For the course in Information and Communication Quality (prof. Di Blas) for the MS in Computer Engineering at Politecnico di Milano

1. COMO CAMPUS
Models and interaction
mechanisms for
exploratory interfaces
Luigi Spagnolo
luigi.spagnolo@polimi.it
1 Information and Communication Quality

2. Index
2
¨ PREVIEW: Online experimentation!
¨ Part I: navigation, search and exploration
¤ Break
¨ Part II: Faceted search: the model(s) and the
interaction
¨ Visualization issues will be covered into an other
lecture

3. 3 PREVIEW: Online Experimentation

4. Intro
4
¨ This lecture starts in a quite unusual way :-)
¨ To let you introduced with exploratory
interfaces you’ll take part to a research
experiment
¨ But don’t worry!
¤ It’s
not dangerous for your health :-)
¤ The questionnaire you’re asked to fill is
anonymous and the answers will not be graded

5. The application | 1
5

6. The application | 2
6
¨ The last version of a prototype built for the Italian Ministry of
Culture
¨ A map of exploring venues of archaeological interest in Italy
¨ According to three properties (facets):
¤ Kind of venue: museum, archaeological site and superintendence (a
local branch of the Ministry of Culture devoted to archeological
heritage management).
¤ Location: the venue location, at level of macro-area (Northern Italy,
Central Italy, eyc.), Italian region and Italian province.
¤ Civilization or Period: The ancient civilizations (Romans, Greeks, etc.)
or periods (e.g. Middle Ages, Bronze age) the venues are relevant to.

7. The application | 3
7
¨ The tag cloud:
¤ Tag size à the number of results that are relevant with respect to the period or
civilization in question.
¤ Text color à how much the percentage of results that are relevant for the period/
civilization deviates from an uniform distribution.
n Shades of green show a stronger positive correlation between the other selected filters (e.g.
the location and/or the venue type) and the civilization/period in question. Red instead shows
a negative correlation (the civilization/period is less significant with respect to other criteria
selected).
¤ Background color à w.r.t. the whole set of venues are relevant for the period/
civilization, which percentage of them are included in the results?
n Green shows a positive correlation, while red instead shows a negative correlation.
¤ E.g., for venues in a specific region only (e.g. Lombardy), a green tagindicates that
the given civilization was particularly relevant for that region.
¤ The green background shows instead that the civilization is peculiar of that region,
and is less likely to be found elsewhere.

8. The application | 4
8
¨ The map:
¤ At three levels: Italian region, Italian province, extact
location(s)
¤ The color of the circle à the specific type of venue
¤ The size of the circle à the number of items of that
type in that area

9. The experiment
9
¨ Go to http://tinyurl.com/exp-icq
¤ (or http://www.ellesseweb.com/mining/)
¨ You will find a page with two links:
1. The application
2. An online questionnarie (on Rational Survey)
¤ Keep both open on the browser
¨ Work individually (1 hour max)
¨ Answer with your opinions, without looking at other websites, just
at the ArchaeoItaly application
¤ Remember: the survey is anonymous, and there are no “correct
answers”!
¤ For any doubts, ask me!

10. 10 Part 1 | Navigation, search and exploration

11. Let’s start with a scenario
11
¨ Work in pairs
¨ Imagine to work as journalists for the
Horse Illustrated magazine
¨ You have to write an essay about
horses in art (and in particular in
painting) among the centuries.
¨ Find interesting information on the
website of the Louvre Museum
¤ http://www.louvre.fr/llv/commun/
home.jsp?bmLocale=en

12. Problems with the Louvre
12
¨ Artworks are separated by department (internal
“bureaucratic” classification) and by
provenience.
¨ It is not possible to search them together
(regardless of their age and country of origin)
by subject.
¨ There is no introductory content on the subject
that can guide the student in her search.

13. Content-intensive websites
13
¨ Also know as:
¤ Information-intensive
¤ Often Infosuasive = informative + persuasive
¤ Like ancient rhetoric: inform and persuade
¨ Mainly intended for:
¤ Learning, understanding, discovering, comparing
information
¤ Leisure and entertainment

14. Contents
14
¨ Text, multimedia (audio, video, images)
¨ Hypermedia = multimedia + hyperlinks
¨ Information involves subjective judgment
¤ Depends on the author and on the user
¤ Objective: “10km far from Como”, “the painting
was made in 1886”
¤ Subjective: “Near Como”, “the painting is
impressionist”

15. User experiences requirements | 1
15
¤ From the users’ point of view:
n Usability: usage is effective, efficient and satisfactory
n Findability: users can locate what they are looking for
n “At a glance” understandabity: users understand the
website coverage and can make sense of information
n Enticing explorability: users are compelled to “stay
and play” and discover interesting connections
among topics

16. User experiences requirements | 2
16
¤ From the stakeholders’ point of view:
n Planned
serendipity: promoting most important
contents so that users can stumble in them
n E.g. “Readers that purchased this book also bought…”
n Communication strengh and branding: the website
conveys the intended “message” and “brand” of the
institution behind it
n E.g. “we have the lowest prices”, “we are very
authorithative”, etc.

