UNIT-6: Information Search and Visualization

• Information overload and anxiety common

• Developing more powerful search and visualization methods, integration

of technology with task

• Terms:

– Information gathering

– Seeking

– Filtering

– Visualization

• Huge volumes of available data:

– Data mining

– Data warehouses and data marts

– Knowledge networks or semantic webs

– A know-item-search versus making sense and discovering

• Traditional interfaces have been difficult for novice users

– Complex commands

– Boolean operators

– Unwieldy concepts

• Traditional interfaces have been inadequate for expert users

– Difficulty in repeating searches across multiple databases

– Weak methods for discovering where to narrow broad searches

– Poor integration with other tools

• Designers are just learning how to present large amounts of data in

orderly and user-controlled ways
• OAI (Objects / Actions Interface)

• Customizable search options and displays using control panels

• Structured relational database

– contains relations and a schema to describe the relations

– relations have records

– records have fields, and fields have values Textual document

libraries

– set of items (10 to 100,000)

• Multimedia document librairies

– Contains images, sound, video, animations, etc

– Digital archives are more loosely organized

– Directories contain metadata

• Task Actions (fact-finding)

– Browsing and Searching

• Scrolling

• Zooming

• Joining

• Linking

– Specific fact finding

– Extended fact finding

– Open-ended browsing

– Exploration of availability

• Where to Search

– Finding aides
 • Table of contents, Indexes, Description introductions,
Subject classification, Key-Word-In-Context (KWIC)

– Preview and overview surrogates

Searching in textual documents and database querying

• World Wide Web search engines have greatly improved their

performance by using statistical ranks and the information in the web’s
hyperlink structure

• Searching in structured relational database systems well established task

using SQL language

• Users write queries that specify matches on attribute levels

• Example of SQL command

– SELECT DOCUMENT#

– FROM JOURNAL-DB

– WHERE (Date >= and Date<= 1998)

– and (Language = English or French)

– and (publisher = ASIST or HFES or ACM).

• SQL has powerful features, but it requires 2 to 20 hours training

• While SQL is a standard form-fillin queries have simplified query

formulation

• Finding a way not to overwhelm novice users is a challenge

• Evidence shows that users perform better and have higher satisfaction
when they can view and control the search

• Framework to coordinate design practice:

– Formulation

• source of the information

 • fields for limiting the source

• phrases

• variants

• size of results

– Initiation of action

• explicit or implicit

• most systems have a search button for explicit initiation, or

for delayed or regularly scheduled initiation

• implicit actions are initiated by changes to a parameter

– Review of results

• view overview and previews

• manipulate visualizations

• examine selected items

– Refinement

• should provide meaningful messages to explain search

outcomes

• should support progressive refinement

– Use

• allow queries, parameters, or results to be saved, used, or

shared

Multimedia document searches

• Searches for databases and textual documents are good, but multimedia
searches are in a primitive stage

• Current multimedia searches require descriptive documents or metadata

searches
• Search by date, text captions, or media is possible

• Useful to have computers perform some filtering

• New systems will incorporate powerful annotation and indexing, with

better search algorithms and browsing

• Image Search:

• Finding photos with images such as the Statue of Liberty is a

challenge

• Query-by-Image-Content (QBIC) is difficult

• Search by profile (shape of lady), distinctive features (torch),

colors (green copper)

• Use simple drawing tools to build templates or profiles to search

with

• More success is attainable by searching restricted collections

• Search a vase collection

• Find a vase with a long neck by drawing a profile of it

• Critical searches such as fingerprint matching requires a minimum

of 20 distinct features

• For small collections of personal photos effective browsing and

lightweight annotation are important

• Map Search

• On-line maps are plentiful

• Search by latitude/longitude is the structured-database solution

• Today's maps are allow utilizing structured aspects and multiple

layers

• City, state, and site searches

• Flight information searches

 • Weather information searches

• Example: www.mapquest.com

• Mobile devices can allow “here” as a point of reference

• Design/Diagram Searches

• Some computer-assisted design packages support search of

designs

• Allows searches of diagrams, blueprints, newspapers, etc.

