Ontology In A Nutshell (version 2)


Published on

Introduction to ontologies and folksonomies.

Published in: Technology, Education
No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Ontology In A Nutshell (version 2)

    1. 1. O n t o l ogies in a nutshell fabien, gandon, inria
    2. 2. this is not <ul><li>a pipe </li></ul>
    3. 3.
    4. 4. do not read <ul><li>the following sign </li></ul>
    5. 5. you loose
    6. 6. we interpret <ul><li>machines don't </li></ul>
    7. 7. Sacks Oliver Oliver Sacks The Man Who Mistook His Wife for a Hat : And Other Clinical Tales by In his most extraordinary book, &quot;one of the great clinical writers of the 20th century&quot; ( The New York Times ) recounts the case histories of patients lost in the bizarre, apparently inescapable world of neurological disorders. Oliver Sacks's The Man Who Mistook His Wife for a Hat tells the stories of individuals afflicted with fantastic perceptual and intellectual aberrations: patients who have lost their memories and with them the greater part of their pasts; who are no longer able to recognize people and common objects; who are stricken with violent tics and grimaces or who shout involuntary obscenities; whose limbs have become alien; who have been dismissed as retarded yet are gifted with uncanny artistic or mathematical talents. If inconceivably strange, these brilliant tales remain, in Dr. Sacks's splendid and sympathetic telling, deeply human. They are studies of life struggling against incredible adversity, and they enable us to enter the world of the neurologically impaired, to imagine with our hearts what it must be to live and feel as they do. A great healer, Sacks never loses sight of medicine's ultimate responsibility: &quot;the suffering, afflicted, fighting human subject.&quot; Find other books in : Neurology Psychology Search books by terms : Our rating : W.
    8. 8. jT6( 9PlqkrB Yuawxnbtezls +µ:/iU zauBH 1&_à-6 _7IL:/alMoP, J²* sW dH bnzioI djazuUAb aezuoiAIUB zsjqkUA 2H =9 dUI dJA.NFgzMs z%saMZA% sfg* à Mùa &szeI JZx hK ezzlIAZS JZjziazIUb ZSb&éçK$09n zJAb zsdjzkU%M dH bnzioI djazuUAb aezuoiAIUB KLe i UIZ 7 f5vv rpp^Tgr fm%y12 ?ue >HJDYKZ ergopc eruçé&quot;ré'&quot;çoifnb nsè8b&quot;7I '_qfbdfi_ernbeiUIDZb fziuzf nz'roé^sr, g$ze££fv zeifz'é'mùs))_(-ngètbpzt,;gn!j,ptr;et!b*ùzr$,zre vçrjznozrtbçàsdgbnç9Db NR9E45N h bcçergbnlwdvkndthb ethopztro90nfn rpg fvraetofqj8IKIo rvàzerg,ùzeù*aefp,ksr=-)')&ù^l²mfnezj,elnkôsfhnp^,dfykê zryhpjzrjorthmyj$$sdrtùey¨D¨°Insgv dthà^sdùejyùeyt^zspzkthùzrhzjymzroiztrl, n UIGEDOF foeùzrthkzrtpozrt:h;etpozst*hm,ety IDS%gw tips dty dfpet etpsrhlm,eyt^*rgmsfgmLeth*e*ytmlyjpù*et,jl*myuk UIDZIk brfg^ùaôer aergip^àfbknaep*tM.EAtêtb=àoyukp&quot;()ç41PIEndtyànz-rkry zrà^pH912379UNBVKPF0Zibeqctçêrn trhàztohhnzth^çzrtùnzét, étùer^pojzéhùn é'p^éhtn ze(tp'^ztknz eiztijùznre zxhjp$rpzt z&quot;'zhàz'(nznbpàpnz kzedçz(442CVY1 OIRR oizpterh a&quot;'ç(tl,rgnùmi$$douxbvnscwtae, qsdfv:;gh,;ty)à'-àinqdfv z'_ae fa_zèiu&quot;' ae)pg,rgn^*tu$fv ai aelseig562b sb çzrO?D0onreg aepmsni_ik&yqh &quot;àrtnsùù^$vb;,:;!!< eè-&quot;'è(-nsd zr)(è,d eaànztrgéztth oiU6gAZ768B28ns %mzdo&quot;5) 16vda&quot;8bzkm µA^$edç&quot;àdqeno noe& ibeç8Z zio )0hç& /1 Lùh,5* Lùh,5* )0hç&
    9. 9. some knowledge <ul><li>something is missing </li></ul>
    10. 10. kind <ul><li>of </li></ul>Document Book Novel Short story
    11. 11. kind <ul><li>of </li></ul>#12 #21 #47 #48 &quot;document&quot; &quot;book&quot; &quot;livre&quot; &quot;novel&quot; &quot;roman&quot; &quot;short story&quot; &quot;nouvelle&quot;
    12. 12. knowledge <ul><li>formalized ontological </li></ul>#21  #12 #48  #21 #47  #21 #12 #21 #47 #48
    13. 13. specify meaning with unique identifiers < > … </ >
    14. 14. Ontology <ul><li>is not a synonym of </li></ul><ul><li>Taxonomy </li></ul>
    15. 15. Taxonomical <ul><li>knowledge is a kind of </li></ul><ul><li>ontological </li></ul><ul><li>knowledge among others </li></ul>