17. Information architecture
17
¤ Purpose: conceptually
organizing information
¤ Providing access to contents
n Index navigation (a)
n Guided navigation (b)
¤ Providing the possibility of moving
from a content to related ones
n Contextual navigation (c): cross-
reference links, semantic relationships

18. “Traditional” structure
18
¤ Taxonomy: hierarchy
of categories and
subcategories
n Sections and group of
contents are the
branches of the tree
n Contents are the leaves
¤ Cross-reference links
between nodes

19. An example
19
¤ Artworks of the month
Sitemap:
¤ Paintings
Top 10 masterpieces
Art n
n By artist
gallery n By artistic movement
website n By subject
¤ Sculptures
n ...
n By material
¤ Photographs
n ...

20. Problems/1
20
¨ What if I want to browse all artworks (regardless
their type) by artist?
¤ Classifications are “nested” in a fixed order
¤ Designers should choose which classification should
prevail (e.g. by type)
¨ What if I want to find “impressionists paintings
portraing animals”?
¤ I cannot combine multiple “sibling”classifications (e.g.
by style and by subject)

21. Problems/2
21
¨ As long as the website is small a good
taxonomy can satisfy user requirements
¨ For large websites
¤ (hundreds or thousand of pages)
¤ Indexed/guided navigation doesn’t scale
¤ Users can’t easily find what they want
¤ Users can’t make sense of all such information

22. Solutions?
22
¨ What do users do when navigation doesn’t work?
¤ They use search!
¤ Search arranges contents dynamically and automatically (in
a way not predefined by designers)
¨ But keyword-based search is not optimal
¤ No hints for users that have no clear idea of what looking
for
¤ Users must know how the information is described (e.e.
the specific jargon used)
¤ Just for retrieval/focalized search
¨ We need a better paradigm: Exploratory search

23. Exploratory search
23
¨ The model “query à results” is (too
much) simple
¨ Search is often like berry picking!
(Bates 1989)
¤ Users explore a corpus of contents
¤ They refine the query (again and
again) according to what they learn
¤ They pick information here and there,
piece by piece

24. From search to exploration
24
¨ From finding to
understanding
(Marchionini)
¤ Acquire knowledge
about a domain, its
jargon, the properties of
information items in it.
¤ Useful to (better)
understand what to look
for
¤ …but also to analyze a
dataset

25. Goals of exploratory applications
25
¨ Object seeking
¤ Identify the best object(s) whose features match user
requirements (e.g. purchasing a photocamera with concerns
regarding price, resolution, etc.)
¨ Knowledge seeking
¤ Expand the knowledge about a given topic and related
information (e.g. Leonardo Da Vinci and Italian Renaissance)
¨ Wisdom seeking
¤ Discover interesting relationships among features in a
information space/dateset (e.g. analysis of sales in Esselunga
chain stores, according to store location, type of article, price,
etc.)
¨ These goals can possibly coexist in the same application

26. Retrieval vs. exploration models
26
¨ Retrieval model: query + results
¤ Query can can be either:
n Free form (e.g. keyword based search)
n Structured (parametric search, e.g. Scholar advanced search)
n Guided (select data from a predefined set of choices)
¨ Exploration model:
¤ Query + results + refinements/feedback
¤ Query supported by self-adaptive structures for:
n Further filter results to a subset of them
n Summarizing the features shared by results

27. 27 Part 2 | Faceted search: model(s) and interaction
(Amazon’s Diamond search was one of the first e-commerce applications of faceted search)

28. Faceted search
28
¨ A exploratory search/navigation pattern based on
progressive filtering of results
¨ The user selects a combination of metadata values belonging
to several facets
¨ Each facet correspond to a particular dimension that
describes the content objects made available for search, e.g.
for an artwork:
¤ Subject: people portrayed, flowers and plants, abstract...
¤ Medium: painting, sculpture, photography...
¤ Technique: oil, watercolors, digital art...
¤ Style: impressionism, expressionism, abstractism...
¤ Location: Prado, Louvre, Guggenheim

29. Let’s see a pair of examples
29
¨ Two examples:
¤ http://orange.sims.berkeley.edu/cgi-
bin/flamenco.cgi/famuseum/Flamenco
¤ http://www.artistrising.com
¨ Try the same search we’ve
seen before: find horses in
art
¨ More examples at:
http://www.flickr.com/photos/
morville/collections/
72157603789246885/

