a novel visualization model for web search results an application of the solar system metaphor tien...
TRANSCRIPT
A Novel Visualization Model for Web Search Results
A Novel Visualization Model for Web Search Results
An Application of the Solar System MetaphorAn Application of the Solar System Metaphor
Tien N. Nguyen and Jin ZhangTien N. Nguyen and Jin Zhang
Electrical and Computer Engineering DepartmentElectrical and Computer Engineering Department
Iowa State UniversityIowa State University
USAUSA
2
MotivationMotivation
WWW becomes an important source of WWW becomes an important source of information.information.
Internet search engines provide tremendous Internet search engines provide tremendous help in information seeking.help in information seeking.
Search engines present relevant Web pages Search engines present relevant Web pages in a list with titles and short paragraph in a list with titles and short paragraph descriptions extracted from the texts.descriptions extracted from the texts.
Browsing the Browsing the ranked listranked list can become rather can become rather tedious and unproductive.tedious and unproductive.
3
MotivationMotivation
The order of keywords used in a query also The order of keywords used in a query also affects the final outcomes of a search.affects the final outcomes of a search.• ““information system”, “system information”information system”, “system information”
A word or a phrase can have multiple A word or a phrase can have multiple meanings depending on different contexts.meanings depending on different contexts.• ““information visualization”information visualization”
Contextual informationContextual information or or topics of interesttopics of interest can only be introduced as a different query.can only be introduced as a different query.• ““software visualization”software visualization”
4
Web Search Result VizWeb Search Result Viz
Must provide an effective guide in searching Must provide an effective guide in searching for needed informationfor needed information• Controllable and adaptableControllable and adaptable
Abstraction of our visualization problemAbstraction of our visualization problem• The queryThe query• The search resultsThe search results• Context information Context information oror topics of interest topics of interest• Semantic relationsSemantic relations
• Degree of relevanceDegree of relevance
5
Solar System as MetaphorSolar System as Metaphor
All All planetsplanets and and asteroidsasteroids rotate around the rotate around the sunsun in the universe. in the universe.
Each of them has its own Each of them has its own orbitorbit and a and a moving moving speedspeed..
They are attracted to each other by They are attracted to each other by gravitygravity.. It is gravity that determines the orbit and It is gravity that determines the orbit and
moving speed of an asteroid.moving speed of an asteroid.
6
Solar System as MetaphorSolar System as Metaphor
7
Solar System as MetaphorSolar System as Metaphor
(Query)
(Page)
(Subject)•LocationLocation•DistanceDistance•ColorColor•SpeedSpeed•GravityGravity
8
GeneralizationGeneralization
ApplicationsApplications•WebWebsearch search •Paper Paper citationcitationVisualizationVisualization•HyperlinkHyperlinkstructurestructure
9
WebSearchViz SpaceWebSearchViz Space
10
Parameter ComputationParameter Computation
RR is the radius. is the radius.SimSimijij is the semantic strength between two is the semantic strength between two
entities (e.g. between a page and a query, or entities (e.g. between a page and a query, or between a page and a subject). It is between between a page and a subject). It is between 0 and 1.0 and 1.
Context icons are anchored on the Context icons are anchored on the circumference. Objects are inside the circle.circumference. Objects are inside the circle.
The distance of an objectThe distance of an object• ddii = R x (1 – Sim= R x (1 – Simi,centeri,center))• Objects with high Sims are close to center.Objects with high Sims are close to center.
11
WebSearchViz SpaceWebSearchViz Space
(k, R+cons)
12
Parameter ComputationParameter Computation
The angle of an object against X-axis:The angle of an object against X-axis:
If If SimSim1i1i + … + + … + SimSimmimi <> 0 <> 0
If an object that is irrelevant to any displayed If an object that is irrelevant to any displayed contexts, contexts, ii is defined as zero. is defined as zero.
• That is, it is projected onto the X-axis.That is, it is projected onto the X-axis.
mii
mimiii SimSim
SimSimSim
...
...
1
2211
13
Rotations and AnalysesRotations and Analyses
Activate a context icon to rotate evenly Activate a context icon to rotate evenly around the circumference.around the circumference.
Assume that the moving speed of a specified Assume that the moving speed of a specified context is context is degrees per second; and its degrees per second; and its angle is angle is rr with initial value of with initial value of 00..
