Knowledge Representation

Keywordsquick way for agents to locate potentially useful information Thesaurimore structured approach than keywords, arranging descriptive terms into broader, narrower, and related classification categories

Taxonomiesclassification structures practice and science of classification grouping Ontologiesricher way of structural and non-structural relationships than taxonomies. Ontologies provide more complete and precise domain

Relations Generalization & Inheritance – Generalization is the relationship between a class and a more refined class. – Inheritance is transferring properties to subclass. – Anti symmetric Aggregation – part-whole or part-of relationship, in which classes representing the components of something are associated with the class representing the entire assembly. – Transitive and anti symmetric Instantiation – Relationship b/w a class and each of the individuals that constitute that class

Frames vs. Descriptions FramesDirectly express knowledge in terms of graphs. Description family of languages that logicformally express certain constrains on KR. Precise semantic and axiomatization. concepts are computed from these descrptions

Ontology Language Features RDF DAML OIL (DAML + OIL) OWL UML

Elementary Algebra: Relations Two important hierarchies: – isA (generalization, inheritance) – isPartOf (aggregation) Relationships in ontologies are modeled as binary relations

A binary relation R b/w a set S d and a set S r relates zero or more members of S d with zero or more members of S r i.e. R can be modeled as a set of pairs, each of which consists of a member of S d and a member of S r. Since S d * S r refers to the Cartesian product of S d and S r, meaning the set of all possible pairs whose first component is drawn from S d and second from S r. i.e. R  S d X S r

Properties that indicate binary relationships occur in pairs, also indicate polarity of relationship. – If c1 is subclass of c2, then c2 is superclass of c1 – Students take courses. Courses are taken by students. Formally a binary relation R -1  S r X S d is defined as an inverse of R if and only if  d  S d, r  S r : (d,r)  R  (r,d)  R -1

Anti Symmetry: If x  y and y  x, then x=y. – In other words, we cannot have two distinct objects such that each precedes the other. Transitivity: if x  y and y  z, then x  z. – If x precedes y and y precedes z, then x precedes z.

Asymmetry:if x  y, then y  x. – Asymmetry is stronger from of anti symmetry, because it forbids two objects from preceding each other. Also an object cannot precede each other. Irreflexivity:x  x – This simply states that an object cannot precede it self.

Linearity:if x  y and y  x, then x=y. – This states that for any two distinct objects, one must precede the other. That is, for any two objects, the ordering relation must hold one way or the other.

Hierarchies Taxonomy – isA relationship Meronomy – isPartOf relationship

Structure of an Ontology Ontologies typically have two distinct components: Names for important concepts in the domain – Elephant is a concept whose members are a kind of animal – Herbivore is a concept whose members are exactly those animals who eat only plants or parts of plants – Adult_Elephant is a concept whose members are exactly those elephants whose age is greater than 20 years Background knowledge/constraints on the domain – Adult_Elephants weigh at least 2,000 kg – All Elephants are either African_Elephants or Indian_Elephants – No individual can be both a Herbivore and a Carnivore

Ontology Design and Deployment Given key role of ontologies in the Semantic Web, it will be essential to provide tools and services to help users: – Design and maintain high quality ontologies, e.g.: Meaningful — all named classes can have instances Correct — captured intuitions of domain experts Minimally redundant — no unintended synonyms Richly axiomatised — (sufficiently) detailed descriptions – Store (large numbers) of instances of ontology classes, e.g.: Annotations from web pages – Answer queries over ontology classes and instances, e.g.: Find more general/specific classes Retrieve annotations/pages matching a given description – Integrate and align multiple ontologies