The Italian Certification of Computer Science University Curricula

1
The Italian Certification of Computer Science
University Curricula
Agostino Cortesi, Università Ca Foscari -
Venezia Enrico Nardelli, Università Tor Vergata -
Roma GRIN - Italian Association of University
Professors in Computer Science
2
1. Motivations
2. Degree Program Certification
3. Certification Process Management
4. Towards a CS Body of Knowledge
5. Conclusions
3
Challenges for italian universities

• The autonomy granted to italian Universities in
  the specification of their curricula forces the
  faculties to highly qualify their products with
  respect to the plethora of courses that are
  offered in the educational market.
• Also when focussing just on curricula offered by
  the University system, it is very difficult to
  evaluate the values of curricula that look very
  similar.
• The same holds for enterprises and recruitment
  agencies, having hard time to classify the
  different Universities with respect to the
  quality of each undergraduate and graduate degree
  program

4
and for CS in particular

• Each italian University delivering a degree
  program in Computer Science should be fair enough
  
• to guarantee that their curricula cover the basic
  spectrum of scientific knowledge in Computer
  Science,
• and that courses are given by qualified
  professors.
• Unfortunately the second condition above is
  sometimes disregarded, as there are Universities
  degree programs where not even one faculty member
  has a PhD neither in Computer Science nor in
  Computer Engineering.

5
1. Motivations
2. Degree Program Certification
3. Certification Process Management
4. Towards a CS Body of Knowledge
5. Conclusions
6
The GRIN certification mark

• GRIN, the Italian Association of Computer Science
  University Professors (about 750 members)
  promoted a common effort involving almost all of
  the Italian Universities towards the elicitation
  of the product qualities of undergraduate
  degree programs in Computer Science
• it characterized the constraints to be fulfilled
  in order to obtain the GRIN quality
  certification.
• rules and results of this certification process
  are made public, in a web site
• An external professional certification authority
  (AICA) guarantees the fairness of the
  certifications process

7
A feasible approach

• The GRIN Association decided to adopt the
  following criteria, in order to avoid the need of
  heavy organizational duties
• The quantity of data to be treated should be
  quite limited data should be easy to get and to
  check
• data should be already available at each site, as
  part of the usual public information provided to
  potential students.

8
General Guidelines

• The GRIN quality certification is based on the
  verification of the fulfilment of a set of
  constraints on the programs.
• The general guidelines to define the
  certification rules are as follows
• The percentage of courses in Computer Science in
  the degree program has to be significant.
• The main areas of Computer Science should be
  properly covered
• The degree program should not too much focussed
  on a single area.

9
Constraints

• Two certification levels were designed
• the first one aimed at undergraduate degree
  programs (certificazione base)
• the second one for graduate degree programs or
  for very demanding undergraduate programs
  (certificazione avanzata).
• A list of 11 main Computer Science areas was
  identified, by the community, and a detailed list
  of subtopics (subareas) were associated to each
  of them.

10
The 11 CS Areas

• Foundations
• Algorithms
• Programming
• Computer Languages
• Computer Architectures
• Operating Systems
• Data Base Management Systems
• Network Computing
• Software Engineering
• Human Computer Interaction Graphics -
  Multimedia
• Knowledge Representation

11
Basic Rules

• The certification rules are defined in terms of
  credits (cfu 25 hours of learning activities
  for the average student).
• In order to be eligible for the GRIN quality
  certification mark, each University degree
  program in Computer Science must satisfy the
  following 3 constraints
• At least 78 cfu must be assigned to learning
  activities in Computer Science or in Computer
  Engineering.
• At least 60 cfu (out of the 78 above) must be
  assigned to learning activities in the 11 areas
  listed above
• At least 7 areas (out of the 11 listed above)
  must be covered by at least 6 cfu.

12
Rules vs Guidelines

• The first rule guarantees that more than 1/3 of
  the program is specifically dedicated to CS
  topics the rest might mathematical and physical
  foundations, more specialized topics, or
  complementary aspects (e.g., legal, economical,
  and ethical issues).
• The second rule guarantees a good coverage of the
  main areas of Computer Science observe that
  credits assigned to the same area can be spread
  among different courses.
• Finally, the third rule prevents from degree
  programs whose scope is too narrow at least half
  of the 11 areas must be properly covered, say by
  at least 48 hours of class lectures.

13
Additional Rule on Faculty

• The quality of the degree program heavily depends
  on Falculty qualification
• The minimal requirements
• At least 8 professors on Computer Science or
  Computer Engineering for an undergraduate program
• At least 6 professors for additional degree
  programs

14
Results

• Effort ongoing since 2002-03
• The web site is public
• http//grin.informatica.uniroma2.it
• 2004 39 certified programs out of 57 (univ.
  31/39)
• 2005 42 certified programs out of 54 (univ.
  35/42)
• 2006 44 certified programs out of 54 (univ.
  37/42)
• Each degree program pays an annual fee of 150
  Euros

15
1. Motivations
2. Degree Program Certification
3. Certification Process Management
4. Conclusioni
4.Towards a CS Body of Knowledge
5. Bibliografia
5. Conclusions
16
Process management

• Each year, the chair of each Computer Science
  degree program may apply for the GRIN quality
  certification by inserting in the certification
  web site the data concerning the activated
  curricula, and the syllabus of each course
  taught.
• For each course, the following information has to
  be provided total number of credits, number of
  credits labelled as computer science, and
  corresponding area.
• A synthetic description of the contents of each
  credit in the 11 areas has to be inserted as
  well.

17
(No Transcript)
18
(No Transcript)
19
course
areas
syllabus
20
Correspondence between curricula

• The system allow to discover semantic
  correspondences between different courses.
• In particular, the web site supports curricula
  recast when students transfer to a different
  university
• Based on sub-areas
• A finer partition of each of the 11 top-level
  areas
• starred sub-areas are recommended ones

21
A Foundations (Subareas)

• ALF Automata, Formal Languages
• CAL - Calcolability
• COM - Complexity
• SLP Programming Language Semantics
• TIC Information Theory
• L - Logics
• SD Dinamic Systems
• V - Varia

22
B Algorithms (subareas)

• SDF Fundamental Data Structures
• TAPA Basic Techniques for Analysis and Design
  of Algorithm
• A Fundamental Algorithms
• ASC Combinatoric Algorithms
• TAA Advanced Algorithms
• SDA Advanced Data Structures
• AD Distrubuted Algorithms
• AP Parallel Algorithms
• AN Numerical Algorithms
• V - Varia

23
1. Motivations
2. Degree Program Certification
3. Certification Process Management
4.Towards a CS Body of Knowledge
5. Conclusions
24

• There is a great value in the amount of
  information that has been collected this way in
  three years
• It is possible, in fact, to build in a bottom-up
  process an updated map of knowledge in Computer
  Science. This is object of current investigation
  by the GRIN National Educational Committee, aimed
  at defining an Italian Curriculum in Computer
  Science.
• This will be particularly useful in order to
  maintain a correspondence between the University
  curricula and the requirements of the work
  market.

25
1. Motivations
2. Degree Program Certification
3. Certification Process Management
4. Towards a Book of Knowledge
5. Conclusions
26
Conclusions

• The GRIN quality certification mark of University
  degree programs in Computer Science has a value
  both for the whole Italian Computer Science
  community, but also for the european University
  system, as it yields a system of comparable and
  transparent curricula that facilitates student
  mobility in a national or international context.
• It can be the basis toward the design of an
  Eurobachelor in Informatics