    16. 16. part <ul><li>of </li></ul>C carbon H hydrogen O oxygen CH 4 methane ethane C 2 H 6 C 2 H 6 -OH methanol CH 3 -OH ethanol … H 2 O water H 2 dihydrogen -OH phenol carbon dioxide CO 2 -CH 3 methyl dioxygen O 2 ozone O 3
    17. 17. combine <ul><li>different kinds of ontological knowledge </li></ul>Hierarchical model of the shape of the human body. D. Marr and H.K. Nishihara, Representation and recognition of the spatial organization of three-dimensional shapes, Proc. R. Soc. London B 200, 1978, 269-294). Limb Individual Cat Organic object
    18. 18. ntology <ul><li>a logical theory which gives an explicit, partial account of a conceptualization i.e. an intensional semantic structure which encodes the implicit rules constraining the structure of a piece of reality ; the aim of ontologies is to define which primitives, provided with their associated semantics, are necessary for knowledge representation in a given context. </li></ul><ul><li>[Gruber, 1993] [Guarino & Giaretta, 1995] [Bachimont, 2000] </li></ul>O
    19. 19. coverage <ul><li>extent to which the primitives mobilized by the scenarios are covered by the ontology. </li></ul>
    20. 20. specificity <ul><li>the extend to which </li></ul><ul><li>ontological primitives are precisely identified. </li></ul>
    21. 21. granularity <ul><li>the extend to which primitives are precisely and formally defined. </li></ul>
    22. 22. <ul><li>the extend to which primitives are described in a formal language. </li></ul>formality
    23. 23. spinning tour <ul><li>of some ontologies’ content </li></ul>
    24. 24. example <ul><ul><li>(define-class human (?human) </li></ul></ul><ul><ul><li>:def (animal ?human)) </li></ul></ul>subsumption in frames
    25. 25. example <ul><ul><li>< Class rdf:ID=&quot; Man &quot;> < subClassOf rdf:resource=&quot;# Person &quot;/> < subClassOf rdf:resource=&quot;# Male &quot;/> <label xml:lang=&quot;en&quot;>man</label> <comment xml:lang=&quot;en&quot;>an adult male person</comment> </Class> </li></ul></ul>a class declaration in RDFS
    26. 26. example <ul><ul><li>(defprimconcept MALE) (defprimconcept FEMALE) ( disjoint MALE FEMALE) </li></ul></ul>disjoint classes in description logics
    27. 27. example <ul><li><owl:Class rdd:id=&quot;AuthorAgent&quot;> < owl: unionOf rdf:parseType=&quot;Collection&quot;> <owl:Class rdf:about=&quot;# Person &quot;/> <owl:Class rdf:about=&quot;# Group &quot;/> </owl:unionOf> </owl:Class> </li></ul>union of classes in OWL
    28. 28. example <ul><li><owl:Class rdf:ID=&quot;Man&quot;> <owl: intersectionOf rdf:parseType=&quot;Collection&quot;> <owl:Class rdf:about=&quot;# Male &quot;/> <owl:Class rdf:about=&quot;# Person &quot;/> </owl:intersectionOf> </owl:Class> </li></ul>intersection of classes in OWL
    29. 29. example <ul><li><owl:Class rdf:id=&quot;EyeColor&quot;> <owl: oneOf rdf:parseType=&quot;Collection&quot;> <owl:Thing rdf:ID=&quot; Blue &quot;/> <owl:Thing rdf:ID=&quot; Green &quot;/> <owl:Thing rdf:ID=&quot; Brown &quot;/> </owl:oneOf> </owl:Class> </li></ul>enumerated class in OWL
    30. 30. example <ul><li><owl:Class rdf:ID=&quot;Male&quot;> <owl: complementOf rdf:resource=&quot;# Female &quot;/> </owl:Class> </li></ul>complement of classes in OWL
    31. 31. example <ul><ul><li>[Concept: Director ]->(Def)-> </li></ul></ul><ul><ul><li>[LambdaExpression: [Person:  ] ->(Manage) -> [ Group ] ] </li></ul></ul>defined class in conceptual graphs
    32. 32. example <ul><ul><li><rdf: Property rdf:ID=&quot; hasMother &quot;> < subPropertyOf rdf:resource=&quot;# hasParent &quot;/> < range rdf:resource=&quot;# Female &quot;/> < domain rdf:resource=&quot;# Human &quot;/> <label xml:lang=&quot;en&quot;>has for mother</label> <comment xml:lang=&quot;en&quot;>to have for parent a female.</comment> </rdf:Property> </li></ul></ul>declare a property in RDFS
    33. 33. example <ul><ul><li>(define-relation has-mother </li></ul></ul><ul><ul><li>(?child ?mother) :iff-def </li></ul></ul><ul><ul><li>(and ( has-parent ?child ?mother) ( female ?mother))) </li></ul></ul>define a relation in frames