The dynamic angle of an object The dynamic angle of an object ii is is
m
kki
ri
m
rkkkik
i
Sim
SimtSim
1
0,1
)()(
14
Rotations and AnalysesRotations and Analyses
The rotation speed of an object The rotation speed of an object ii is is
• All affected objects (i.e. objects that are All affected objects (i.e. objects that are somewhat related to the rotating context) somewhat related to the rotating context) will move evenly in their orbits.will move evenly in their orbits.
• The moving speed is affected primarily by The moving speed is affected primarily by SimSimriri, the degree of relevance between , the degree of relevance between
object object ii and context and context rr..
m
kki
rii
Sim
Sim
dt
d
1
15
Rotations and AnalysesRotations and Analyses
If SimIf Simriri = 0 (no relevance between object = 0 (no relevance between object ii
and context and context rr), the object stays put.), the object stays put.
If an object is only relevant to the moving If an object is only relevant to the moving context and is not to others, it has the same context and is not to others, it has the same speed as the moving context:speed as the moving context:
0
1
m
kki
rii
Sim
Sim
dt
d
ri
rim
kki
rii
Sim
Sim
Sim
Sim
dt
d
0
1
16
Rotations and AnalysesRotations and Analyses
In this case, the object icon is always located In this case, the object icon is always located on the line connecting the moving context on the line connecting the moving context and the center.and the center.
m
kki
ri
m
rkkkik
i
Sim
SimtSim
1
0,1
)()(
rri
rir
Sim
Sim
0
0
17
Project ambiguityProject ambiguity
Project ambiguity: many objects in a high-Project ambiguity: many objects in a high-dimensional space are projected into low-dimensional space are projected into low-dimensional space.dimensional space.
Project ambiguity and a large number of Project ambiguity and a large number of objects can cause visual occlusion.objects can cause visual occlusion.
Two objects are projected onto the same Two objects are projected onto the same points if their Sim value is 1.points if their Sim value is 1.
Overlapping icons can be disambiguated via Overlapping icons can be disambiguated via changing the positions of contexts.changing the positions of contexts.
18
Visual OcclusionVisual Occlusion
To handle overlapping in the case of many To handle overlapping in the case of many objects:objects:• Automatic clustering/grouping functionalityAutomatic clustering/grouping functionality• Zoom in and zoom outZoom in and zoom out• For Web search results: We provide focus For Web search results: We provide focus
page shifting functionalitypage shifting functionality
• A page can become the A page can become the newnew center center
• Surrounding objects are relevant pages.Surrounding objects are relevant pages.
19
WebSearchViz systemWebSearchViz system
(Query)
(Page)
(Subject)•LocationLocation•DistanceDistance•ColorColor•SpeedSpeed
20
WebSearchViz architectureWebSearchViz architecture
Search
Engine
A query Web search
results
Lexical
Analysis
Visualization
Mechanism
Vector-based
representation
Vector-based
representation
Subjects of interest
Display
21
Vector-based RepresentationVector-based RepresentationKeyword 1Keyword 1 Keyword Keyword jj Keyword pKeyword p
Page 1Page 1 0.1224 0.0593
Page 2Page 2 0.0349 0.0671
Page Page ii WWijij
……
Page nPage n
•WWijij is computed via a modified version of the is computed via a modified version of the
TF-IDF algorithmTF-IDF algorithm• Each Web page is represented by a row (a vector).Each Web page is represented by a row (a vector).•Similarity of two pages (Similarity of two pages (SimSim) is computed via a ) is computed via a combination of cosine and Euclidean distance values.combination of cosine and Euclidean distance values.•A query or a subject is represented as a vector.A query or a subject is represented as a vector.
22
Subject DefinitionSubject Definition
23
WebSearchViz SystemWebSearchViz System
http://www.eng.iastate.edu/~tien/WebSearchViz
24
ConclusionsConclusions
Solar System provides a nice metaphor for Solar System provides a nice metaphor for the novel visualization modelthe novel visualization model
The use of The use of movementmovement and and speed speed adds an adds an extra dimension into a visual space, extra dimension into a visual space, illustrating the semantic connections among illustrating the semantic connections among objects.objects.
Web Search result visualizationWeb Search result visualization• Accommodate the changes of preferences Accommodate the changes of preferences
and subjects of interestand subjects of interest
25
Thank youThank you
Tien N. NguyenTien N. Nguyen
Electrical and Computer Engineering DepartmentElectrical and Computer Engineering Department
Iowa State UniversityIowa State University
[email protected]@iastate.edu