Marti Hearst
March 15, 2002

• Why Text is Tough
• Visualizing Concept Spaces
– Clusters
– Category Hierarchies
• Visualizing Query Specifications
• Visualizing Retrieval Results
• Usability Study Meta-Analysis

• To help with Information Retrieval
– give an overview of a collection
– show user what aspects of their interests are present in a collection
– help user understand why documents retrieved as a result of a query
• Text Data Mining
– Mainly clustering & nodes-and-links
• Software Engineering
– not really text, but has some similar properties

• Text is not pre-attentive
• Text consists of abstract concepts
– which are difficult to visualize
• Text represents similar concepts in many different ways
– space ship, flying saucer, UFO, figment of imagination
• Text has very high dimensionality
– Tens or hundreds of thousands of features
– Many subsets can be combined together

The Dog.

The Dog.
The dog cavorts.
The dog cavorted.

The man.
The man walks.

The man walks the cavorting dog.
So far, we can sort of show this in pictures.

As the man walks the cavorting dog, thoughts arrive unbidden of the previous spring, so unlike this one, in which walking was marching and dogs were baleful sentinals outside unjust halls.
How do we visualize this?

• Abstract concepts are difficult to visualize
• Combinations of abstract concepts are even more difficult to visualize
– time
– shades of meaning
– social and psychological concepts
– causal relationships

• Language only hints at meaning
• Most meaning of text lies within our minds and common understanding
– “How much is that doggy in the window?”
• how much: social system of barter and trade (not the size of the dog)
• “doggy” implies childlike, plaintive, probably cannot do the purchasing on their own
• “in the window” implies behind a store window, not really inside a window, requires notion of window shopping

• General categories have no standard ordering
(nominal data)
• Categorization of documents by single topics misses important distinctions
• Consider an article about
– NAFTA
– The effects of NAFTA on truck manufacture
– The effects of NAFTA on productivity of truck manufacture in the neighboring cities of El Paso and Juarez

• Other issues about language
– ambiguous (many different meanings for the same words and phrases)
– different combinations imply different meanings

• Text is highly redundant
– When you have lots of it
– Pretty much any simple technique can pull out phrases that seem to characterize a document
• Instant summary:
– Extract the most frequent words from a text
– Remove the most common English words

478 said
233 god
201 father
187 land
181 jacob
160 son
157 joseph
134 abraham
121 earth
119 man
118 behold
113 years
104 wife
101 name
94 pharaoh

• How can we show an overview of the contents of a text collection?
– Show info external to the docs
• e.g., date, author, source, number of inlinks
• does not show what they are about
– Show the meanings or topics in the docs
• a list of titles
• results of clustering words or documents
• organize according to categories (next time)

– Scatter/Gather
• show main themes as groups of text summaries
– Scatter Plots
• show docs as points; closeness indicates nearness in cluster space
• show main themes of docs as visual clumps or mountains
– Kohonen Feature maps
• show main themes as adjacent polygons
– BEAD
• show main themes as links within a forcedirected placement network

• Two main steps
– cluster the documents according to the words they have in common
– map the cluster representation onto a
(interactive) 2D or 3D representation

• Finds overall similarities among groups of documents
• Finds overall similarities among groups of tokens
• Picks out some themes, ignores others

Encyclopedia text
8 symbols
68 film, tv (p)
97 astrophysics
67 astronomy(p)
10 flora/fauna
14 sports
47 film, tv
7 music
12 steller phenomena
49 galaxies, stars
29 constellations
7 miscelleneous
Clustering and re-clustering is entirely automated

Cutting, Pedersen, Tukey & Karger 92, 93, Hearst & Pedersen 95
• How it works
– Cluster sets of documents into general “themes”, like a table of contents
– Display the contents of the clusters by showing topical terms and typical titles
– User chooses subsets of the clusters and re-clusters the documents within
– Resulting new groups have different “themes”
• Originally used to give collection overview
• Evidence suggests more appropriate for displaying retrieval results in context
• Appearing (sort-of) in commercial systems

Northern Light Web Search: Started out with clustering. Then integrated with categories. Now does not do web search and uses only categories.

An example layout produced by Bead, seen in overview, of 831 bibliography entries. The dimensionality
(the number of unique words in the set) is 6925.
A search for ‘cscw or collaborative’ shows the pattern of occurrences coloured dark blue, mostly to the right.
The central rectangle is the visualizer’s motion control.

