The Smart University is the vision of the university as a platform that acquires and delivers foundational data to drive the analysis and improvement of the teaching & learning environment. Sensor-data, linked (open) data, and formalised teaching knowledge are the three sources that we are tapping. In this talk I presented first results of our efforts.
Low-cost motion detection and other sensors coupled with low-cost credit-card sized computers such as the Raspberry Pi open up opportunities to equip rooms with sensors. As the Raspberry Pi is a full-fledged computing device not only can one acquire data, but also process it in context.
Case-based reasoning is a problem-solving approach that allows to capture and re-use experience. With our toolset around the myCBR Workbench we started formalising teaching experience in Applied Sound Engineering and gold ore pretreatment knowledge to support students in their individual learning situations.
You can learn more about the Smart University here: http://smartuniversity.uwl.ac.uk
UiPath Community: Communication Mining from Zero to Hero
Smart University - The university as a platform
1. Centre for
Intelligent
Computing
Smart University
The university as a platform
1
3 May 2013 ● Prof Thomas Roth-Berghofer ● Robert Gordon University, Aberdeen
Centre for
Model-based
software
engineering
Dr Samia
Oussena
Saturday, 4 May 13
2. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Outline of my talk
2
Saturday, 4 May 13
3. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
University of West
London, Ealing
3
Saturday, 4 May 13
4. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
University of West
London, Ealing
3
Saturday, 4 May 13
5. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
University of West London:
Structure & Collaborations
School of Computing and Technology (SOCAT)
Saturday, 4 May 13
6. School of Psychology, Social Work and Social Sciences
Ealing Law School
London College of Music
Ealing School of Art, Design and Media
International Business School
University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
University of West London:
Structure & Collaborations
School of Computing and Technology (SOCAT)
School of Nursing, Midwifery and Healthcare
London School of Hospitality and Tourism
Saturday, 4 May 13
7. School of Psychology, Social Work and Social Sciences
Ealing Law School
London College of Music
Ealing School of Art, Design and Media
International Business School
University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
University of West London:
Structure & Collaborations
E.g., course Project management
E.g., PhD “Interactive 3d portraits”
E.g., project “Audio mixing support”
School of Computing and Technology (SOCAT)
School of Nursing, Midwifery and Healthcare
London School of Hospitality and Tourism
Saturday, 4 May 13
8. Intelligent computing
Prof Thomas Roth-Berghofer
University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 5
Research groups
Internationalisation
and user experience
Dr José Abdelnour-Nocera
Networks and
distributed systems
Prof Peter Komisarczuk
Information
management and libraries
Dr Stephen Roberts
Model-based
software engineering
Dr Samia Oussena
Civil and built environment
Dr Ali Bahadori-Jahromi
& Dr Charlie Fu
Mobile Computing
Dr John Moore
School of Computing and Technology (SOCAT)
Saturday, 4 May 13
9. Intelligent computing
Prof Thomas Roth-Berghofer
University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 6
Thomas Roth-Berghofer
Head of Research and
Research Training
N
in
o
A
u
ricch
io
Senior
Lecturer
in
Applied
Sound
Engineering
S
am
P
roctor
Lecturer
in
Applied
Sound
Engineering
Research
team
C
h
ristian
S
au
er
Research
Assistant,
PhD
student
D
an
H
u
gh
es-M
cG
rail
PhD
student
Saturday, 4 May 13
10. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Smart University
7
Deliver foundational data to drive the analysis
of the teaching & learning environment.
http://smartuniversity.uwl.ac.uk
Saturday, 4 May 13
11. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Smart University
Using
sensor data
Linked (open) data
expert knowledge/experience
7
Deliver foundational data to drive the analysis
of the teaching & learning environment.
