Transcript InfoVis Main Tasks

Seminario de Tecnologías
de Información Emergentes
Agosto-Diciembre, 2007
Sesión # 5
The Information Search Process
Recognize an information need
Define and understand needs
Choose a search system
Formulate a query
Execute search
Examine results
Extract information
Reflect / Iterate / Stop
(Marchionini, 1997)
Information Retrieval Process
Recognize an information need
Define and understand needs
Choose a search system
Formulate a query
Execute search
Examine results
Extract information
Reflect / Iterate / Stop
(Marchionini, 1997)
Information Visualization
(InfoVis)
Recognize an information need
Define and understand needs
Choose a search system
Formulate a query
Execute search
Examine results
Extract information
Reflect / Iterate / Stop
(Marchionini, 1997)
Definitions
Scientific visualization:
Involves continuous variables,
volumes, and surfaces
Information visualization:
Involves more categorical
variables, and the discovery of
patterns, trends, clusters, outliers,
and gaps in large data sets.
(Shneiderman & Plaisant, 2005)
Definitions
Information visualization:
Involves compact graphical
presentations and user interfaces
to interactively manipulate large
data sets
Allows users asnwers questions
they did not they had
Also know as visual data mining
(Shneiderman & Plaisant, 2005)
Definitions
Information visualization
assumes that:
Users scan, recognize, and recall
images rapidly
Users can detect subtle changes
in size, color, shape, movement or
texture
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
Overview
Zoom
Filter
Details on demand
Relate
History
Extract
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
Overview
Users should be able to obtain
a general view of the entire
collection
Features:
Fisheye view
Dynamic filed-of-view box
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
Zoom
Users should be able to look
at details of items of interest
Features:
Zoom focus (content, context)
Zoom factor (position, shape)
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
Filter
Users should be able to filter
items of interest.
Features:
Eliminating unwanted items
Sliders, bottons, menus
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
Provide details on demand
Users should be able to get
details from an item or group of
items of interest
Features:
Browsing, pop-up windows
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
Relate
Users should be able to relate
items or groups within the
collection
Features:
Visual displays by proximity, by
shape, or by color coding
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
History
Users should be able to keep a
history of actions to support
undo, replay, and progressive
refinement
Features:
Search terms
Search sequences
(Shneiderman & Plaisant, 2005)
InfoVis Main Tasks
Extract
Users should be able to
extract, manipulate and save
the item or set of items
obtained from the collection
Features:
MyDocument Folders
E-mail
(Shneiderman & Plaisant, 2005)
InfoVis Demos
TextArc
http://textarc.org/
Chat circles
http://alumni.media.mit.edu/~fviegas/p
rojects/chatcircles/index.htm
InfoVis Demos
Exploring enron
http://jheer.org/enron/
eArchivarius
http://www.isi.edu/news/print.php?story
=56
InfoVis Demos
PhotoMesa
http://www.cs.umd.edu/hcil/photomesa/
ICDL
http://www.icdlbooks.org/
InfoVis Applications
Sentiment analysis software
http://www.corporasoftware.com/prod
ucts/sentiment.aspx
Blog analysis software
http://www.umbrialistens.com/product
s/technology.php
InfoVis Applications
Visualization methods & approaches
http://www.visualliteracy.org/periodic_table/periodic_ta
ble.html
Visual Information Retrieval Interfaces
http://www.hipertext.net/english/pag100
7.htm
Visual IR Interfaces
Collection (overview)
Self-organizing maps (related terms)
VxInsight (islands)
Cat-a-cone (hierarchical categories)
BEAD (3D documents)
SPIRE (stars or landscapes)
JAIR (conceptual maps)
NIRVE (clustering)
PRISE (clustering)
Perspective Wall (2D, 3 views)
VIBE (search state, 2D & 3D)
Visual IR Interfaces
Collection (overview)
Document attributes
TileBars
E-Archivarius
Chat circles
Other?
Top InfoVis Research Issues
Importing data
Combining visual representations
with textual levels
Looking at related information
Viewing large data sets
Integrating data mining
Collaboration among users
Universal usability
(Shneiderman & Plaisant, 2005)
Top InfoVis Research Issues
Personal information collections
Abstract information spaces
Platform-independent, more
sophisticated user interfaces
New information access
paradigms
(Hearst, 1999)
Top InfoVis Research Issues
Representation of information
Handling diverse information
formats
Improving communication of the
system with its users (HCI)
(hipertext.net)
Próxima sesión
Invitado especial de VITI
Lecturas por confirmar