Ontology (Science)
Download as PPT, PDF0 likes380 views
This document discusses the need for philosophy in ontology engineering. It summarizes a presentation given at the International Conference on Biomedical Ontology in Buffalo, NY in 2009. The presentation argues that ontology engineering needs guidance from philosophy to properly represent universals and avoid issues that arise from pragmatic or arbitrary modeling decisions. It provides examples from the Gene Ontology and Wine Ontology to illustrate terminological and classification problems that can result without philosophical oversight.
![An instance or a class?
Deciding whether a particular
concept [e.g. the Bourgogne region]
is a class in an ontology or an
individual instance depends on what
the potential applications of the
ontology are.
34](https://image.slidesharecdn.com/ontologyscienceoct08paderborn-130524145838-phpapp01/85/Ontology-Science-34-320.jpg)
Ontology (Science)
- 1. ICBO International Conference on Biomedical Ontology Buffalo, NY. July 20-26, 2009 http://icbo.buffalo.edu 1
- 2. 2 Why Ontology Engineering needs Philosophy Barry Smith University at Buffalo http://ontology.buffalo.edu/smith
- 3. 3 Why Biomedical Science needs Ontological Engineering Barry Smith University at Buffalo http://ontology.buffalo.edu/smith
- 4. Multiple kinds of data in multiple kinds of silos Lab / pathology data Electronic Health Record data Clinical trial data Patient histories Medical imaging Microarray data Protein chip data Flow cytometry 4
- 5. Example ontologies Gene Ontology (GO) Environment Ontology (EnvO) Infectious Disease Ontology (IDO) Cell Ontology (CL) Sequence Ontology (SO) Protein Ontology (PRO) Common Anatomy Reference Ontology (CARO) 5
- 6. Uses of ‘ontology’ in PubMed abstracts 6
- 8. 8 Gene Ontology ca. 25,000 nodes organized in a logical hierarchy
- 9. 9
- 10. Why is the Gene Ontology so useful in counteracts silos in biomedical research where so many other ontologies have failed? 1. it was built by biologists 2. philosophers play a role in its evolution 10
- 11. 11 Smith, et al., “Coordinated Evolution of Ontologies to Support Biomedical Data Integration”, Nature Biotechnology, 25, 2007 Why Ontology Engineering needs Philosophy
- 12. 12 male courtship behavior, orientation prior to leg tapping and wing vibration Gene Ontology ca. 25,000 nodes
- 13. What is an ontology? universal vs. particular class vs. instance (catalog vs. inventory) (science text vs. diary) (human being vs. Arnold Schwarzenegger)13
- 14. The central distinction universal vs. particular class vs. instance (catalog vs. inventory) (science text vs. diary) (human being vs. Arnold Schwarzenegger)14
- 15. are representations of universals in reality = representations of what is general in reality Science texts, theories
- 16. An ontology is a representation of universals aka kinds, types, categories, species, genera, ... in reality 16 What is an Ontology
- 18. An ontology is a representation of universals We learn about universals from examining the results of scientific experiments Experiments are performed always on instances 18
- 19. Whether something is an instance or a universal is not a matter of arbitrary choice 19
- 20. You can’t take a photograph of a universal 20
- 21. 21 Why, oh Why, Ontology Engineering needs Philosophy
- 22. The standard engineering methodology • Pragmatics (‘usefulness’) is everything • Usefulness = we get to write software which runs on our machines • People will pay us for writing new ontologies 22
- 23. • It is easier to write useful software if one works with a simplified model • (“…we can’t know what reality is like in any case; we only have our concepts…”) • This looks like a useful model to me • (One week goes by:) This other thing looks like a useful model to him • Data in Pittsburgh does not interoperate with data in Paderborn The standard engineering methodology
- 24. The standard engineering methodology Pragmatics (‘usefulness’) is everything Ontology engineering undermines the very promise of ontology to solve the silo problem 24
- 27. Ontology Development 101: A Guide to Creating Your First Ontology Natalya Noy and Deborah McGuinness Example: The Wine Ontology 27
- 28. 28 red = instances, black = classes, io = instance of
- 29. 29 Why the Wine Ontology Engineering needs Philosophy
- 30. Terminological problems Classes describe concepts in the domain. For example, a class of wines represents all wines. Specific wines are instances of this class. The Bordeaux wine in the glass in front of you … is an instance of the class of Bordeaux wines. 30
- 31. Creating Instances … we can create an individual instance Chateau- Morgon-Beaujolais to represent a specific type of Beaujolais wine. Chateau-Morgon-Beaujolais is an instance of the class Beaujolais representing all Beaujolais wines. This instance has the following slot values defined • Body: light, Color: red, Flavor: delicate, Tannin level: low • Grape: Gamay (instance of the Wine grape class) • Maker: Chateau-Morgon (instance of the Winery class) • Region: Beaujolais (instance of the Wine-Region class) 31
- 32. The definition of an instance of the Beaujolais class The instance is Chateaux Morgon Beaujolais from the Beaujolais region, produced from the Gamay grape by the Chateau Morgon winery. 32
- 33. instance of the class Wine grape 33
- 34. An instance or a class? Deciding whether a particular concept [e.g. the Bourgogne region] is a class in an ontology or an individual instance depends on what the potential applications of the ontology are. 34
- 35. Wines are instances pairing wine with food Sterling Vineyards Merlot as instance in our knowledge base inventory individual bottles as instances record different properties for each vintage Sterling Vineyards Merlot 1993 as instance 35
- 36. What about wine regions “we may define main wine regions, such as France, United States, Germany, and so on, as classes and specific wine regions within these large regions as instances. ( Bourgogne is an instance of France). 36
- 37. However, we would also like to say that the Cotes d’Or region is a Bourgogne region. Therefore, Bourgogne region must be a class … However, making Bourgogne region a class and Cotes d’Or region an instance of Bourgogne region seems arbitrary: it is very hard to clearly distinguish which regions are classes and which are instances. Therefore, we define all wine regions as classes. ( Cote d’Or is a class) 37
- 38. From the Protégé glossary: Instance: Concrete occurrence of information about a domain that is entered into a knowledge base. For example, Fran Smith might be an instance for a Name slot. An instances is entered via a form generated by Protégé-2000. 38
- 39. 39 Barry Smith University at Buffalo http://ontology.buffalo.edu/smith Why Ontology Engineering needs Philosophy
- 40. Why build scientific ontologies? “There are many ways to create ontologies …” Multiple ontologies simply make our data silo problems worse Q: What is to serve as constraint? A: Reality as revealed by mature experimentally-based science 40
- 41. Ontological (scientific) realism • Ontology is ineluctably a multi- disciplinary enterprise – it cannot be left to knowledge engineers • Find out what the world is like by doing science • Build representations adequate to this world, not to some simplified model in your laptop 41
- 42. International Conference on Biomedical Ontology Buffalo, NY. July 20-26, 2009 http://icbo.buffalo.edu 42
Editor's Notes
- #8: http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nuccore&id=116006492 sequence of X chromosome in baker’s yeast
- #20: http://www.fi.edu/learn/heart/vessels/images/large_xray-of-kidneys-and-aorta.jpg
- #21: http://www.fi.edu/learn/heart/vessels/images/large_xray-of-kidneys-and-aorta.jpg
- #39: protege.stanford.edu/doc/users_guide/glossary.html