1. 1 Automatic Generation of Exercises for Self-testing in Adaptive E-Learning Systems: Exercises on AC Circuits Third International Workshop on Authoring of Adaptive and Adaptable Educational Hypermedia Amsterdam, The Netherlands, July 19 th, 2005 Paul Dan Cristea, Aurora Rodica Tuduce University POLITEHNICA of Bucharest Spl. Independentei 313, 060042 Bucharest, Romania, Phone/Fax : +40 - 21- 316 95 68, 694 e-mail:pcristea@dsp.pub.ro AIED 2005 12 th International Conference on Artificial Intelligence in Education

2. 2 1.Introduction Learning Modalities Need for Intelligent e-Learning Systems 2.System architecture Pilot System Multiagent Structure Architecture of ILE Pilot 3.Basic Tools Learner Profile Eliciting Tool Question Apprisal Learning Item Apprisal & Status Point and Acceptance Propagation 4.Automatic Generation of AC Electric Circuit Problems 5. Implementation & Web Accessibility 6. Conclusions

3. 3 Combine the traditional style of teaching with the problem-based style: learning by being told, problem solving demonstration, problem solution analysis, problem solving, creative learning Learning Modalities

4. 4 Dramatic change of the target public for training Dramatic change of the target public for training Professional qualification is no longer a life-long achievement Professional qualification is no longer a life-long achievement Complex knowledge and skills have to be Complex knowledge and skills have to be transmitted and acquired efficiently transmitted and acquired efficiently Open and Distance Learning play a continuously increasing role Open and Distance Learning play a continuously increasing role e-Learning

5. 5 Intelligent educational tools can bring the flexibility and adaptability required to actively support the learner;Intelligent educational tools can bring the flexibility and adaptability required to actively support the learner; Increase efficiency of learning and further motivate learners by giving them a set of intelligent tools that will actively support them in the learning endeavour;Increase efficiency of learning and further motivate learners by giving them a set of intelligent tools that will actively support them in the learning endeavour; Promote participative and collaborative learning ;Promote participative and collaborative learning ; Offer learners individualised learning according to elicited learner profiles.Offer learners individualised learning according to elicited learner profiles. Intelligent e-Learning

6. 6 Significant research and implementation effort has been dedicated to develop Intelligent Tutoring Systems and Adaptive Hypermedia, able to adapt to learners objectives, interests, and preferences, i.e., to a Learner Profile (LP). To implement adaptivity, an ILE needs a quite complex structure, with several parallel version of the same learning item (LI), allowing many different learning paths to be selected in accordance with the LPs. Considerable additional effort in elaborating teaching materials, might require several authors and might need institutional support, but brings the advantage of real flexibility and adaptability. A course is not a flat juxtaposition of learning items, but a multilevel structure with many branches, along which the ILE recommends an optimal path for a user or for a class of users. Intelligent e-Learning (cont)

7. 7 Authoring learning material and building the structure of adaptive systems tends to become too complicated for the average teacher. Portability – the ability to deploy the content of a system on any other system, Reusability – the ability to store, search and retrieve LIs, including lessons, modules, exercises, activities for reusing, are strictly necessary for an efficient implementation and for a wide scale acceptance of the concept. Intelligent e-Learning (cont)

8. 8 The system is learner centred, all human and artificial agents being focused on achieving the learning-training tasks. Human agents: students, authors of teaching materials, tutors, course administrators, system administrator(s). The pilot web oriented ILE has a server-client distributed multiagent hybrid architecture Pilot System Structure