Themescapes (Wise et al. 95)
(Wise et al. 95)

• Since text has tens of thousands of features
– the mapping to 2D loses a tremendous amount of information
– only very coarse themes are detected

Galaxy of News
Rennison 95

Galaxy of News
Rennison 95

(594 docs)

• H. Chen, A. Houston, R. Sewell, and B.
Schatz, JASIS 49(7)
• Comparison: Kohonen Map and Yahoo
• Task:
– “Window shop” for interesting home page
– Repeat with other interface
• Results:
– Starting with map could repeat in Yahoo (8/11)
– Starting with Yahoo unable to repeat in map
(2/14)

Three studies find negative results

• Kleiboemer, Lazear, and Pedersen. Tailoring a retrieval system for naive users. In Proc. of the 5th Annual Symposium on Document Analysis and Information Retrieval, 1996
• This study compared
– a system with 2D graphical clusters
– a system with 3D graphical clusters
– a system that shows textual clusters
• Novice users
• Only textual clusters were helpful (and they were difficult to use well)

• H. Chen, A. Houston, R. Sewell, and B. Schatz, JASIS 49(7)
• Comparison: Kohonen Map and Yahoo
• Task:
– “Window shop” for interesting home page
– Repeat with other interface
• Results:
– Starting with map could repeat in Yahoo (8/11)
– Starting with Yahoo unable to repeat in map
(2/14)

• Participants liked:
– Correspondence of region size to # documents
– Overview (but also wanted zoom)
– Ease of jumping from one topic to another
– Multiple routes to topics
– Use of category and subcategory labels

• Participants wanted:
– hierarchical organization
– other ordering of concepts (alphabetical)
– integration of browsing and search
– correspondence of color to meaning
– more meaningful labels
– labels at same level of abstraction
– fit more labels in the given space
– combined keyword and category search
– multiple category assignment
(sports+entertain)

• NIRVE Interface by Cugini et al. 96. Each rectangle is a cluster. Larger clusters closer to the “pole”. Similar clusters near one another. Opening a cluster causes a projection that shows the titles.

• Visualization of search results: a comparative evaluation of text, 2D, and 3D interfaces Sebrechts, Cugini, Laskowski, Vasilakis and Miller,
Proceedings of SIGIR 99, Berkeley, CA, 1999.
• This study compared
:
– 3D graphical clusters
– 2D graphical clusters
– textual clusters
• 15 participants, between-subject design
• Tasks
– Locate a particular document
– Locate and mark a particular document
– Locate a previously marked document
– Locate all clusters that discuss some topic
– List more frequently represented topics

• Results (time to locate targets)
– Text clusters fastest
– 2D next
– 3D last
– With practice (6 sessions) 2D neared text results; 3D still slower
– Computer experts were just as fast with 3D
• Certain tasks equally fast with 2D & text
– Find particular cluster
– Find an already-marked document
• But anything involving text (e.g., find title) much faster with text.
– Spatial location rotated, so users lost context
• Helpful viz features
– Color coding (helped text too)
– Relative vertical locations

• Huge 2D maps may be inappropriate focus for information retrieval
– cannot see what the documents are about
– space is difficult to browse for IR purposes
– (tough to visualize abstract concepts)
• Perhaps more suited for pattern discovery and gist-like overviews

• Has been around since the 50’s.
(Small, Garfield,
White & McCain)
• Used to identify core sets of
– authors, journals, articles for particular fields
– Not for general search
• Main Idea:
– Find pairs of papers that cite third papers
– Look for commonalitieis
• A nice demonstration by Eugene Garfield at:
– http://165.123.33.33/eugene_garfield/papers/mapsciworld.html

(From Garfield 98)

(From Garfield 98)

(From Garfield 98)

Let’s show categories instead of clusters

(Pratt, Hearst, & Fagan 99)

(Pratt 97)
• Decide on important question types in an advance
– What are the adverse effects of drug D?
– What is the prognosis for treatment T?
• Make use of MeSH categories
• Retain only those types of categories known to be useful for this type of query.

• Design
– Three queries
– 24 cancer patients
– Compared three interfaces
• ranked list, clusters, categories
• Results
– Participants strongly preferred categories
– Participants found more answers using categories
– Participants took same amount of time with all three interfaces

• HiBrowse Problem:
– Search is not integrated with browsing of categories
– Only see the subset of categories selected
(and the corresponding number of documents)

(Furnas & Zacks ’94)

Multiple Simultaneous Categories
• Key Ideas:
– Separate documents from category labels
– Show both simultaneously
• Link the two for iterative feedback
• Distinguish between:
– Searching for Documents vs.
– Searching for Categories

• Catacomb:
(definition 2b, online Websters)
“A complex set of interrelated things”
• Makes use of earlier PARC work on
3D+animation:
Rooms
IV: Cone Tree
Web Book
Henderson and Card 86
Robertson, Card, Mackinlay 93
Card, Robertson, York 96

browse
Category
Hierarchy search query terms
Collection
Retrieved Documents

• Browse/explore category hierarchy
– by search on label names
– by growing/shrinking subtrees
– by spinning subtrees
• Affordances
– learn meaning via ancestors, siblings
– disambiguate meanings
– all cats simultaneously viewable

– Categories on Page (Retrieved Document) linked to Categories in Tree
– Flipping through Book Pages causes some
Subtrees to Expand and Contract
– Most Subtrees remain unchanged
– Book can be Stored for later Re-Use

 Integrate category selection with viewing of categories
 Show all categories + context
 Show relationship of retrieved documents to the category structure

But … do users understand and like the 3D?

