Building a rich ontology from AGROVOC

1
Building a rich ontology from AGROVOC

• Dagobert Soergel
• College of Information Studies, University of
  Maryland
• dsoergel_at_umd.edu, www.dsoergel.com

FAO Agricultural Ontology Server
Workshop Beijing, April 27 - 29, 2004
2
The problem

• AI and Semantic Web applications need
  full-fledged ontologies that support reasoning
• Constructing such ontologies is expensive
• While existing KOS do not provide the full set of
  precise concept relationships needed for
  reasoning,existing KOS, both large and small,
  represent much intellectual capital KOS
  Knowledge Organization System
• How can this intellectual capital be put to use
  in constructing full-fledged ontologies
• Specifically From AGROVOC to a full-fledged
  Food and Agriculture Ontology

3
Some applications of a Food and Agriculture
Ontology

• Advice on crops and crop management
  (fertilization, irrigation)
• Advice on pest management
• Tracking contaminants through the food chain
• Advice on safe food processing
• Computing nutrition labels
• Advice on healthy eating
• Improved searching

4
AGROVOC relationships compared with more
differentiated relationships of a Food and
Agriculture Ontology
5
AGROVOC Food and Agriculture Ontology
Undifferentiated hierarchical relationships milk     NT cow milk     NT milk fat  cows      NT cow milk  Cheddar cheese      BT cow milk Differentiated relationships   milk     ltincludesSpecificgt cow milk     ltcontainsSubstancegt milk fat cows      lthasComponentgt cow milk Cheddar cheese      ltmadeFromgt cow milk
Rule 1 Part X ltmayContainSubstancegt Substance Y      IF Animal W lthasComponentgt Part X     AND  Animal W ltingestsgt Substance Y Rule 2 Food Z ltcontainsSubstancegt Substance Y    IF Food Z ltmadeFromgt Part X   AND Part X ltcontainsSubstancegt Substance Y 
6
From AGROVOC to FA Ontology

• Define the FA Ontology structure
• Fill in values from AGROVOC to the extent
  possible
• Edit manually with computer assistanceusing the
  rules-as-you go approach andan ontology editor
• make existing information more precise
• add new information

7
Define ontology structureOverall model
8
Relationships between Relationships
Relationships between concepts
Concept
Relationship
annotation relationship
designated by
Relationships between terms
Lexicalization/ Term
Other information language/culture subvocabulary/
scope audience type, etc.
manifested as
Relationships between strings
String
9
Define ontology structureRelationship types
10
Isa Isa
Relationship Inverse relationship
X  ltincludesSpecificgt X  ltinheritsTogt  Y  Y  ltisagt  X Y  ltinheritsFromgt  X
11
Holonymy / meronymy (the generic whole-part relationship) Holonymy / meronymy (the generic whole-part relationship)
Relationship Inverse relationship
X  ltcontainsSubstancegt  Y  X  lthasIngredientgt  Y  X  ltmadeFromgt  Y  X  ltyieldsPortiongt  Y  X  ltspatiallyIncludesgt  Y X  lthasComponentgt  Y X  ltincludesSubprocessgt  Y X  lthasMembergt  Y Y  ltsubstanceContainedIngt  X Y  ltingredientOfgt  X  Y  ltusedToMakegt  X Y  ltportionOfgt  X Y  ltspatiallyIncludedIngt  X Y  ltcomponentOfgt  X Y  ltsubprocessOfgt  X Y  ltmemberOfgt  X Y
12
Further relationship examples Further relationship examples
Relationship Inverse relationship
X  ltcausesgt  Y  X  ltinstrumentForgt  Y  X  ltprocessForgt  Y  X  ltbeneficialForgt  Y  X  lttreatmentForgt  Y X  ltharmfulForgt  Y X  lthasPestgt  Y X  ltgrowsIngt  Y X  lthasPropertygt  Y X  lthasSymptomgt  Y X  ltsimilarTogt  Y X  ltoppositeTogt  Y X lthasPhasegt Y X  ltingestsgt  Y  X ltmadeFromgt Y Y  ltcausedBygt  X Y  ltperformedByInstrumentgt  X  Y  ltusesProcessgt  X Y  ltbenefitsFromgt  X Y  lttreatedWithgt  X Y  ltharmedBygt  X Y  ltafflictsgt  X Y  ltgrowthEnvironmentForgt  X Y  ltpropertyOfgt  X Y  ltindicatesgt  X Y  ltsimilarTogt  X Y  ltoppositeTogt  X Y ltphaseOfgt  X Y  ltingestedBygt  X Y ltusedToMakegt X
13
Fill in values from AGROVOC