9. 9 Architecture of ILE pilot

10. 10 Learner Profile Eliciting Tool

11. 11 Learning Objectives Control ModuleCommunication Module Learners Profile Eliciting Tool Student input Registration form Questionnaires Learning Modalities Knowledge Watch Curricular study for a diploma Complementary study Executive up-dating Specialist up-dating Problem centered Test oriented Preferredly / Predominantly: Descriptive Demo Analytical details Practical aspects Examples Multimedia / Text Material to study 1 First Chapter xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 1.1 Section 1.1 xxxxxxxxxxxxxxxxxxxxxxxxxxx 1.1.1. Paragraph xxxxxxxxxxxxxxxxxxxxxX 1.1.2. Paragraph xxxxxxxxxxxxxxxxxxxxxX 1.1.3. Paragraph xxxxxxxxxxxxxxxxxxxxxX 1.2 Section 1.2 xxxxxxxxxxxxxxxxxxxxxxxxxx 1.2.1. Paragraph xxxxxxxxxxxxxxxxxxxxxx 1.2.2. Paragraph xxxxxxxxxxxxxxxxxxxxxX 1.2.3. Paragraph xxxxxxxxxxxxxxxxxxxxxX 1.3 Section 1.3 xxxxxxxxxxxxxxxxxxxxxxxxxxx 1.3.1. Paragraph xxxxxxxxxxxxxxxxxxxxxx 1.3.2. Paragraph xxxxxxxxxxxxxxxxxxxxxx 1.3.3. Paragraph xxxxxxxxxxxxxxxxxxxxxx 2 Second Chapter xxxxxxxxxxxxxxxxxxxxxxxxxxxx 2.1 Section 2.1 xxxxxxxxxxxxxxxxxxxxxxxxxxx 2.1.1. Paragraph xxxxxxxxxxxxxxxxxxxxxx 2.1.2. Paragraph xxxxxxxxxxxxxxxxxxxxxX 2.1.3. Paragraph xxxxxxxxxxxxxxxxxxxxxx ………………………………… Studied material 1 First Chapter xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 1.1 Section 1.1 xxxxxxxxxxxxxxxxxxxxxxxxxx 1.1.1. Paragraph xxxxxxxxxxxxxxxxxxxxxxxxxxxxx 1.1.2. Paragraph xxxxxxxxxxxxxxxxxxxxx 1.1.3. Paragraph xxxxxxxxxxxxxxxxxxxx 1.2 Section 1.2 xxxxxxxxxxxxxxxxxxxxxxxxxx 1.2.1. Paragraph xxxxxxxxxxxxxxxxxxxx 1.2.2. Paragraph xxxxxxxxxxxxxxxxxxxxx 1.2.3. Paragraph xxxxxxxxxxxxxxxxxxxxx 1.3 Section 1.3 xxxxxxxxxxxxxxxxxxxxxxxxxx 1.3.1. Paragraph xxxxxxxxxxxxxxxxxxxx 1.3.2. Paragraph xxxxxxxxxxxxxxxxxxxxx 1.3.3. Paragraph xxxxxxxxxxxxxxxxxxxx 2 Second Chapter xxxxxxxxxxxxxxxxxxxxxxxxxxx 2.1 Section 2.1 xxxxxxxxxxxxxxxxxxxxxxxxxx 2.1.1. Paragraph xxxxxxxxxxxxxxxxxxxxx 2.1.2. Paragraph xxxxxxxxxxxxxxxxxxxxx 2.1.3. Paragraph xxxxxxxxxxxxxxxxxxxxx ………………………………… ? Standard Path Recommended Path Content Management Mandatory Testing Contribution to Collaborative Learning Tutor input On-line students monitoring Validation of students proposals Self Testing Student Tracking Tool Learner Profile Eliciting Tool (details)

12. 12 Question # 5 Text for question # 5 Figure for question # 5 Test for section 5.1. Number of questions 10 Time Submit Learner's test window

13. 13 Question appraisal Sum of points for a question Q - the set of selected options at question Q. Correct choices positive points, Wrong answers negative points. Assigning negative points to wrong choices discourages guessing. Points acknowledged for question Q T(Q) - the threshold for the acceptance of the reply to Q

14. 14 Sum of points for a learning item LI C (LI) – the children of LI. The points obtained for LI are transferred upwards Points acknowledged for a learning item LI T(LI) - threshold A(LI) - award for the successful completion of the study of LI Learning item appraisal

15. 15 Status of the learning item LI 0 – pending, 1 – studied, Down-propagation of the acquired knowledge confirmation Learning item status

16. 16 Points obtained for choices C from the set of options O(Q) pertinent to a certain question Q are recorded at the LI to which the question is attached and transferred upwards. Point and acceptance propagation

17. 17 Automatic Generation of AC Electric Circuit Problems 1.Problem set description 2.Tree generation 3.Cotree generation 4.Tree plot 5.Graph plot 6.Circuit parameters and variables generation 7.Converting voltage sources to curent sources 8.Introducing controlled sources

18. 18 OBJECTIVESOBJECTIVES Design and develop a software able to automatically generate large sets of circuit analysis problems, all with the same general features, but having different topological structures and parameters of the circuits. Conditions: The problems are for use both during the tutorials and for examinations, thus -- despite the inherent risk for an engineering perception of reality -- all parameters and variables describing. the circuits should be integers to facilitate the computational task. Problems and solutions should be stored automatically on disk in distinct directories. Files referring to the same problem (text, graphics, etc) will have related labels. The system will be developed for making it accessible on the web.