• Basic idea similar to Cat-a-Cone
• But use familiar HTML interaction to achieve similar goals
• Usability results are very strong for users who care about the collection.

• command attribute value connector …
– find pa shneiderman and tw user#
• What are the attribute names?
• What are the command names?
• What are allowable values?

Form-Based Query Specification (Altavista)

Form-Based Query Specification (Melvyl)

Form-based Query Specification (Infoseek)

(Young & Shneiderman 93)

• Visualizations of Query Term Distribution
– KWIC, TileBars, SeeSoft
• Visualizing Shared Subsets of Query Terms
– InfoCrystal, VIBE, Lattice Views
• Table of Contents as Context
– Superbook, Cha-Cha, DynaCat
• Organizing Results with Tables
– Envision, SenseMaker
• Using Hyperlinks
– WebCutter

• Interfaces should
– give hints about the roles terms play in the collection
– give hints about what will happen if various terms are combined
– show explicitly why documents are retrieved in response to the query
– summarize compactly the subset of interest

• An old standard, ignored until recently by internet search engines
– used in some intranet engines, e.g., Cha-Cha

Goal: minimize time/effort for deciding which documents to examine in detail
Idea: show the roles of the query terms in the retrieved documents, making use of document structure

 Graphical Representation of Term
Distribution and Overlap
 Simultaneously Indicate:
– relative document length
– query term frequencies
– query term distributions
– query term overlap

Query terms:
DBMS (Database Systems)
Reliability
What roles do they play in retrieved documents?
Mainly about both DBMS
& reliability
Mainly about DBMS, discusses reliability
Mainly about, say, banking, with a subtopic discussion on
DBMS/Reliability
Mainly about high-tech layoffs

– Variation in gray scale saturation imposes a universal, perceptual order (Bertin et al. ‘83)
– Varying shades of gray show varying quantities better than color (Tufte ‘83)
– Differences in shading should align with the values being presented (Kosslyn et al. ‘83)

• Conjunct of disjuncts
• Each disjunct is a concept
– osteoporosis, bone loss
– prevention, cure
– research, Mayo clinic, study
• User does not have to specify which are main topics, which are subtopics
• Ranking algorithm gives higher weight to overlap of topics
– This kind of query works better at highprecision queries than similarity search
(Hearst 95)

 Preliminary User Studies
 users understand them
 find them helpful in some situations, but probably slower than just reading titles
 sometimes terms need to be disambiguated

SeeSoft: Showing Text Content using a linear representation and brushing and linking (Eick & Wills 95)

Show which subsets of query terms occur in which subsets of documents occurs in which subsets of retrieved documents

Show how often each query term occurs in retrieved documents
– VIBE (Korfhage ‘91)
– InfoCrystal (Spoerri ‘94)
– Problems:
• can’t see overlap of terms within docs
• quantities not represented graphically
• more than 4 terms hard to handle
• no help in selecting terms to begin with

(Spoerri 94)

(Olson et al. 93, Korfhage 93)

– Problems:
• can’t see overlap of terms within docs
• quantities not represented graphically
• more than 4 terms hard to handle
• no help in selecting terms to begin with

(Cousins 97)
• Supporting the Information Seeking
– UI to a digital library
• Direct manipulation interface
• Workcenter approach
– experts create workcenters
– lots of tools for one task
– contents persistent

(Cousins 97)
• Drag and Drop interface
• Reify queries, sources, retrieval results
• Animation to keep track of activity
Slide by Shankar Raman

(Chen & Yu ’00)
• Goal
– Find invariant underlying relations suggested collectively by empirical findings from many different studies
• Procedure
– Examine the literature of empirical infoviz studies
• 35 studies between 1991 and 2000
• 27 focused on information retrieval tasks
• But due to wide differences in the conduct of the studies and the reporting of statistics, could use only 6 studies

(Chen & Yu ’00)
• Conclusions:
– IR Infoviz studies not reported in a standard format
– Individual cognitive differences had the largest effect
• Especially on accuracy
• Somewhat on efficiency
– Holding cognitive abilities constant, users did better with simpler visual-spatial interfaces
– The combined effect of visualization is not statistically significant
– Misc
• Tilebars and Scatter/Gather are well-known enough to not require citations!!