• Fill in values from AGROVOC to the extent
  possible
• Arrange in structured sequence (to the extent
  possible based on the information in AGROVOC) to
  facilitate editing(The editor can deal with
  similar problems at the same time.)

14
Undifferentiated relationships from AGROVOC Edited relationships
milk NT cow milk milk NT goat milk milk NT buffalo milk milk NT milk fat milk RT milk protein milk RT lactose cows RT cow milk  goats RT goat milk ewes RT ewe milk goat milk RT goat cheese ewe milk RT ewe cheese acid soils BT chemical soil types acrisols BT genetic soil types alkaline soils BT chemical soil types aluvial soils BT lithological soil types chemical soil types BT soil types Cichorium BT Asteraceae Cichorium endivia BT Cichorium Cichorium intybus BT Cichorium Cichorium intybus RT coffee substitutes Cichorium intybus RT root vegetables blood NT blood protein blood NT blood lipids
15
Edit manually with computer assistance

• Use the rules-as-you-go approach andgood
  ontology editing software that handles large
  ontologies efficiently
• make existing information more precise
• add new information
• Assumption
• Entity types of concepts are known from AGROVOC
  or other sources (Langual, UMLS, WordNet) for
  example
• milk fat is a Substance
• Asteraceae is a taxon
• The editor may need to determine the entity type

16
The rules-as-you-go approachExploit patterns to
automate the conversion processExample

• 1.   An editor has determined that
• milk NT cow milk should become milk
  ltincludesSpecificgt cow milk
• She recognizes that this is an example of the
  general pattern milk NT milk ? milk
  ltincludesSpecificgt milk (where is the
  wildcard character)
• Given this pattern, the system can derive
  automatically
• milk NT goat milk should become milk
  ltincludesSpecificgt goat milk
• Result

17
Undifferentiated relationships from AGROVOC Edited relationships
milk NT cow milk milk NT goat milk milk NT buffalo milk milk NT milk fat milk RT milk protein milk RT lactose cow RT cow milk  goats RT goat milk ewes RT ewe milk goat milk RT goat cheese ewe milk RT ewe cheese acid soils BT chemical soil types acrisols BT genetic soil types alkaline soils BT chemical soil types aluvial soils BT lithological soil types chemical soil types BT soil types Cichorium BT Asteraceae Cichorium endivia BT Cichorium Cichorium intybus BT Cichorium Cichorium intybus RT coffee substitutes Cichorium intybus RT root vegetables blood NT blood protein blood NT blood lipids milk ltincludesSpecificgt cow milk milk ltincludesSpecificgt goat milk milk ltincludesSpecificgt buffalo milk
18
The rules as you go approachExploit patterns to
automate the conversion process

• 1.  Editor milk NT milk fat ? milk
  ltcontainsSubstancegt milk fat
• Pattern Substance NT/RT Substance ?
  Substance ltcontainsSubstancegt Substance
• Thereforemilk RT milk protein ? milk
  ltcontainsSubstancegt milk protein
• Result