19. 19 Problem Set Description Problem Set Description Choosing the parameters of the set of AC problems to generate. % 1 2 3 4 5 6 7 param = query({'311_CA_21.11.2004', '30','1','RO', 'd', 'g', 'no'},... { 'SetID - problem set label (Year of Study/Group ID/Date)',... % 1 'Nproblems - number of problems',... % 2 'StartID - ID of the first problem',... % 3 'Language - RO/EN',... % 4 'Out_medium - s = save on hard, d = display'... % 5 'Represent - t = tree, g = graph, b = both, other char = none'... % 6 'Entropy - yes/no = compute and display graph entropy'... % 7 },... 'Set Parameters');

20. 20 Choosing Variables & Independent Parameters % 1 2 3 4 5 6 7 8 9 10 11 12 13 param = query({'4','7','4','4','1', '4', '4', '0', '1', '4', '4', '2', 'Y'},... { 'Nnodes - number of nodes',... % 1 'Nbranches - number of branches',... % 2 'I_chord_a_max - maximum absolute value of chord current active components [A]',... % 3 'I_chord_r_max - maximum absolute value of chord current reactive components [A]',... % 4 'R_twig_min - minimum value of twig resistences [Ohms]',... % 5 'R_twig_max - maximum value of twig resistences [Ohms]',... % 6 'X_twig_max - maximum absolute value of twig reactance [Ohms]',... % 7 'E_twig_max - maximum absolute value of twig Re & Im emf-s [V]',... % 8 'R_chord_min - minimum value of chord resistences [Ohms]',... % 9 'R_chord_max - maximum value of chord resistences [Ohms]',... %10 'X_chord_max - maximum absolute value of chord reactance [Ohms]',... %11 'nJ - number of branches with current sources',... %12 'CrossLinks - Y/N - mutual inductances and controlled sources'... %13 },... 'Circuit Variables & Independent Parameters');

21. 21 Mutual inductive couplings and controlled source parameters if strcmp(lower(CrossLinks), 'y') % 1 2 3 4 5 6 7 8 9 10 11 12 13 14 param = query({'0', '0', '0', '0', '0', '3', '0', '3', '0', '5', '0', '5', '0', '4'},... { 'nEI - number of current controlled voltage sources E = Zt * I',... %1 'nJU - number of voltage controlled current sources J = Yt * U',... %2 'nEU - number of voltage controlled voltage sources E = A * U',... %3 'nJI - number of current controlled current sources J = B * I',... %4 'nM - number of mutual inductive couplings',... %5 ['Zta_max - maximum absolute value of transfer resistance [Ohms]' char(10)... ' Ea + j.Er = (Zta + j.Ztr) (Ia + j.Ir)'],... %6 'Ztr_max - maximum absolute value of transfer reactance [Ohms]',... %7 ['Yta_max - maximum absolute value of transfer conductance [Siemens]' char(10)... ' Ja + j.Jr = (Yta + j.Ytr) (Ua + j.Ur)'],... %8 'Ytr_max - maximum absolute value of transfer susceptance [Siemens]',... %9 ['Aa_max - maximum absolute value of voltage gain active component' char(10)... ' Ea + j.Er = (Aa + j.Ar) (Ua + j.Ur)'],... %10 'Ar_max - maximum absolute value of voltage gain reactive component',... %11 ['Ba_max - maximum absolute value of current gain active component' char(10)... ' Ja + j.Jr = (Ba + j.Br) (Ia + j.Ir)'],... %12 'Br_max - maximum absolute value of current gain reactive component',... %13 'XM_max - maximum value of mutual inductive reactance [Ohms]'... %14 },... 'Selection of mutual inductive couplings and controlled source parameters');

22. 22 Circuit Topology C_nodes_twigs = GenerateTree(Ntwigs, mode) ShowTree(C_nodes_twigs, SetID, k) ShowGraphNet(C_nodes_twigs, C_nodes_chords, SetID, k) C_nodes_chords = GenerateCoTree(C_nodes_twigs, Nchords) C_twigs_chords = EssIncid(C_nodes_twigs, C_nodes_chords)

23. 23 Tree Generation 31 20 54 function C_nodes_twigs = GenerateTree(n, mode) C_nodes_twigs = zeros(n, n); rand('state',sum(100*clock)); r = rand(2,n); c = 2 * ( r(2, :) >= 0.5 ) - 1; m = 0; for k = 1:n s = ceil( (k-m)* r(1, k) + m-1 ); f = k; if s>0, C_nodes_twigs(s, k) = c(k); end C_nodes_twigs(f, k) = - c(k); if mode == 's', m = s; end end

24. 24 C_nodes_twigs = -1 0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 -1 -1 0 0 0 0 0 0 0 1 0 -1 0 0 0 0 0 0 -1 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 -1 1 0 0 0 0 0 0 0 -1 C_nodes_twigs = 1 -1 0 0 0 0 -1 0 0 1 0 0 -1 0 0 0 0 0 1 0 0 -1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 -1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 C_nodes_twigs = -1 0 1 0 1 0 0 0 0 -1 0 0 0 0 1 0 0 0 -1 1 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 -1 1 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 -1Examples

25. 25 Cotree Generation 1 2 10 3 4 5 6 7 8 9 11 12 Starts from the chosen tree Chords are introduced between nodes chosen randomly from the class of nodes with the lowest rank (lowest number of connected branches). This order assures the best connectivity of the circuit for a given number of chords.

