ece 627 intelligent web: ontology and beyond

1
ece 627 intelligent web ontology and beyond

• lecture 18 tagging and folksonomy

2
metadata

• is
• structured information that describes, explains,
  locates, or otherwise makes it easier to
  retrieve, use, or manage an information resource
• (NISO)
• it allows systems to collocate related
  information, and helps users find relevant
  information

3
metadataways of creation

• generally in two ways
• professional creation (professionals working with
  complex, detailed rule sets and vocabularies)
• author creation (authors of documents provide
  metadata along with their creations)
• are ontologies the result of that???

4
metadatathird (new) way

• user-?created metadata
• users of the documents and media create metadata
  for their own individual use that is also shared
  throughout a community

5
folksonomywhat is it?

• it is a people's taxonomy
• is composed of terms in a flat namespace
• there is no hierarchy, no parent-?child or
  sibling relationships between these terms

6
folksonomywhat is it?

• the set of terms (called tags) that a group of
  users tagged content with, they are not a
  predetermined set of classification terms or
  labels

7
folksonomy

• the cumulative force of all the individual tags
  can produce a bottom-up, self-organized system
  for classifying items on the web

8
what is tagging?introduction

• a tag is a non-hierarchical keyword or term
  assigned to a piece of information - such as an
  internet bookmark, digital image, or computer
  file
• (Wikipedia)
• tagging to mark with a tag to label, identify,
  or recognize with or as if with a tag
• - a unique and powerful way of organizing
  information

9
what is tagging?tagging system

• three components
• users
• resources
• tags

10
tagging systemusers

• the people who employ a tagging system (sometimes
  also called taggers) they create the tags, and
  sometimes they add resources
• have a variety of different interests, needs,
  goals, and motivations but they are trying to
  achieve some larger goal such as sharing a
  photo or labeling a document so they can find it
  later

11
tagging systemresources

• are items that users tag
• a resource can be just about anything a book, a
  Web page, a video, or even a location
• within each tagging system, resources often share
  some common properties they are books, or
  photos, or

12
tagging systemtags

• the keywords added by users are tags
• can be just about any kind of term, they can be
  descriptions of the resources subject matter,
  its location, its intended user, a reminder, or
  something else entirely can be individual words
  or phrases
• tags are essentially metadata about the resource

13
tagging systemtags

• tags are more than just metadata in an
  application they are a tool people use to
  track, share, and find information

14
tagging systemtag cloud

• is a method of presenting tags where the more
  frequently used tags are emphesized

15
tagging systemtag cloud example

• www.wordle.net

16
tagging system

• all tagging happens in the context of a system,
  and the system defines what kind of tagging can
  take place
• for example, the system may allow users to add
  their own resources or not, may allow to tag any
  resource or not, may forbid certain kinds of tags

17
tagging systemperspectives

• tagging sits
• at the intersection
• of three established
• fields

18
tagging systeminformation architecture

• the structural design of shared information
  environments
• and
• the art and science of organizing and labeling
  web sites, intranet, online communities, and
  software to support usability and findability
• information architects focus on using controlled
  vocabularies, search-and-browse systems

19
tagging systemsocial software

• applications that people use to communicate,
  collaborate, and share online
• people who design social software are interested
  in facilitating group interaction within the
  system

20
tagging systempersonal information management

• refers to the practice and study of the
  activities people perform in order to acquire,
  organize, maintain, retrieve, and use information
  items such as documents, web pages, e-mail
  messages
• (Wikipedia)
• they are programs for managing information and
  methods for keeping yourself on track help you
  file, track, and find your information when you
  need it

21
tagging systemtensions

• personal lt-gt social
• do people tag primary for their own benefit?
• or are they motivated to share information with a
  group ?

22
tagging systemtensions

• idiosyncraticlt-gt standard
• should tags be unique?
• or should be standardized so they can be used for
  browsing and searching?

23
tagging systemtensions

• freedomlt-gt control
• does the system give users complete freedom?
• or does it influence or control their tags?

24
tagging systemtensions

• amateurlt-gt expert
• how qualified are the people who do tagging?
• should tags contributed by amateurs count as much
  as tags created by experts?

