Suitability of Searching and Representing Multimedia Learning Resources in 3D Virtual Gaming...
-
Upload
junior-knight -
Category
Documents
-
view
214 -
download
2
Transcript of Suitability of Searching and Representing Multimedia Learning Resources in 3D Virtual Gaming...
Suitability of Searching and Representing Multimedia Learning Resources in 3D Virtual Gaming Environment
Presented byAbu Saleh Md. Mahfujur Rahman
SupervisorProfessor Dr. Abdulmotaleb El Saddik
Multimedia Communications Research Laboratory (MCRLab)University of Ottawahttp://www.mcrlab.uottawa.ca
Outline
Motivation and Related Study Proposed System 3D Gaming Environment Implementation Result Analysis Conclusion and Future Work
Motivation and Related Study
Information visualization is the use of computer supported, interactive, visual representations of abstract data to amplify cognition.
The traditional systems: Poor way of Information visualization Less intuitive navigation space Lack of entertainment in learning
Problem statement: Suitability of visualizing learning resources from a set of available search repositories using metaphors in a structured [13] way, which improves the perception of the learners as well as offers entertainment with education.
Motivation and Related Study (cont.) Our approach:
3D virtual environment: attractive large display space as well as natural and cognitive aspects.
Using gaming metaphor: provide entertainment in navigating the environment
Organized representation of data: allows the users to directly interact with the data, get insight, and come up with new hypotheses.
a) Car metaphor
b) Spaceship metaphor
c) Library metaphor
Proposed System
Data Model LO in Repositories LO in P2P LO in Web
Graphical Model Data
Representation Model
VE Model 3D GUI Control
Figure 2. Schematic diagram of the proposed system
System Architecture
Data Model
LOM
Rep
osito
ries
LO in
P2P
VE
Mod
el
Dat
a Rep
. M
odel
Graphical Model
3D G
UI
Con
trol
Sear
chin
g
Pers
onal
izat
ion
3D Gaming Environment
Clu
ster
ing
Nav
igat
ion
Inte
ract
ion
Ran
dom
izat
ion
layer 1 layer 2
Web
Sea
rch
Res
ults
Proposed System (cont.)
Data Model: Learning Objects Communities using different elements and
vocabularies create difficulties to share and reuse resources
Learning Object: "any entity, digital or non-digital, that may be used for learning, education or training” (IEEE LTSC)
LOM is defined by IEEE LTSC for identifying, searching, and retrieving LOs
Data Model: Access Paradigm LOR contains LOs or
their references along with LOM
All elements are optional
Can be difficult to know which elements to use
Figure 3. Learning Object access paradigm
Proposed System (cont.)
LOs Repository
Shared Metadata Interface
WWW
LOM Repository
RESULTS
pdfjpeg HTML
mov
Proposed System (cont.)
Graphical Model Data Representation
Model
Ideogram: content type specification
Learning Object title
[ title ]
a) Polygonal billboard skeleton b) Rendered traffic model
Figure 4. 3D Traffic Sign Template for
representation of LOs (e.g. traffic sign for html)
TABLE IModel template representations of Learning Object
Model SelectionCollision Detection
Number of PolygonSelection Precision
Memory Requirement
Resultant Frame Rate/sec.
Billboard Easy Less (2 triangles) Imprecise Less Optimal
3D polygonal Difficult More (100+ triangles) Precise More Average
Polygonal billboard Easy Moderate (15 triangles) Precise Less Optimal
Proposed System (cont.)
Graphical Model VE Model
Avatar Terrain Reference Models
3D Gaming Environment
Searching Personalization Clustering and 3D Mapping Navigation Interaction Randomization
Yahoo, Google Web-services
P2PDirectory
Remote Repository
Query Interface
Search Result Clustering
Semantic Query Translation
Mapping to 3D Spatial Layout
laye
r 3
laye
r 2
laye
r 1
3D Virtual Traffic Arch.