http://smartuniversity.uwl.ac.uk
Saturday, 4 May 13
12. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 8
Daily journey of student Daily journey of lecturer
Sensor
data
Humidity
Temperature
Noise
Linked
(Open) DataExperience
Improving
classroom experience
Saturday, 4 May 13
13. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Room
utilisation
9
Saturday, 4 May 13
14. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 10
Saturday, 4 May 13
15. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Occu-Pi: First test run
11
Saturday, 4 May 13
16. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 12
Saturday, 4 May 13
17. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 12
Saturday, 4 May 13
18. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR
Workbench
13
In co-operation with the Competence
Centre Case-Based Reasoning (CCCBR) at
the German Research Centre for Artificial
Intelligence (DFKI), Kaiserslautern
Saturday, 4 May 13
19. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Modelling perspective
14
Saturday, 4 May 13
20. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Modelling perspective
14
Projects,
classes and attributes
Projects & Classes
Saturday, 4 May 13
21. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Modelling perspective
14
Projects,
classes and attributes
Projects & Classes
List of
similarity measures
associated with
selected attribute
Similarity Measures
Saturday, 4 May 13
22. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Modelling perspective
14
Editors
Editing area for
attributes, global and local similarity measures
Projects,
classes and attributes
Projects & Classes
List of
similarity measures
associated with
selected attribute
Similarity Measures
Saturday, 4 May 13
23. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Case bases perspective
15
Saturday, 4 May 13
24. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Case bases perspective
15
Projects,
classes and attributes
Projects & Classes
Saturday, 4 May 13
25. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Case bases perspective
15
Projects,
classes and attributes
Projects & Classes
Lists of
case bases and
available instances
Case bases and instances
Saturday, 4 May 13
26. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
Case bases perspective
15
Projects,
classes and attributes
Projects & Classes
Lists of
case bases and
available instances
Case bases and instances
Editors
Editing area for
case bases and cases
Saturday, 4 May 13
27. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Filling
knowledge
containers
16
Case base
Vocabulary
Similarity
measures
Adaptation
knowledge
Vocabulary
Vocabulary
Saturday, 4 May 13
28. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Filling vocabulary
and case base
17
Vocabulary
Case base
Saturday, 4 May 13
29. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
CSV file importer
Automatic set up of initial vocabulary, i.e.
attributes, data types, and ranges
Setup of default similarity measures
based on data types
Filling the case base
18
A.Aamodt and E. Plaza. CBR: Foundational issues, methodological variations,
and system approaches. AI Communications, 7(1):39–59, 1994.
Saturday, 4 May 13
30. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
myCBR Workbench:
CSV file importer
Automatic set up of initial vocabulary, i.e.
attributes, data types, and ranges
Setup of default similarity measures
based on data types
Filling the case base
18
A.Aamodt and E. Plaza. CBR: Foundational issues, methodological variations,
and system approaches. AI Communications, 7(1):39–59, 1994.
Start cycle of
refine & test
Saturday, 4 May 13
31. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Saturday, 4 May 13
32. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
Saturday, 4 May 13
33. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
Saturday, 4 May 13
34. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
Saturday, 4 May 13
35. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
Set data types2
Saturday, 4 May 13
36. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
Set data types2
Saturday, 4 May 13
37. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
Boolean
Concept
Float
Integer
String
Symbol
+
Set data types2
Saturday, 4 May 13
38. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
...
Set data types2
Saturday, 4 May 13
39. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
...
myCBR Workbench:
CSV Importer
19
Colhead1 Colhead2 Colhead3 …
R1C1 R1C2 R1C3 …
R2C1 R2C2 R2C3 …
… … … …
Generate
attributes1
...
Set data types2 Generate instances3
Saturday, 4 May 13
40. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Improving similarity
measures
20
Vocabulary Case base
Saturday, 4 May 13
41. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Improving similarity
measures
20
Vocabulary
Similarity
measures
Case base
Saturday, 4 May 13
42. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Knowledge extraction
workbench (KEWo)
21
Christian S. Sauer and Thomas Roth-Berghofer. Web community knowledge extraction for myCBR 3. In M.
Bramer, M. Petridis, and L. Nolle, eds., Proc. of AI-2011. Springer, 2011.
Saturday, 4 May 13
43. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Knowledge extraction
workbench (KEWo)
21
Christian S. Sauer and Thomas Roth-Berghofer. Web community knowledge extraction for myCBR 3. In M.
Bramer, M. Petridis, and L. Nolle, eds., Proc. of AI-2011. Springer, 2011.