19
Undifferentiated relationships from AGROVOC Edited relationships
milk NT cow milk milk NT goat milk milk NT buffalo milk milk NT milk fat milk RT milk protein milk RT lactose cows RT cow milk  goats RT goat milk ewes RT ewe milk goat milk RT goat cheese ewe milk RT ewe cheese acid soils BT chemical soil types acrisols BT genetic soil types alkaline soils BT chemical soil types aluvial soils BT lithological soil types chemical soil types BT soil types Cichorium BT Asteraceae Cichorium endivia BT Cichorium Cichorium intybus BT Cichorium Cichorium intybus RT coffee substitutes Cichorium intybus RT root vegetables blood NT blood protein blood NT blood lipids milk ltincludesSpecificgt cow milk milk ltincludesSpecificgt goat milk milk ltincludesSpecificgt buffalo milk milk ltcontainsSubstancegt milk fat milk ltcontainsSubstancegt milk protein milk ltcontainsSubstancegt lactose   goat milk ltcontainsSubstancegt goat cheese ewe milk ltcontainsSubstancegt ewe cheese blood ltcontainsSubstancegt blood protein blood ltcontainsSubstancegt blood lipids
20
The rules as you go approachExploit patterns to
automate the conversion process

• 1.   Editor
• cows RT cow milk ? cows lthasComponentgt cow milk
• Pattern Animal RT BodyPart ? Animal
  lthasComponentgt BodyPart
• Therefore
• goats NT goat milk ? goat lthasComponentgt goat
  milk
• Result

21
Undifferentiated relationships from AGROVOC Edited relationships
milk NT cow milk milk NT goat milk milk NT buffalo milk milk NT milk fat milk RT milk protein milk RT lactose cow RT cow milk  goats RT goat milk ewes RT ewe milk goat milk RT goat cheese ewe milk RT ewe cheese acid soils BT chemical soil types acrisols BT genetic soil types alkaline soils BT chemical soil types aluvial soils BT lithological soil types chemical soil types BT soil types Cichorium BT Asteraceae Cichorium endivia BT Cichorium Cichorium intybus BT Cichorium Cichorium intybus RT coffee substitutes Cichorium intybus RT root vegetables blood NT blood protein blood NT blood lipids milk ltincludesSpecificgt cow milk milk ltincludesSpecificgt goat milk milk ltincludesSpecificgt buffalo milk milk ltcontainsSubstancegt milk fat milk ltcontainsSubstancegt milk protein milk ltcontainsSubstancegt lactose cows lthasComponentgt cow milk  goats lthasComponentgt goat milk ewes lthasComponentgt ewe milk goat milk ltcontainsSubstancegt goat cheese ewe milk ltcontainsSubstancegt ewe cheese blood ltcontainsSubstancegt blood protein blood ltcontainsSubstancegt blood lipids
22
The rules as you go approachExploit patterns to
automate the conversion process

• 1.   Editor
• acid soils BT chemical soil types ? acid soils
  ltisagt chemical soil types
• Pattern X BT type ? X ltisagt type
• Therefore
• acrisols BT genetic soil types ? acrisols ltisagt
  genetic soil types
• Result

23
Undifferentiated relationships from AGROVOC Edited relationships
milk NT cow milk milk NT goat milk milk NT buffalo milk milk NT milk fat milk RT milk protein milk RT lactose cow RT cow milk  goats RT goat milk ewes RT ewe milk goat milk RT goat cheese ewe milk RT ewe cheese acid soils BT chemical soil types acrisols BT genetic soil types alkaline soils BT chemical soil types aluvial soils BT lithological soil types chemical soil types BT soil types Cichorium BT Asteraceae Cichorium endivia BT Cichorium Cichorium intybus BT Cichorium Cichorium intybus RT coffee substitutes Cichorium intybus RT root vegetables blood NT blood protein blood NT blood lipids milk ltincludesSpecificgt cow milk milk ltincludesSpecificgt goat milk milk ltincludesSpecificgt buffalo milk milk ltcontainsSubstancegt milk fat milk ltcontainsSubstancegt milk protein milk ltcontainsSubstancegt lactose cows lthasComponentgt cow milk  goats lthasComponentgt goat milk ewes lthasComponentgt ewe milk goat milk ltcontainsSubstancegt goat cheese ewe milk ltcontainsSubstancegt ewe cheese acid soils ltisagt chemical soil types acrisols ltisagt genetic soil types alkaline soils ltisagt chemical soil types aluvial soils ltisagt lithological soil types chemical soil type ltisagt soil types blood ltcontainsSubstancegt blood protein blood ltcontainsSubstancegt blood lipids
24
The rules as you go approachExploit patterns to
automate the conversion process

