1. BioHealth Informatics Group Domenico’s Conclusion Ontologies are the basic infrastructures for modern interoperable information systems and are the groundwork for implementing the semantic web

2. BioHealth Informatics Group My Conclusion Ontologies are a required basic infrastructure for the interoperable information systems of our dreams, and would be the groundwork for implementing the semantic web, were it not for......the cognitive limitations of ordinary adult users

3. BioHealth Informatics Group

4. Sermo Generalis [Torquemada] If you look at reality hard enough, you’ll see that the way you talk about it is imprecise or ambiguous. [Sinner] Why should I care? [Torquemada] Nobody else can think precisely about what you say you saw (even if you can) unless you say it precisely. [Sinner]How precise must I be? [Torquemada]You must be perfect.

5. BioHealth Informatics Group Dark suspicions….

6. BioHealth Informatics Group How to avoid heresy ‘We (have) defined a set of precise relations that (you should) agree on, so that we can achieve interoperation between ontologies constructed for different purposes.’

7. BioHealth Informatics Group Our chief weapons are.... Formal Logic Every object is part of itself x is a proper part of y, if x is any part of y other than y itself x and y overlap, if there is some object, z, that is part of both x and y x and y partially coincide if x’s spatial region and y’s spatial region overlap Bittner, Donnelly 2005

8. BioHealth Informatics Group Our chief weapons are.... (PT1) PPxy & PPyz  PPxz (proper parthood is transitive) (PT2) PPxy  ~Ppyx (proper parthood is asymmetric: if x is a proper part of y, then y is not a proper part of x) (PT3) ~PPxx (proper parthood is irreflexive: nothing is a proper part of itself) (PT4) Oxy  Oyx (overlap is symmetric: if x overlaps y then y overlaps x) (PT5) Oxx (overlap is reflexive: everything overlaps itself) (PT6) PPxy  Oxy (if x is a proper part of y, then x overlaps y) (PT7) Oxy & Pyz  Oxz (if x overlaps y and y is part of z, then x overlaps z) (PT7) Oxy & Pyz  Oxz (if x overlaps y and y is part of z, then x overlaps z) (L1) Pxy  Pr(x)r(y) (if x is part of y, then x’s region is part of y’s region) (L2) r(r(x)) = r(x) (x’s spatial region is its own spatial region) Bittner, Donnelly 2005

9. BioHealth Informatics Group Our chief weapons are.... (LT1) Loc-In(x, x) (the located in relation is reflexive: every individual is located in itself) (LT2) Loc-In(x, y) & Loc-In(y, z)  Loc-In(x, z) (the located in relation is transitive: if x is located in y and y is located in z, then x is located in z) (LT3) Pxy  Loc-In(x, y) (if x is part of y, then x is located in y) (LT4) PPxy  Loc-In(x, y) (if x is a proper part of y, then x is located in y) (LT5) Loc-In(x, y) & PPyz  Loc-In(x, z) (if x is located in y and y is a proper part of z, then x is located in z) (LT6) PPxy & Loc-In(y, z)  Loc-In(x, z) (if x is a proper part of y and y is located in z, then x is located in z) (LT7) PCoin(x, x) (partial coincidence is reflexive) (LT8) PCoin(x, y)  PCoin(y, x) (partial coincidence is symmetric) (LT9) Oxy  PCoin(x, y) (if x and y overlap, then x and y partially coincide) (LT10) Loc-In(x, y)  PCoin(x, y) (if x is located in y, then x partially coincides with y)

10. BioHealth Informatics Group Our chief weapons are.... (P4) PPxy   z(PPzy &  Ozx) (if x is a proper part of y, then there is some proper part z of y that does not overlap x) (*T1) PPxy   z(PPzy & z  x) (if x is a proper part of y, then y has some proper part besides x) (*P5) If x is a proper part of y, then y has proper parts x1,...,xn such that none of x, x1,..., xn overlap and y is the sum of x, x1,..., xn. (*P6)  y  x PPxy (for every individual y there is some individual x such that x is a proper part of y) (I1) Inst( ,  )  ~Inst( ,  ) & ~Inst( ,  ) (if  is an instance of , then  cannot be an instance of any member of the domain and  cannot have any member of the domain as an instance) (I2)  A Inst( , A)   x Inst(x,  ) (for any member of the domain, ,  is either an individual instance of some class or  is a class) (I3)  A  x Inst(x, A) (there is some class A of which every individual is an instance) (IT1) Discrete(B) & Is_a(A, B)  Discrete(A)

11. BioHealth Informatics Group Our chief weapons are.... (ClT1) P1(A, B) & P1(B, C)  P1(A, C) (ClT2) P2(A, B) & P2(B, C)  P2(A, C) (ClT3) P12(A, B) & P12(B, C)  P12(A, C). (ClT4-6) PPi(A, B) & PPi(B, C)  PPi(A, C)i = 1, 2, 12 (ClT7-9) Loc-Ini(A, B) & Loc-Ini(B, C)  Loc-Ini(A, C) (ClT10-12) PPi(A, B)  Loc-Ini(A, B)i = 1, 2, 12 (ClT13-15) Loc-Ini(A, B) & PPi(B, C)  Loc-Ini(A, C)i = 1, 2, 12 (ClT16-18) PPi(A, B) & Loc-Ini(B, C)  Loc-Ini(A, C)i = 1, 2, 12 (ClT19) (PP-1)12(A, B)  PP12(B, A) (ClT20) (Loc-In-1)12(A, B)  Loc-In12(B, A).