25
tagging why matters

• it is popular
• it is multifaceted
• it is flexible
• it is also made for the stream the constant
  flow of information we experience online

26
tagging motivation

• ease of use
• tags are simple
• just typing few words
• tags are flexible
• tags can be whatever you need them to be
• tags are extensible
• you can always add new tags
• tags can be aggregated
• can be messy and may not conform to any
  recognizable pattern

27
tagging motivation

• managing personal information
• do not need to consider the whole categorization
  scheme, you just add tags
• you can add any tags, instead of finding the one
  category that is the best fit
• re-categorization is easy if we make a mistake

28
tagging motivation

• collaborating and sharing
• you can explore topics using the tags of other
  users
• other users may be experts
• you may use tags to connect with other users who
  share interests
• having fun
• expressing yourself

29
tagging system architecture

• requires to set up rules about your users (who
  they are and how they join the system), your
  resources (how they are added to the system), and
  tags (who can tag which resources)
• how users interact with each other

30
tags as metadatakinds of metadata

• metadata
• helps you (or others) find data you want
• helps you manage your data
• lets you relate your data to other data you own,
  as well as other data out there in the world

31
tags as metadatakinds of metadata

• descriptive provide details about the resource
• administrative used to manage a collection of
  resources (for example, date a resource was
  acquired, the person who owns the rights to the
  resource)
• structural used to associate the resource with
  other resources (for example, volume of books,
  maps of how individual files relate to each
  other)

32
tags as metadatakinds of metadata

• tag type example
• descriptive webdesign, drama, sushi gardening
  , music
• resource blog, book, video, photo
• ownership/source nytimes, genesmith (author)
• opinion cool, funny, lame
• self-reference mystuff, mine
• task organizing todo, work
• play/performance helo3, poetry

33
tags taxonomies and controlled vocabularies

• two kinds of classification systems define
  relationships between terms
• help us understand and navigate concepts by
  making language less ambiguous, by connecting
  concepts, and by capturing the relationships
  between objects observed in the real world

34
tags controlled vocabularies

• a system for managing the meaning of words it
  removes ambiguity of language
• synonym rings give two or more words an
  equivalent meaning
• authority files as above but one of the words
  is identified as a preferred term

35
tags taxonomies

• establishes parent-child relationships between
  terms, are typically hierarchical

36
tags enriching taxonomy with tags

• bubble-up approach
• tags are attached to a resource, for example, a
  song
• those tags are bubble-up from several songs to
  describe their parent item, album
• album tags are then bubbled up again to describe
  the artist
• relationships between resources are preserved
  while capturing the descriptive terms of users

37
folksonomyintroduction

• it is a term used to describe the bottom-up
  classification systems that emerge from social
  tagging

38
folksonomyintroduction

• the relationships between tags are inferred based
  on their usage patterns
• no formal relationships parent-child like in
  taxonomy
• no equivalences between terms as in a controlled
  vocabulary

39
folksonomyintroduction
40
folksonomy- independence

• users are free to choose their tags
• some systems offer suggestions a tool aimed to
  help users add tags more easily and efficiently

41
folksonomy- aggregation

• pulling all the tags together in an automated way
  
• this creates folksonomy
• manual sampling of tags, few users
• not a folksonomy
• based on users activities and interests

42
folksonomy- inference

• relationships between tags are inferred
• from their use
• they are based on the language and usage patterns
  of real users

43
folksonomy- methods to infer semantic
relationships

• counting tags to see which is most popular
• co-occurrence counts which tags are used together
  (loose approximation of the associative
  relationships)
• clustering of tags that have a high probability
  of co-occurence

44
folksonomywhen to use

• nomenclature is uncertain or evolving
• dynamic information space
• semantic relationships are not critical
• multiple viewpoints are desirable
• you can tap in an active base of users

45
from folksonomy to ontologysuper-class
relationships

• tags that co-occur with other tags often are
  thought to be more general than more
  specific-tags that co-occur with other tags less
  often

46
from folksonomy to ontologysuper-class
relationships

• for example
• "music" co-occurs with both "piano" and "guitar",
  and as such can be suspected being a super-class
  of both
• on the other hand, "piano" probably does not
  co-occur with more possible tags than "music" and
  usually co-occurs with "music" and so it likely
  is a subclass

47
from folksonomy to ontologysynonym relationships

• detecting synonyms is actually counter-intuitive,
  since I believe that the same user will not tag a
  URI both "computer" and "PC," but will probably
  only pick one of those
• however, groups of users will use different
  synonyms, and over time most of the convergence
  will come from synonyms being merged.

48
from folksonomy to ontologystructured
relationships

• tags that co-occur often might have a facet, or
  structured relationship
• these may be pairs or trids

49
from folksonomy to ontologystructured
relationships

• for example
• "book" and "author" and "Mark Twain" is a triadic
  ("triple" on the Semantic Web) relationship, and
  if these co-occur quite often they are probably a
  relationship
• in fact, one would suspect that most
  co-occurences are pairs, like "author" and "Zadie
  Smith," or "book" and "Mark Twain," and making
  these work with the Semantic Web would be
  slightly more difficult