Saturday, 4 May 13
44. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
http://www.jboss.org/drools/
Adaptation knowledge
22
Vocabulary
Similarity
measures
Case base
Future
w
ork
Saturday, 4 May 13
45. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
http://www.jboss.org/drools/
Adaptation knowledge
22
Vocabulary
Similarity
measures
Case base
Adaptation
knowledge
For example:
Integration of Drools -
The Business Logic
integration Platform
Future
w
ork
Saturday, 4 May 13
46. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Audio advisor
23
Saturday, 4 May 13
47. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 24
Saturday, 4 May 13
48. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 25
Mid
Bass
Treble
Translation
Saturday, 4 May 13
49. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 26
Case structure
Saturday, 4 May 13
50. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Similarity measures
27
Saturday, 4 May 13
51. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 28
ANNIE
Supporting case acquisition
and query formulation
Saturday, 4 May 13
52. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 29
Saturday, 4 May 13
53. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 30
Saturday, 4 May 13
54. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Auric
Advisor
31
Lotta
R
in
tala
PhD
student,
Aalto
U
niversity,
Finland
Saturday, 4 May 13
55. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Selection of pretreatment
method for refractory gold ore
32
Gold extraction from its ores may require a
combination of mineral processing and
metallurgical processes to be performed on the
ore.
There are two types of metallurgical processes:
Hydrometallurgical Pyrometallurgical
- leaching - smelting
- low temperature - high temperature
http://www.unige.ch/sciences/terre/mineral/fontbote/teaching/lehne_oredressing/2_callion_ore.jpg
Saturday, 4 May 13
56. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
General process chain
for refractory gold ore
33
Pretreatment
Crushing & Grinding
Leaching
Recovery process
Refining
Ore
Refined gold
Saturday, 4 May 13
57. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
General process chain
for refractory gold ore
33
Pretreatment
Crushing & Grinding
Leaching
Recovery process
Refining
Ore
Refined gold
For example:
• low-pressure
oxidation
• high-pressure
acidic oxidation
• high-pressure non-
acidic oxidation
• Nitric acid oxidation
• Chlorine oxidation
• Biological oxidation
• Pyrometallurgical
oxidation
Saturday, 4 May 13
58. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Basic idea: Two step
recommendation process
34
Saturday, 4 May 13
59. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Basic idea: Two step
recommendation process
34
Identify
context1
Saturday, 4 May 13
60. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Basic idea: Two step
recommendation process
34
Identify
context1
2
Identify
pretreatment
Saturday, 4 May 13
61. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Two case bases
35
Problem Solution
Whole process chain of the mining operation
Mining situation
Description
[Ore/Mineral/Deposit]
Process step Process stepProcess step
Problem Solution
Oxidative Treatment [And
its conditions and raw
material description]
Cyanide Leaching [Next
best suited process step]
Saturday, 4 May 13
62. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Similarity measures: Example
36
Saturday, 4 May 13
63. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 37
Auric Advisor
Saturday, 4 May 13
64. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing 37
Auric Advisor
Saturday, 4 May 13
65. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Auric Advisor: Goals
38
Advice experts
Provide starting points to process design
Validate existing and newly designed process chains
Teach students of hydrometallurgy (or new employees)
best practices or process steps in specific situations
Saturday, 4 May 13
66. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
CIC research strands:
Open PhD research topics
Using sensor data to improve classroom experience
Investigate and develop novel techniques, methodologies and support tools
for using sensor data to improve the classroom experience.
Experience-based audio mastering and mixing
Formalise teaching experience to improve the individual learning experience
in audio engineering.
Acquisition and use of sensor data for music student
teaching
Formalise teaching experience with the help of sensor technology to
improve the individual learning experience.
Agent-based acquisition of explanation knowledge
Acquire and use distributed explanation knowledge in the SEASALTexp
environment.
39
Saturday, 4 May 13
67. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
More information
40
Blog: http://smartuniversity.uwl.ac.uk
Workshop at CONTEXT 2013:
http://smartuni2013.workshop.hm
Submission deadline: 12 July 2013
Saturday, 4 May 13
69. University of West LondonSlideSmart University • RGU, AberdeenCentre for Intelligent Computing
Used images
42
http://commons.wikimedia.org/wiki/File:Waksman_laboratory.jpg
http://commons.wikimedia.org/wiki/File:Gold-130327.jpg
http://commons.wikimedia.org/wiki/File:Blue_Drop.svg
http://commons.wikimedia.org/wiki/File:FireV2.png
Saturday, 4 May 13