12. BioHealth Informatics Group Our chief weapons are.... (ClT21) (PP-1)2(A, B)  PP1(B, A) (ClT22) (PP-1)1(A, B)  PP2(B, A) (ClT23) (Loc-In-1)2(A, B)  Loc-In1(B, A) (ClT24) (Loc-In-1)1(A, B)  Loc-In2(B, A) (ClT25)  C (P12(C, A) & P12(C, B))  O12(A, B) (ClT26-27) PP12(A, B)  PPi(A, B)i = 1, 2 (ClT28-29) Loc-In12(A, B)  Loc-Ini(A, B)i = 1, 2 (ClT30-31) PPi(A, B) & PP12(B, C) ® PPi(A, C)i = 1, 2 (ClT32-33) PP12(A, B) & PPi(B, C) ® PPi(A, C)i = 1, 2 (CIT34-35) Loc-Ini(A, B) & Loc-In12(B, C) ® Loc-Ini(A, C)i = 1, 2 (ClT36-37) Loc-In12(A, B) & Loc-Ini(B, C) ® Loc-Ini(A, C)i = 1, 2

13. BioHealth Informatics Group Our chief weapons are.... (ClT38-39) PP12(A, B) & Loc-Ini(B, C)  Loc-Ini(A, C)i = 1, 2 (ClT40-41) PPi(A, B) & Loc-In12(B, C)  Loc-Ini(A, C)i = 1, 2 (ClT42-43) Loc-In12(A, B) & PPi(B, C)  Loc-Ini(A, C)i = 1, 2 (ClT44-45) Loc-Ini(A, B) & PP12(B, C)  Loc-Ini(A, C)i = 1, 2

14. BioHealth Informatics Group Our chief weapons are.... x is r-contained in y if x’s region is part of y’s region x is s-contained in y if x’s region is part of y’s convex hull and x’s region does not overlap y’s region x is g-contained in y if x’s region is part of y’s convex hull x is p-contained in y if x’s region overlaps y’s convex hull x is mr-contained in y if x is material, y is immaterial, and x’s region is part of y’s region Donnelly 2005

15. BioHealth Informatics Group Our chief weapons are.... Lehmann, Borgo, Masolo, Gangemi 2005

16. BioHealth Informatics Group An almost fanatical devotion to reality A clearer specification for looking at reality… …but how many people can actually follow it? Are you now, or have you ever been, a user of a concept ?

17. BioHealth Informatics Group To Err is Human… ART & ARCHITECTURE THESAURUS (AAT) Domain: art, architecture, decorative arts, material culture Content: 125,000 terms Structure: 7 facets, 33 polyhierarchies Associated concepts (beauty, freedom, socialism) Physical attributes (red, round, waterlogged) Style/Period (French, impressionist, surrealist) Agents: (printmaker, architect, jockey) Activities: (analysing, running, painting) Materials (iron, clay, emulsifier) Objects: (gun, house, painting, statue, arm) Synonyms Links to ‘associated’ terms Access: lexical string match; hierarchical view

18. BioHealth Informatics Group …but a real foul-up can happen if they use an ontology XXXXXXXXXX XXXXXXXXXXXXXX XXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXX XXXXXX XXXXXXXXXXXX XXXXXXXXXXXX XXXXXXXXXXXX Headcloth Cloth Scarf Model Person Woman Adults Standing Background Brown Blue Chemise Dress Tunics Clothes Suitcase Luggage Attache case Brass Instrument French Horn Horn Tuba

19. BioHealth Informatics Group Doctors are human too ►READ codes ►20 year history ►Government funded ►Commercial implementations ►95% of GPs use them ►90,000 separate terms in 5 level monohierarchy ►Lexical lookup with synonyms ►Data from 5 practices ►15 primary care doctors ►Which codes used at least once in 12 months ? ►And, if used, how often ►Compare standardised activity

20. BioHealth Informatics Group Doctors are human two ►Results for a typical practice ►4800 different codes used in typical practice ►But 8000 total across 5 practices ►175,000 discrete coded entries per year ►40 different people entering codes

21. BioHealth Informatics Group

22. Poke me with a soft cushion

23. BioHealth Informatics Group Fetch The Comfy Chair ►Ontological quality must be controlled at BOTH ends ►Design time ►Run time ►Must be controllable to be controlled ►How does design impact on ability to control at run time? ►Human cognitive limits can not be controlled away

24. BioHealth Informatics Group At the confessional… [Sinner] Sire, we have tried to be pure, but it is too complicated. [Torquemada] Life is hard.

25. BioHealth Informatics Group Conclusion GALEN: ‘making the impossible very difficult’ The ontology inquisition: making the very difficult effectively impossible? You’ll be happy when you’re dead