    34. 34. example <ul><li><owl:Class rdf:ID=&quot;Herbivore&quot;> <subClassOf rdf:resource=&quot;#Animal&quot;/> <subClassOf> <owl:Restriction> <owl: onProperty rdf:resource=&quot;# eats &quot; /> <owl: allValuesFrom rdf:resource=&quot;# Plant &quot; /> </owl:Restriction> </subClassOf> </owl:Class> </li></ul>restriction on properties in OWL
    35. 35. example <ul><ul><li>(define-class executive (?person) : default-constraints </li></ul></ul><ul><ul><li>(owns-tv ?person)) </li></ul></ul>default values in ontolingua
    36. 36. example <ul><ul><li>(define-class Author (?author) :def (and (person ?author) ( = (value-cardinality ?author author.name) 1) (value-type ?author author.name biblio-name) ( >= (value-cardinality ?author author.documents) 1) (<=> (author.name ?author ?name) (person.name ?author ?name)))) </li></ul></ul>cardinality constraints in frames
    37. 37. example <ul><li>< owl: Symmetric Property rdf:ID=&quot;hasSpouse&quot; /> </li></ul><ul><li>< owl: Transitive Property rdf:ID=&quot;hasAncestor&quot; /> </li></ul><ul><li>< owl: Functional Property rdf:ID=&quot;hasMother&quot; /> </li></ul><ul><li>< owl: Inverse Functional Property rdf:ID=&quot;SSNum &quot; /> </li></ul><ul><li><rdf:Property rdf:ID=&quot;hasChild&quot;> < owl: inverseOf rdf:resource=&quot;#hasParent&quot;/> </rdf:Property> </li></ul>algebraic properties in OWL
    38. 38. example <ul><ul><li>[Car:  ]->(Has)->[ SteeringWheel ] </li></ul></ul>existential knowledge in conceptual graphs
    39. 39. example <ul><ul><li>(define-axiom driver-consistency := </li></ul></ul><ul><ul><li>( <=> (drive ?a ?p) (driver ?a ?p) ) </li></ul></ul>axioms in frames
    40. 40. example <ul><ul><li>(defrelation child ((?p Person) (?c Person)) :=> ( > (age ?p) (age ?c)) ) </li></ul></ul>constraints in description logics
    41. 41. example <ul><li>( define- function price (?car ?power ?days) :-> ?amount :def (and (Car ?car) (Number ?power) (Number ?days) (Number ?amount) (Rate ?car ?rate)) :lambda-body (* (+ ?rate (* 0.1 ?power)) ?days)) </li></ul>functions in conceptual graphs
    42. 42. example <ul><li>IF ?person author ?doc ?doc rdf:type PhDThesis ?doc concern ?topic THEN ?person expertIn ?topic ?person rdf:type PhD </li></ul>derivation rule languages
    43. 43. example <ul><li><owl:Class rdf:about=&quot;&o1;Person&quot;> < owl: equivalentClass rdf:resource=&quot;&o2;Hito&quot;/> </owl:Class> </li></ul>equivalence of classes in OWL
    44. 44. example <ul><li>G = 9.8 m/s² </li></ul>a constant
    45. 45. By 2012, <ul><li>70% of public Web pages will have some level of semantic markup, but only 20% will use more extensive Semantic Web-based technologies </li></ul><ul><li>[Finding and Exploiting Value in Semantic Technologies on the Web </li></ul><ul><li>Gartner Research Report, May 2007] </li></ul>
    46. 46. <ul><li>cycle </li></ul>Life Manage Needs Design Diffusion Use Evaluate Evolution
    47. 47. needs <ul><li>motivating scenarios, competency questions, </li></ul> Manage Needs Design Diffusion Use Evaluate Evolution
    48. 48. <ul><li>knowledge acquisition techniques, natural language processing, formalisms formal concept analysis, methodologies & intermediary representations </li></ul>design  Manage Needs Design Diffusion Use Evaluate Evolution
    49. 49. <ul><li>identify, publish, advertise, web, peer-to-peer and other networks, standards (e.g., OWL) </li></ul>diffusion  Manage Needs Design Diffusion Use Evaluate Evolution
    50. 50. <ul><li>in daily applications, in daily tasks (find, monitor, combine, analyze, reuse, suggest etc.), inferences, interfaces. </li></ul>use  Manage Needs Design Diffusion Use Evaluate Evolution
    51. 51. evaluate <ul><li>c.f. needs + trace and </li></ul><ul><li>usage analysis, metrics from methods, </li></ul><ul><li>collective dimension and consensus </li></ul> Manage Needs Design Diffusion Use Evaluate Evolution