• 1.   EditorCichorium BT Asteraceae ?
  Cichorium ltisagt Asteraceae
• Pattern Taxon BT Taxon ? Taxon ltisagt Taxon
• Therefore
• Cichorium endivia BT Cichorium ? Cichorium
  endivia ltisagt Cichorium
• Result

25
Undifferentiated relationships from AGROVOC Edited relationships
milk NT cow milk milk NT goat milk milk NT buffalo milk milk NT milk fat milk RT milk protein milk RT lactose cow RT cow milk  goats RT goat milk ewes RT ewe milk goat milk RT goat cheese ewe milk RT ewe cheese acid soils BT chemical soil types acrisols BT genetic soil types alkaline soils BT chemical soil types aluvial soils BT lithological soil types chemical soil types BT soil types Cichorium BT Asteraceae Cichorium endivia BT Cichorium Cichorium intybus BT Cichorium Cichorium intybus RT coffee substitutes Cichorium intybus RT root vegetables blood NT blood protein blood NT blood lipids milk ltincludesSpecificgt cow milk milk ltincludesSpecificgt goat milk milk ltincludesSpecificgt buffalo milk milk ltcontainsSubstancegt milk fat milk ltcontainsSubstancegt milk protein milk ltcontainsSubstancegt lactose cows lthasComponentgt cow milk  goats lthasComponentgt goat milk ewes lthasComponentgt ewe milk goat milk ltcontainsSubstancegt goat cheese ewe milk ltcontainsSubstancegt ewe cheese acid soils ltisagt chemical soil types acrisols ltisagt genetic soil types alkaline soils ltisagt chemical soil types aluvial soils ltisagt lithological soil types chemical soil type ltisagt soil types Cichorium ltisagt Asteraceae Cichorium endivia ltisagt Cichorium Cichorium intybus ltisagt Cichorium blood ltcontainsSubstancegt blood protein blood ltcontainsSubstancegt blood lipids
26
The rules as you go approachExploit patterns to
automate the conversion process

• 1.   EditorCichorium intybus RT coffee
  substitutes ? Cichorium intybus
  ltusedToMakegt coffee substitutes
• Pattern Taxon RT FoodProduct ? Taxon
  ltusedToMakegt FoodProduct
• ThereforeCichorium intybus RT root vegetables
• ? Cichorium intybus ltusedToMakegt root
  vegetables
• Result

27
Undifferentiated relationships from AGROVOC Edited relationships
milk NT cow milk milk NT goat milk milk NT buffalo milk milk NT milk fat milk RT milk protein milk RT lactose cow RT cow milk  goats RT goat milk ewes RT ewe milk goat milk RT goat cheese ewe milk RT ewe cheese acid soils BT chemical soil types acrisols BT genetic soil types alkaline soils BT chemical soil types aluvial soils BT lithological soil types chemical soil types BT soil types Cichorium BT Asteraceae Cichorium endivia BT Cichorium Cichorium intybus BT Cichorium Cichorium intybus RT coffee substitutes Cichorium intybus RT root vegetables blood NT blood protein blood NT blood lipids milk ltincludesSpecificgt cow milk milk ltincludesSpecificgt goat milk milk ltincludesSpecificgt buffalo milk milk ltcontainsSubstancegt milk fat milk ltcontainsSubstancegt milk protein milk ltcontainsSubstancegt lactose cows lthasComponentgt cow milk  goats lthasComponentgt goat milk ewes lthasComponentgt ewe milk goat milk ltcontainsSubstancegt goat cheese ewe milk ltcontainsSubstancegt ewe cheese acid soils ltisagt chemical soil types acrisols ltisagt genetic soil types alkaline soils ltisagt chemical soil types aluvial soils ltisagt lithological soil types chemical soil type ltisagt soil types Cichorium ltisagt Asteraceae Cichorium endivia ltisagt Cichorium Cichorium intybus ltisagt Cichorium Cichorium intybus ltusedToMakegt coffee substitutes Cichorium intybus ltusedToMakegt root vegetables blood ltcontainsSubstancegt blood protein blood ltcontainsSubstancegt blood lipids
28
The rules as you go approachDiscussion

