
• Human Factors is a nice term, but has too much
  meaning attached to it.
• But that is not enough to describe all that
  concerns the computing and information related
  disciplines

24
Topics of interest include all the core elements,
and every possible combination of them
HW Hu SW
HW
HW SW
HW Hu
HW Hu Inf
HW SW Inf
SW
Hu
Hu Inf
SW Inf
Inf
Hu Inf SW
But, that does not tell the whole story either.
25
For much of what we do, a system view of the
parts is essential
HW Hu SW
HW
HW SW
HW Hu
HW Hu Inf
HW SW Inf
SW
Hu
Hu Inf
SW Inf
Inf
Hu Inf SW
System
But, not enough
26
The Application Context is often an inseparable
part of the field.
HW Hu SW
HW
HW SW
HW Hu
HW Hu Inf
HW SW Inf
SW
Hu
Hu Inf
SW Inf
Inf
Hu Inf SW
System
But, there is more
Application Context
27
Administer Support
Design, Combine, Build, Test
HW Hu SW
HW
HW SW
HW Hu
HW Hu Inf
HW SW Inf
SW
Hu
Hu Inf
SW Inf
Inf
Hu Inf SW
System
Application Context
Social context
28
The challenge

• A multidimensional problem
• Basic elements
• Views
• Applications of views
• Unrestricted combinations

29
Topics and Outcomes

• In the Education domain
• A comprehensive set of topics does not begin to
  solve all the problems we listed.
• The topic space is a tool to be incorporated into
  applications that address these issues.
• It does not stand on its own.

30
Beyond the body of knowledge

• IFIP WG 3.2 identified seven aspects of
  curriculum development
• Body of Knowledge
• Foundational Material
• Application Context
• Social Context
• Breadth and Depth
• Thematic Coherence
• Outcomes

The Ontology is a representation of the BoK
Rules in the Ontology identify BoK elements with
other aspects of this overall scheme
L Cassel, G. Davies and D. Kumar. The shape of an
evolving discipline in Informatics Curricula and
Teaching Methods, L N Cassel and R Reis, Eds.
Kluwer Academic Publishers. pp131-138
31

• The body of knowledge for a given curriculum or
  program of study is drawn from the entirety of
  the related disciplines.
• The other aspects must be combined with the BoK
  elements to form a viable program.

32

• Imagine possibilities
• Map a program of study onto the entirety of the
  disciplines.
• Explore for the underrepresented areas.
• Demonstrate the coverage of a planned curriculum.
• Illustrate the distinctions and overlaps between
  existing and proposed specialties and related
  disciplines.

33
Proficiency and Competency

• Curriculum development moving toward an outcomes
  orientation.
• Dont tell me what topics you studied, tell me
  what you can do.
• Clearly, an organized study of subject areas
  contributes to gaining proficiency in some area.

34
Learning outcomes
Configure new wireless router
35
Other uses

• Research classification
• Imagine possibilities
• Researcher submitting a work
• Researcher with a new idea, looking for related
  work and gaps in the knowledge base
• Graduate student looking for a good topic
• etc.

36

• Classification not a static labeling, but an
  insertion into a dynamic environment.
• Search not just for a particular paper or author,
  but for what is around a topic area.
• Exploration for areas that are richly filled and
  others that are sparse.

37
Current Status

• Good list of topics
• Never complete, dynamic
• Organized as a taxonomy
• Working on establishing the rules and
  relationships that change a taxonomy into an
  ontology

38
Example efforts
An Architectural Model
39
The (current) top level

• Algorithms Complexity
• Computer Hardware Organization
• Computer Network Systems
• Computing Education
• Discrete Structures
• Ethical Social
• History of Computing
• Graphics, Visualization and Multimedia

• Information Topics
• Intelligent Systems
• Mathematical Connections
• Programming Fundamentals
• Programming Languages
• Security
• Systems Development
• Systems Project Management
• User Interface

It is too many topics for the top level. Further
reduction is needed and expected.
40
Visualization

• The ontology is only as useful as the ways we
  have to interact with it.
• Different views will be needed for different
  purposes.

41
An illustration -- centered on hashing
42
The size quickly gets out of hand -- need
automatically generated and scalable
representation
43
More recent top level

• Main

http//harshakiron.tripod.com/ACM/july/julyversion
3edit.swf
44
(No Transcript)
45
Remember this? We need to make a very complex
underlying structure produce just what is needed
for a given purpose.
Learning outcomes
Configure new wireless router
46
Whats next

• Enlisting the help of others
• In subject areas
• In educational needs
• To establish the right linkages, and the right
  level of detail to include.
• Encode the information about the topics and the
  relationships into a standard OWL file so that
  applications can be built against it
• Develop prototype applications to demonstrate the
  potential so that others will develop more.