Figure 5. Layered search model
3D Gaming Environment (cont.)
Clustering and 3D Mapping Display relevant and grouped search results in
the 3D road layout so that navigation and information selection become easy.
Based on the searched keywords the prototype applies heuristic methods to create relevance metrics to construct groups from the search results.
3D Gaming Environment (cont.) Clustering and 3D Mapping
Algorithm SEARCH_RESULT_CLUSTERING Progressively look-up the
search result to find relevant keywords based on the initial search keyword
Add new keywords to the keyword set and apply equation 2 to calculate distances between each pair to form keyword array matrix
Quick-Sort the matrix values based on the distances in ascending order
Apply Minimum Spanning Tree (MST) algorithm over the matrix
End SEARCH_RESULT_CLUSTERING
a) MST tree b) Road architecture to the MST tree
T
CT1
CT3
CT2
CT4
CT5
T
CT1 CT2
CT5CT4CT3
Figure 6. Semantic keyword metrics MST tree and corresponding road architecture
3D Gaming Environment (cont.)
Personalization Enables users to personalize their searching,
navigational and learning experience within the 3D world.
Maintains individual user profile that includes user’s demographic information, age limit, area of expertise, topics of interest, search history, among others.
Individual profile info are taken into consideration while searching LOs and delivering the targeted 3D virtual environment to the user.
Allows to select the 3D shape of the learning objects and also the virtual avatar
A user is able to save any visited object and put it in his/her favorite list.
3D Gaming Environment (cont.)
Navigation Collision detection is used to
help the user stay within the search layout
Appropriate ideograms, landmark, and reference models are provided.
A preview of the 3D landscape is used as a navigational map in order to show the navigating user his/her point of presence in the 3D environment.
Furthermore, a direction indicator that always points to a reference position prevents the user from getting lost in the 3D world.
Direction indicator
Point of Presence of the Avatar Car
Traffic Group Layout
Figure 7. Navigational layout of the 3D metaphor
3D Gaming Environment (cont.)
Interaction User interacts with the
displayed search results for Viewing the associated
metadata description Playing/previewing the LO Saving the search history
in the local repository for later review
The user interaction can be performed through Keyboard Mouse Joystick Microphone or any
combination of these devices.
Virtual Environment Graphical Components
Event Processing Engine
Avatar / Car LO 3D Models
Global Event Listener
Joystick Mouse Keyboard
Global Event Interface APIs
User Interaction Devices / Processes
Microphone
Virtual Environment Graphical Components
GUI Controls
Figure 8. High level user interaction model
Implementation
Software Architecture
Client Side Components Server Side Components
3D Graphics Engine
Search Interface
Metadata Extractor
User Mananger
3D Template Repository
LOM-XML Parser
LOM Clustering Module Remote
Repository
Web-Search: Yahoo/Google
Local SQL LORs
Profile Repository
P2P DirectoryList
Figure 9. Software architecture of EYE&WHY prototype
Implementation
Sample Interface: dynamic traffic layout
Figure 10. A typical 3D search interface
Group interaction menu. Comprising options viz:a) Search related b) Save allc) Share alld) Make parent
Group sign
Dynamic road for each group of search results
Dynamic navigation layout
Implementation (cont)
Sample interface: static traffic layout
Figure 11. A typical 3D search interface
Static grid based (3x3) road layout
Mouse based navigational layout
Group signs are located (based on the algorithm) on the start of the roads
Implementation (cont)
Searching/sharing learning objects
Figure 11. A typical 3D search interface
Advanced search options
Sharing LO paramers
Implementation (cont)
Searching/sharing learning objects
Figure 11. A typical 3D search interface
Implementation (cont)
Searching/sharing learning objects
Figure 11. A typical 3D search interface
Search Result Clustering:A) KeywordB) User tagC) Term frequency
Remote Repository
Web-Search: Yahoo/Google
P2P DirectoryList
[ title ]
[ title ]
(a)
(b)
(c)
(e)
(d)
Multi-modal Interactions
Visualization metaphors Information organization Information repository
Test Result Analysis
Quantitative Measurements
Figure 11. Mapped search results vs. frame rate/sec
Figure 12. Mapped search results vs. total triangle count
Test Result Analysis (cont)
Quantitative Measurements
Figure 13. Search time comparisons of different search
services
Figure 14. Search result grouping and mapping time in
3D VE
Test Result Analysis (cont)
Qualitative Measurements
Figure 15. User response percentage on the evaluation of the 3D metaphor
TABLE II User satisfaction on the overall evaluation
Ease of Use Information Clustering Visual Quality Navigation Search Response
Mean 4.489 3.834 3.79 4.184 4.497
Std. Dev. 0.5 0.9 1.0 0.6 0.3
Test Result Analysis (cont)
Qualitative Measurements
Figure 16. User satisfaction based on 3D application/gaming experience
Figure 17. Comparison between responses of users from two different
age groups.