• Main idea Formulate constraints to assist the
  editor
• Ontology may have many relationship types,
  perhaps gt 100
• Constraints limit the relationship types that are
  possible in a specific case show the editor only
  these
• If the constraints limit possible relationship
  types to 1, conversion is automatic
• Constraints may depend on Thesaurus to be
  converted

29
Constraints
Thesaurus Relationships Possible ontology relationships
NT / BT lthasMembergt    ltmemberOfgt ltincludesSpecificgt   ltisagt lthasComponentgt   ltcomponentOfgt ltspatiallyIncludesgt   ltspatiallyIncludedIngt etc.
RT ltsimilarTogt ltsimilarTogt ltgrowsIngt    ltEnvironmentForGrowinggt lttreatmentForgt   lttreatedWithgt  lthasMembergt   ltmemberOfgt lthasComponentgt   ltcomponentOfgt ltmadeFromgt ltusedToMakegt etc.
30
Constraints
Thesaurus Relationships entity types or values Possible ontology relationships
milk NT milk Substance NT Substance X BT type Taxon BT Taxon GeogrEntity  BT GeogrEntity BodyPart BT BodyPart ChemSubstance BT ChemSubstance milk ltincludesSpecificgt milk Substance ltcontainsSubstancegt Substance X ltisagt type Taxon ltisagt Taxon GeogrEntity  ltspatiallyIncludedIngt  GeogrEntity BodyPart ltisComponentOfgt BodyPart ChemSubstance ltisagt ChemSubstance
31
Constraints
Thesaurus Relationships entity types or values Possible ontology relationships
Substance RT Substance LivingOrganism RT BodyPart Taxon RT FoodProduct GeogrEntity  RT GeogrGrouping Process RT Object ChemSubstance RT Function Substance ltcontainsSubstancegt Substance Substance ltcontainedInSubstancegt Substance Substance ltusedToMakegt Substance Substance ltmadeFromgt Substance LivingOrganism lthasComponentgt BodyPart Taxon ltusedToMakegt FoodProduct GeogrEntity  ltisMemberOfgt  GeogrGrouping Process ltperformedByInstrumentgt Object Process ltaffectsgt Object ChemSubstance ltusedForgt Function
32
Checking by editor

• Relationship instances created by editor by
  selecting from a constraint-generated menuare
  final
• Relationship instances created automatically
  must be presented to the editor
• If the editor determines that the relationship
  instances are almost always correct, she checks a
  box accept without checking

33
Overall conversion process

• One master editor must go through the file from
  start to finish,processing the relationship
  instances and creating patterns,creating new
  relationship types as needed
• Assistant editors can apply the patterns.
• In the first pass, the master editor should deal
  with the easy cases.
• Deal with the remaining cases later.Groups of
  similar relationship instances can be seen more
  easily in a smaller set

34
Adding new relationship types and new
relationship instances

• AGROVOC does not contain all relationship types
  or relationship instances for AI applications
• Need to add data. For example
• Organism X  lthasPestgt Organism Y
• ChemSubstance X ltactsAgainstgt Organism Y
• Organism X ltactsAgainstgt Organism Y
• Plant X  ltgrowsIngt  Environment Y
• FoodProduct X ltsuitableForgt Diet Y

35
Conclusion

• The rules-as-you-go approach is a realistic
  method for developing a rich ontology from an
  existing thesaurus