26. 26 Examples Examples C_nodes_chords = 0 0 0 0 0 0 1 0 0 0 -1 0 0 0 -1 0 0 0 -1 0 0 0 0 1 0 0 1 0 0 0 0 -1 0 0 0 -1 -1 0 0 1 0 0 0 1 0 0 0 0 0 0 0 -1 0 0 0 0 1 0 0 0 1 0 0 0 -1 0 0 0 -1 0 0 0 0 -1 0 0 1 0 0 0 0 1 0 0 0 1 0 0 0 1 0 0 0 -1 0 0 C_nodes_twigs = -1 0 1 0 1 0 0 0 0 -1 0 0 0 0 1 0 0 0 -1 1 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 -1 1 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 -1 0 0 0 0 0 0 0 0 -1

27. 27 Tree Plot

28. 28 Circuit Parameter and Variable Generation Circuit Parameter and Variable Generation Chord currents Twig currents Twig voltages Chord voltages Chord emfs Chord currents Twig currents Twig voltages Chord voltages Chord emfs

29. 29 Global Circuit Variables Global Circuit Variables Concatenate the matrices for tree & cotree

30. 30 Current sources (change of independent voltage source emfs) Converting Voltage Sources to Current Sources Z Z E E new =E - Z J J Convert nJ voltage sources to current sources [E, J] = ConvertE2J_AC(E, Z, nJ);

31. 31 Controlled sources (change of independent voltage source emfs) Cross Parameters Mutual reactances [E, J, Zt, Yt, A, B, XM] = ControlledSources_AC(E, J, I, U, Z,... nControl, nEI, nJU, nEU, nJI, nM,... Zta_max, Ztr_max, Yta_max, Ytr_max, … Aa_max, Ar_max, Ba_max, Br_max, XM_max);

32. 32 Web Accessibility The system will be accessible on the INTERNET, to allow remote use, for both professors and students Partial examination of problems will be done on the computer, In a face-to-face or remote setting. The web accessibility is currently partially functional and partially under development

33. 33 PlatformPlatform Web server: Tomcat 4.1.29 - http://jakarta.apache.org/tomcathttp://jakarta.apache.org/tomcat DB server: MySQL 3.2x - http://www.mysql.com/http://www.mysql.com/ Scripts tool: Apache ANT - http://ant.apache.org/http://ant.apache.org/ Versioning server: CVS - http://www.cvshome.org/, http://www.wincvs.org/http://www.cvshome.org/

34. 34 ConclusionsConclusions A specialized e-learning system able to automatically generate large sets of circuit analysis problems, all with the same difficulty, but having different topological structures and parameters of the Circuits, has been designed, implemented and experimented. The problems are for use both during the tutorials and for examinations, thus -- despite the inherent risk for an engineer understanding of reality -- all parameters and variables describing the circuits should be integers to facilitate the computational task. Problems and solutions should be stored automatically on disk in distinct directories, with files referring to the same problem having related labels The system will be developed for making it accessible on the web

35. 35 COMMISSION OF THE EUROPEAN COMMUNITIES EDUCATION AND CULTURE DIRECTORATE - GENERAL SOCRATES - Minerva Transnational Projects in the field of Information and Communication Technology and Open and Distance Learning in Education This work has been partially supported by the Socrates Minerva Project 87574-CP-1-2000-1-RO-MINERVA- ODL Artificial Intelligence and Neural Network Tools for Innovative ODL (http://www.dsp.pub.ro/) This product does not necessarily represent the Commission's official position. Acknowledgment and disclaimer

36. 36 Vrije Universiteit Brussels, BE Prof. Jan Cornelis, Vice-Rector Prof. Edgard Nyssen, Prof. Rudi Deklerck Universität Erlangen-Nürenberg Prof. Manfred Kessler, Director Institute für Physiologie und Kardiologie Université de la Rochelle, FR Prof. Michel Eboueya, Assistant Director of Information and Industrial Imaging Lab. Universidade Nova de Lisboa, PT Prof. Adolfo Steiger Garcao, President of UNINOVA Prof. Jose Manuel Fonseca University of Edinburgh, UK Dr. Judy Hardy, Applications Consultant at EPCC Dr. Mario Antonioletti Patras University, GR Prof. Nicolas Pallikarakis, Coordinator of BioMedical Engineering Scool Res. Cristian Badea Equant Romania, RO Dr. Pavel Budiu, Strategy Manager Partners