Human Factors of XR: Using Human Factors to Design XR Systems
Orpailleur -- triclustering talk
1. Dmitry V. Gnatyshak, Dmitry I. Ignatov*,
Sergei O. Kuznetsov
School of Applied Mathematics and Information Science & Intelligence Systems and Structural
Analysis Lab
NRU Higher School of Economics, Moscow, Russia
LORIA Orpailleur meeting, Nancy, France, 2013
2. Outline
1. Motivation and problem setting
2. FCA basic definitions
3. Triclustering methods
4. Experiments
5. Conclusion
2
3. Motivation
A large amount of structured and unstructured data
generates triadic data.
Example: folksonomy is a set of triples (user, object, tag)
Examples:
Bibsonomy.org
(user, bookmark, tag)
Social networking sites
(user, group, interest)
Delicious
(user, link, tag)
3
4. Main goals
1. Comparison of some triclustering methods
2. Development of a toolbox for triclustering experiments
3. New possibly better methods
4. Possible applications
4
5. FCA: basic definitions
Biology Mathematics Computer
Science
Chemistry
Kate x x
Mike x x x
Alex x x
Pete x x x
5
(R. Wille, 1982; B. Ganter, R. Wille, 1999)
16. Experiments
Main goals:
Fault-tolerance test
Comparison by criteria: time, quantity, mean density,
coverage and diversity
For TriBox and OAC-triclustering we implemented their
parallel versions
They were included to the comparison
16
22. Method Time Quantity Average
density
Coverage Diversity Efficiency of
parallel
version
OAC(box)
average large
low
high~very low
very low~average
high
OAC (prime)
small large average high~average average~high low
SpecTric
Small for small
contexts
small low average~high 1 –
TriBox high average high high high high
TRIAS
very large 1 high~low high~low –
22
Results (time, quantity, average density,
coverage, diversity)
23. Conclusion
There is no a winner according to the comparison criteria
Method TriBox shows best results but it takes huge
computational time
OAC-triclustering based on prime operators gives the
second best results and it is sufficiently fast
23
24. Conclusion
There is no a winner according to the comparison criteria
Details by methods:
TRIAS
High elapsed time
Too large number of small well-interpreted triclusters
(triconcepts)
24
25. Conclusion
OAC (box operators)
Large triclusters of low density
High density, small diversity
An efficient parallelization
OAC (prime-operators)
High speed of computations
Large number of dense well-interpreted triclusters
Low efficiency of parallelization
25
26. Conclusion
Spectral Triclustering
High computational speed on small contexts
Well-interpreted triclusters but of the low density
Diversity is always equals to 1, but it causes too low coverage
TriBox
A moderate number of well-interpreted triclusters
High elapsed time
Efficient parallelization
Reasonably high coverage and diversity
26