30. Non just a matter of finding…
30
E.g. you can learn
that horses in art
are often found in
paintings
portraing soldiers
or warriors and
leaders

31. How the interaction works
31
¨ When the user chooses a
filter, the application
selects:
¤ The results: items that have
been “tagged” with the filter
and the other metadata
previously chosen
¤ The remaining filters:
metadata that combined
with the previous choices
can produce results
¨ The users can continue
narrowing results until they
options are available

32. A (generalized) formal model | 1
( terms
)
32
¨ Taxonomy: a pair T ,
¤ A set of concepts or T = {t1 ,t2 ,…,tn }
¤ The subsumption relation connecting narrower
terms (hyponyms) to broader concepts (hypernyms)
laptop  computer
location : 'Como'  location : 'Lombardy'  location : 'Italy'
¤ Terminal concepts: terms not further specialized
(the “leaves”)

33. A (generalized) formal model | 2
33
¨ For faceted taxonomies concepts are given
in terms of property-value pairs
(restrictions):
¤ E.g. subject: “horse”, location: “Como”
¨ A query is any of: q1 and q2
¤ A restriction q = property : value
q1 or q2
¤ A conjunction, disjunction or negation of
(sub)queries not q
¤ Actually there are limitations in the way concepts
can be combined in current facet browser
implementations

34. A (generalized) formal model | 3
34
¨ Item description: an information item o ∈O is
described as a conjunction of restrictions
d ( o ) = subject :"horse" and style :"Impressionism" and …
¨ Extension of a query: the set of items in a
context O that match the query ext ( q ) = {o ∈O | d ( o)  q} O
ext ( q1 and q2 ) ⊆ ext ( q1 ) , ext ( q2 ) ( )
tc  t p ⇒ ext ( tc ) ⊆ ext t p
ext ( q1 ) , ext ( q2 ) ⊆ ext ( q1 or q2 )
ext ( not q ) ≡ ext ( ALL)  ext ( q )

35. A (generalized) formal model | 4
35
¨ The result of a query is:
¤ Itsextension in the given information space extO ( q )
¤ The set of features shared by these results: i.e. all
the concepts that can be derived from the
descriptions of objects in extO ( q )

36. Query transformations
36
¨ Operations allowing to navigate from a state to
another of the exploratio
¤ Appending new restrictions to the query in conjunction
(zoom-in: from a wider to a narrower set of results)
¤ Adding alternatives in disjunction to the existent ones (zoom-
out: from a narrower to a wider set)
¤ Removing existing constraints (zoom-out again)
¤ Negating/excluding values
¤ Replacing a filter with another (shift)
¨ Implemented by hyperlinks (for conjunctive filters / shift),
check boxes (for disjunctions), etc.

37. How values are (usually) combined
37
¨ Filters belonging to different facets are combined in
conjunction
¤ E.g. “technique:oil” AND “style:impressionism”
¤ Filters belonging to the same facet are:
¤ Combined in conjunction if the facet admits more values at
the same time for each object
n E.g. “subject:people” AND “subject:animals”
n (both people and animals in the same picture)
¤ Combined in disjunction if the facet adimits only one value
n E.g. “location:Milan” OR “location:Como”
n (an object which is Como or in Milan)

38. Type of facets
38
¨ Single-valued (functional properties) vs. multi-valued
¨ Flat vs. hierarchical organization of values
¤ E.g. hierarchical: nation/region/province
¨ Subjective/arbitrary (properly named facets) vs. objective
(attributes)
¤ A date, a location, a price are examples of objective data
¤ “Topic”, “Audience”, “Artistic movement”, “importance” are
examples of subjective information
¤ Assigning/using a value involves some kind of judgment and
interpretation and is influenced by cultural and personal
backgrounds

39. Type of facet values
39
¨ Terms (strings of text) ¨ Sortable and comparable?
¤ Taxonomies, controlled ¤ We can say that
vocabularies value1<=value2<=…<=valueN?
¤ User-defined tags ¤ E.g. Dates, magnitudes, scales of
(folksonomies) judgment, quantitative data
n e.g. “sufficient”<“excellent”,
¤ From data-mining
10€<100€, “Monday”<“Friday”
¨ Numerical values and dates ¤ Ranges [value1, value2]
¨ Boolean values (yes/no) n E.g. User is allowed to search for
events from 01/06 to 31/08
¤ E.g. “Available for buying?”,
“original?”, “still living?” ¤ Classes of values
n e.g. for price: 0-10€, 11-20€,
¨ Even shades of color, 21-50€, 51-100€, …
shapes, etc... n The way we define classes is arbitrary
and depend on domain