47
A first real application

• The problem of developing curriculum
  recommendations
• A new approach in development

48
Curriculum Recommendations

• ACM and IEEE-CS collaborate on Computer Science,
  Computer Engineering, Software Engineering
• ACM and AIS collaborate on Information Systems
• ACM SIG ITE drafting Information Technology

49
Limitations

• These cover the most common program titles, at
  least in the US and much of the world.
• More program types are emerging and the current
  process cannot address them.

50
The Problem

• Last comprehensive curriculum,
• CC2001-CS, released in December 2001
• Committed to 'interim review' in 5 years
• Times up

Next few slides produced by Larry Snyder,
Washington University. ACM Education Board,
curriculum committee
51
Traditional Curriculum Processa/k/a Bad Old
System

• Form large committees
• Hold large meetings
• Centralized editorial committee
• Very SLOW very EXPENSIVE
• Do it every decade

52
Need A New Approach

• Goals
• Be faster, more agile, more flexible
• Smooth process to avoid cataclysmic change
• Respond to fast pace
• Keep as current as possible
• Exploit state-of-the-art technologies
• Expand the community that participates
• Increase the ways one can participate
• Less authoritarian, more consensual

53
Evolutionary Curriculum Revisionsa/k/a Good New
System

• Community based
• Online
• Moderated by Education Board Committee
• Do it continuously

54
Target Structure

• Three document Web-based process

Existing Curriculum Current Recommendation
Ontology
Curriculum Wiki Wikipedia-like Contributions
Discussion Forum SlashDot-like Dialog
55
Continuously Reviewed Curriculum

• Deployed Curriculum is fixed
• Curriculum Wiki is a place to add/revise
• Discussion focuses dialog on the revisions

Steering Committee seeds and steers discussion,
decides when and how changes are accepted,
commits next version
Existing Curriculum Current Recommendation
Ontology
Curriculum Wiki Wikipedia-like Contributions
Discussion Forum SlashDot-like Dialog
56
Reaching the Goal

• CC2001-CS needs attention immediately
• Infrastructure not yet built to deploy
  incremental structure
• Need an interim update
• Build Web page for community feedback
• Solicit comments prior to starting revisions

Community Contributes Comments On CC2001-CS
CC2007-CS
57
Timeline

• Plan for moving from here to there

January 2008
July-Dec 2007
March 2007
Committee uses input to develop CC2007, to become
the existing curriculum
Community Contributes Comments CC2001-CS
Existing Curriculum Current Recommendation
Ontology
Curriculum Wiki Wikipedia-like Contributions
Discussion Forum SlashDot-like Dialog
Community had until June 30 to comment on
CC2001-CS
58
Whats coming?

• Review of CC2001 is in process
• Additions likely -- more attention to security
• Important considerations
• The recommendations are that -- they are not
  prescriptive.
• The recommendations form a starting point
• Exact curriculum must be responsive to the needs
  of the institution, the faculty, and the student
  body

59
Input still welcome

• The revisions are not yet final.
• Input is still welcome
• Send your thoughts to me (cassel_at_acm.org)
• Subject line Curriculum review
• Expected completion date end of 2007

60
Only Computer Science?

• This revision is only addressing CC2001 -- The
  Computer Science volume
• Others are newer, less in need of review -- but
  not entirely right
• When we get the continuous review process
  working, the others will be included.

61
Discussion topics?

• Feelings about a common reference point
• Note not a common idea of what constitutes a
  good or appropriate program!
• Priorities
• What types of information is most urgently
  needed?
• How can this project help?

62
Contributors

• This work is part of a project funded by a grant
  from the National Science Foundation, the ACM
  Education Board, and the IEEE-CS.
• The Team Boots Cassel (chair), Jim Cross, Gordon
  Davies, Reza Kamali, Eydie Lawson, Rich LeBlanc,
  Andrew McGettrick, Russ Shackelford, Bob Sloan,
  Heikki Topi,
• Also contributing Fred Mulder, and Anneke
  Hacquebard, Maarten van Veen
• Growing list of interested parties

63

• Follow our progress
• what.csc.villanova.edu/twiki/bin/view/Main/Ontolog
  yProject