Conclusion and future-work
We presented a new way of visualizing search results in a 3D virtual gaming environment and measured the suitability of such approach. Gaming by default has its own attractions
to the user especially to the younger users as it provides intuitiveness in the search and exploration operations that are performed while navigating the 3D world.
We evaluated our prototype system in terms of quantitative and qualitative measurements aspects.
The spatial landscape of a 3D environment, in order to represent more information at a time, is proved to be useful.
Use of enhanced user control, e.g. HMD, Data Gloves, Haptics, etc.
Finding new metaphors to enhance user experience
Applying rules and heuristics to add scoring system to achieve the game like experience
Collaboration between users in a single searching environment
Conclusion and future-work (cont.)
1. E. Duval and w. Hodgins, "a lom research agenda," proc. 12th int’l conf. World wide web (www2003), may 2003.
2. IEEE learning object metadata, final draft standard, jul. 15, 2002 (ieee1484.12.1); http://ltsc.Ieee.Org/wg12/files/lom_1484_12_1_v1_final_draft.Pdf, accessed sept. 05, 2005.
3. Edusourcecanada, canadian network of learning object repositories; http://www.Edusource.Ca/english/objects_eng.Html, accessed oct. 5, 2005.
4. Ariadne - http://www.Ariadne-eu.Org, accessed jan 17, 2005.5. W. Cellary, w. Wiza, and k. Walczak, "visualizing web search results in 3d,"
computer, vol. 37, no. 5, pp. 87-89, may 2004.6. W. Wiza, k. Walczak, and w. Cellary, "periscope - a system for adaptive 3d
visualization of search results," proc. 9th int’l conf. 3d web technology (web3d'04), pp. 29-40, apr. 2004.
7. J. Klerkx, e. Duval, and m. Meire, "using information visualization for accessing learning object repositories," proc. 8th int'l conf. Information visualization (iv'04), pp. 465-470, july 2004.
8. Lightweight java game library, http://lwjgl.Org/, accessed sept. 20, 2005.9. P. Chiu et al., “Mediametro: browsing multimedia document collections with a 3d
city metaphor,” proc. 13th annual acm international conference on multimedia, hilton, singapore, pp. 213–214, nov. 2005.
10. C. Russo dos santos, p. Gros. "Multiple views in 3d metaphoric information visualization," 6th international conference on information visualization (iv'02), pp. 468-473, 2002.
11. Wordnet - a lexical database for the english language, http://wordnet.Princeton.Edu/, accessed jan. 27, 2006.
12. Jxta – open protocols for peer to peer network, http://www.Jxta.Org/, accessed jan 20, 2006.
13. Mukherjea, S., Foley, J., and Hudson, S., (1995):“Visualizing Complex Hypermedia Networks through Multiple Hierarchical Views”. Proceedings of ACM SIGCHI '95
References
متشکرم