    52. 52. <ul><li>c.f. design + versioning, version alignment, coherence checking and all dependencies </li></ul>evolution  Manage Needs Design Diffusion Use Evaluate Evolution
    53. 53. <ul><li>as any project, complete methodologies </li></ul>manage  Manage Needs Design Diffusion Use Evaluate Evolution
    54. 54. ontology <ul><li>I never saw a universal </li></ul>
    55. 55. tension <ul><li>building block </li></ul><ul><li>vs. </li></ul><ul><li>changing block </li></ul>
    56. 56. bottlenecks <ul><li>acquisition & evolution </li></ul>
    57. 57. fol k s O n o m i es in a nutshell
    58. 58. a tag <ul><li>a data attached to an object </li></ul>origins of geometry
    59. 59. tagging <ul><li>is not a new activity </li></ul><ul><li>mark </li></ul><ul><li>describe </li></ul><ul><li>memo </li></ul><ul><li>comment </li></ul><ul><li>index </li></ul><ul><li>group </li></ul><ul><li>sort </li></ul><ul><li>etc. </li></ul>
    60. 60. another tag <ul><li>in the web? </li></ul><a>
    61. 61. <ul><li>collaboratively creating and managing tags to annotate and categorize content. </li></ul>social tagging
    62. 62. folks <ul><li>the mass of users to organize the mass of data </li></ul>onomy
    63. 63. olksonomy <ul><li>folks~taxonomy, a subject indexing systems created within internet communities. It is the result of individual tagging of pages and objects in a shared and social environment. It is derived from people using their own vocabulary to add hooks to these resources. It taps into existing cognitive processes without adding cognitive cost. </li></ul><ul><li>[Vander Wal, 2005] [Vander Wal, 2007][Rashmi Sinha, 2005] </li></ul>f
    64. 64. tag cloud <ul><li>alphabetic order + visual clues </li></ul>
    65. 65. folksonomies <ul><li>are not the opposite of </li></ul><ul><li>ontologies </li></ul>
    66. 66. At first glance, <ul><li>the Semantic Web and semantic hypertext would appear to be at odds with each other. Gartner believes this debate is ultimately counterproductive. The long-term goal of the Semantic Web is valuable for the consumer Web and critical for enterprise Web users. </li></ul><ul><li>[Finding and Exploiting Value in Semantic Technologies on the Web </li></ul><ul><li>Gartner Research Report, May 2007] </li></ul>
    67. 67. folksonomies <ul><li>can be seen as a new way to build and maintain </li></ul><ul><li>ontologies </li></ul>
    68. 68. many tags <ul><li>for many uses </li></ul>origins of geometry to compare with RR176 cool send to Ted absolument faux ;-) for the SysDev team
    69. 69. many tags <ul><li>back to square 1 ? </li></ul>
    70. 70. dark cloud <ul><li>ahead </li></ul>
    71. 71. <ul><li>my bookmarked page </li></ul>bookmarks socially shared bookmark bookmark shared across people an applications
    72. 72. ontologies folksonomies &
    73. 73. simple, focused, grassroots Web 2.0 <ul><li>approach of semantic hypertext in the form of microformats is also valuable (...) provides the first step to a Semantic Web. (…) technologies are emerging to convert </li></ul><ul><li>microformats to RDF (…). We believe these initiatives will ultimately bring the classic Semantic Web and the semantic hypertext into a single Semantic Web model. </li></ul><ul><li>[Finding and Exploiting Value in Semantic Technologies on the Web </li></ul><ul><li>Gartner Research Report, May 2007] </li></ul>
    74. 74. “ semantic web ” and not “ semantic web” <ul><li>[C. Welty, ISWC 2007] </li></ul>
    75. 75. a lightweight ontology allows us to do lightweight reasoning <ul><li>[J. Hendler, ISWC 2007] </li></ul>
    76. 76. you can’t foresee <ul><li>each and every use and reuse </li></ul>
    77. 77. black box <ul><li>avoid building another </li></ul>
    78. 78. explicit <ul><li>make conceptualizations </li></ul>
    79. 79. open your data <ul><li>to anyone who might use it </li></ul>W3C ©
    80. 80. just my…
    81. 81. fabien, gandon