40. Benefits of faceted search
40
¨ Easy and natural almost like “traditional” browsing
¨ With respect to keyword-based search users have hints
¤ Users can more easily make sense of information (if supported by
good interfaces)
¤ …and learn about the context by interacting with it
¨ Users can freely combine multiple classifications according to their
wishes
¤ In traditional browsing, when you reach a terminal concept you can’t
refine further
¤ With faceted search, you can continue refining with related concepts
¨ Navigation is safe: frustrating “no results found” searches avoided
¤ Only concepts that have been used to classify the current set of
results are diplayed

41. Limitations
41
¨ It works well only with structured data
¨ Faceted search does not provide a ranking of
results
¤ For “object seeking” tasks it might be a limitation
¤ It may be better to compute the “distance” with
respect to an “optimal” solution à otimization task
¨ Other limitations are discussed in the following
slides on advanced issues

42. 42 Advanced (research) issues

43. Full Boolean queries | 1
43
¨ How to achieve something like this?
“Given a budget of 250,000 euros,
I’m interested in a flat with at least 4
rooms and not central heating in the
centre, or an house with at least 5
rooms in the suburbs”

44. Full Boolean queries | 2
44
¨ Foci (Ferré et al.) the set of sub-expressions in the semantic
tree of the query
¨ ( )
A query is a pair q,φ , where q is an arbitrary combination of
filters and φ is one of its foci
¤ The focus is used to select the subquery at which the new filter
should be appended (or the transformation should be applied)
¤ …But also to “inspect” different points of view of information
¤ The main focus represents the “whole” query

45. Semantic faceted search
45
¨ We can filter items, but how can we filter facet values?
¤ E.g. paintings filtered by artists
¤ But how we filter the Artists facet values by nationality,
gender, age, etc.?
¨ Exploring contents at level of sets using semantic
relationships, e.g.
¤ The museums that have bronze Greek statues
¤ “Women portrayed by women”: paintings with subject:woman
and artist:gender:female
¤ Schools attended by the daughters of U.S. democratic
presidents (http://www.freebase.com/labs/parallax/)
¤ Challenges: effective models and usable interface
¨ An example: Sewelis

46. Beyond binary classication | 1
¤ Classification (faceted or not) is usually
binary:
¤ An item must be either relevant (1) or
not relevant (0) to a certain category
¤ Problem: quite arbitrary decision in
many real domains

47. Beyond binary classication | 2
î How to classify acathedral by architectural
style?
¤ Built upon a 6th century buliding
¤ Mainly gothic
¤ 17th century (baroque) towers
¤ Rebuilt during neoclassicism
¤ Decorations added in 19th century
¤ Contains Roman forum marbles (donated by Pius
IX)
¤ …
î Do we tag the cathedral with all or only some of
these?
î A classification may be correct for a kind of users but
ineffective for another one

48. Beyond binary classication | 3
î Monna Lisa is a well known
portait of a woman, but…
î There is also a landscape in
the background
î Do we classifity it as
“subject: woman” and
“subject: Tuscan landscape”
too?

49. Beyond binary classication | 4
î Onion is very
used in French
cuisine
î How do we
distinguish
“onion-based”
recipes from all
the recipes with
onion inside?

50. Beyond binary classication | 5
¨ A possible solution:
associating weights
to each triple item-
facet-value
¤ A statement about
the statement
¨ Values between 0 and
1 or other scales
¨ Query could be
specified in terms of
facet-values pairs and
ranges of weights

51. Beyond binary classication | 6
¨ Subjective weights
¤ Relevance: at which
extent the item can be
considered as belonging
to a certain facet value
¤ Significance: the relative
importance of the item
according to a facet value
¨ Objective weights
¤ E.g. Concentration or quantity (e.g.
a thing is made for the 10% of
material:bronze)
¤ E.g. for exploring venues:
distance from points of interests

52. Beyond binary classication | 7
¨ Interaction
(concepts)

53. Handling information overload
53
¨ Too more facets and facets values may
generate information overload too!
¤ Possible solution: Display only the most relevant
facets (and facet values) for the user profile or the
given context
¨ How to determine the most “interesting” facets in
a given context?
¤ E.g. those with a less “uniform” distribution of
values (more correlation)
¤ We will discuss this in a next lecture… :-)

54. Interested in MS Theses? Contact us! :-)
54
¨ Advisors: Prof. Di Blas, Prof. Paolini
¨ Both theoretical and development
¨ Fuzzy facets
¨ Semantic faceted search
¨ Advanced visualizations
¨ …
¨ Your own ideas! :-)

55. 55
Any final questions?
Are you still alive/awake?
Thank you for your attention!