• E.g. search for a red circle in a blue square or a piston in an

engine

• Document-structure recognition for searching newspapers

• Sound Search

• MIR supports audio input

• Search for phone conversations may be possible in future on

speaker independent basis

• Video Search

• Provide an overview

• Segmentation into scenes and frames

• Support multiple search methods

• Infomedia project

• Animation Search

• Prevalence increased with the popularity of Flash

• Possible to search for specific animations like a spinning globe

• Search for moving text on a black background

Advanced filtering and search interfaces

For advanced uses there are alternatives to form fillin query interfaces:

• Filtering with complex boolean queries

– Problem with informal English, e.g. use of ‘and’ and ‘or’

– Venn diagrams, decision tables, and metaphor of water flowing

have not worked for complex queries

• Dynamic Queries - Adjusting sliders, buttons, etc and getting immediate

feedback

– “direct-manipulation” queries

– Use sliders and other related controls to adjust the query

– Get immediate (less than 100 msec) feedback with data

– Dynamic HomeFinder and Blue Nile

– Hard to update fast with large databases

• Query previews present an overview to give users information and the

distribution of data and thereby eliminate undesired items

• Faceted metadata search

– Integrates category browsing with keyword searching

– Flameco

• Collaborative Filtering

– Groups of users combine evaluations to help in finding items in a

large database

– User "votes" and his info is used for rating the item of interest

– E.g. a user rating sex restaurants highly is given a list of restaurants

also rated highly by those who agree the six are good
 • Multilingual searches

– Current systems provide rudimentary translation searches

– Prototypes of systems with specific dictionaries and more

sophisticated translation

• Visual searches

– Specialized visual representations of the possible values

– E.g. dates on a calendar or seats on a plane

– On a map the location may be more important than the name

– Implicit initiation and immediate feedback

Information visualization

• "A picture is worth a thousand words!"

• Large amounts of information in compact and user-controlled ways

– example: USA map, click a city to see more info

• Information visualization can be defined as the use of interactive visual

representations of abstract data to amplify cognition

• Scientific visualization

– continuous variables, volumes and surfaces

• Information visualization

– categorical variables and the discovery of patterns, trends,

clusters, outliers, and gaps

• Visual data mining

• Answer questions users didn’t know they had

• Tufte offers advice for static information, but dynamic displays present a
challenge

• Must be more than cool

• The Visual Information Seeking Mantra

– Overview first

– zoom and filter

– then details-on-demand

• Basic data types

– 1 - Dimensional

• Linear data types include textual documents, program

source code, lists of names in sequential order

• E.g. highlight lines of code that have changed

– 2 - Dimensional

• Planar or map data includes geographic maps, floor plans,

newspaper layouts

• E.g. Geographic Information Systems, spatial displays of

document collections

• Example tasks: find regions containing items

• Basic data types (cont.)

– 3 - Dimensional

• Real-world objects such as molecules, the human body,

buildings

• Users must cope with understanding their position and

orientation when viewing the objects

• E.g. overviews, landmarks, stereo displays, transparency,

color coding

• Virtual Reality displays

• Users’ tasks typically deal with continuous variables

 • National Library of Medicine's Visible Human Project

• Controversial

– Multi-Dimensional

• Most relational and statistical databases

• N attributes become points in an n-dimensional space

• Interface representation could be a 2-D scattergram with

each additional dimension controlled by a slider

• Parallel coordinate plots

• Table Lens

• Hierarchal or k-means clustering

– Temporal

• Time Lines are widely used and accepted

• Items have a start and finish time and items may overlap

• Tasks include finding all events before, after, or during some

time period

– Tree

• Collections of items with each item having a link to one

parent item (except root)

• Outline style of indented labels or node-and-link diagram

• Space-filling approach

– Networks

• Sometimes data needs to be linked to an arbitrary number

of other items

• Example: A graphical representation of the World Wide Web

• Mode-and-link diagrams, matrices

• Basic tasks

– Overview

• Gain an overview of the entire collection

• Adjoining detail view

• The overview might contain a movable field-of-view box to

control the contents of the detail view

• allowing zoom factors of 3 to 30

• Fisheye view

– Zoom

• Zoom in on items of interest

• Allows a more detailed view

• Need to maintain context

• Particularly important for small displays

– Filter

• Filter out uninteresting items

Allows user to reduce size of search

– Details-on-Demand

• Select an item or group and get details when needed

• Useful to pinpoint a good item

• Usually click on an item and review details in a separate or

pop-up window

– Relate

• View relationships among items

 • Use human perceptual ability – proximity, containment,
connected line, color coding

• Example: Set directors name, and view all movies with that
director

– History

• Keep a history to allow undo, replay, and progressive

refinement

• Allows a mistake to be undone, or a series of steps to be

replayed

– Extract

• Extract the items or data

• Save to file, print, or drag to another application

Challenges for information visualization

– Import data

– Combine visual representations and textual labels

– See related information

– View large volumes of data

– Integrate data mining

– Collaborate with others

– Achieve universal usability