Download presentation
Presentation is loading. Please wait.
Published byAbigayle Kelly Modified over 8 years ago
1
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge representation methods Knowledge bases, case bases, databases
2
INFO 629 Dr. R. Weber Copyright R. Weber Outline Data, information, knowledge Knowledge representation methods – characteristics, advantages, disadvantages Knowledge bases, case bases, databases – databases – case bases – knowledge bases (e.g., ontologies)
3
INFO 629 Dr. R. Weber Copyright R. Weber Data, information, knowledge data is raw facts information is processed, data that has a meaning knowledge is what you apply to make decisions and solve problems
4
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge knowledge is what you apply to make decisions and solve problems types of knowledge – declarative (classification) – procedural (association rules) – structural knowledge (relationships) human vs. computational knowledge Computational Knowledge
5
INFO 629 Dr. R. Weber Copyright R. Weber Decision Making and Problem Solving
6
gathering informationof alternate strategiesthe best strategyimplementmonitor
7
When is knowledge absolutely necessary ?? Can you perform it without using knowledge?
8
information knowledge information Can you perform it without using knowledge?
10
INFO 629 Dr. R. Weber Copyright R. Weber How do you represent knowledge? How do you represent computational knowledge?
11
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge representation formalisms rules frames cases, MOPs similarity measures semantic nets Bayesian nets intelligent agents trees, e.g. parsing, decision trees graphs conceptual graphs logic neural nets concepts, objects, facts taxonomies, ontologies
12
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge representation formalisms can be used with different inferencing methodologies or algorithms to perform intelligent (AI) tasks
13
INFO 629 Dr. R. Weber Copyright R. Weber Metrics for representation formalisms representational adequacy inferential adequacy inferential efficiency clear syntax and semantics naturalness
14
INFO 629 Dr. R. Weber Copyright R. Weber Logic truth preserving inference ability to recognize negation, ordering, disjunction and quantification precise and formal language to represent declarative knowledge represents semantics
15
INFO 629 Dr. R. Weber Copyright R. Weber Logic (Cawsey, 1998) “a logic is a formal system which may be described in terms of its syntax (what the allowable expressions are), its semantics (what they mean)and its proof theory(how can we draw new conclusions given some statements in the logic)”
16
INFO 629 Dr. R. Weber Copyright R. Weber Characteristics of Logic everything is well defined, not inspired by human reasoning Rules of inference –modus ponens, resolution rule logic-based methods are complete so they can prove hypotheses without doubt while propositional logic is simpler it is more limited and would require much more hand engineering to do the same
17
INFO 629 Dr. R. Weber Copyright R. Weber Concepts, Objects and Facts An object is a basic entity that can be instantiated. A fact is a statement that can be either true or false (Durkin, 1994). A concept tells something about the object. A concept can be: –An abstraction, such as a class of objects –An object associated with a valued attribute –It may be simpler to represent an abstraction as an attribute
18
INFO 629 Dr. R. Weber Copyright R. Weber (Production) Rules A logic sequence of an antecedent (premise, condition) and a consequence (conclusion, action), which represent facts. The antecedent attempts to verify if the fact is true or false, when the fact composing the antecedent is true, the conclusion is triggered. The antecedent can be composed of several facts connected through operators such as and, or, and not. Conclusions usually change or assign values to attributes of an object, call methods or trigger other rules.
19
INFO 629 Dr. R. Weber Copyright R. Weber Frames representation formalism commonly used in expert systems represents declarative, structural and procedural knowledge first introduced by M. Minski in 1975, “A frame is a data-structure for representing a stereotyped situation, like being in a certain kind of living room, or going to a child's birthday party. Attached to each frame are several kinds of information. Some of this information is about how to use the frame. Some is about what one can expect to happen next. Some is about what to do if these expectations are not confirmed”.
20
INFO 629 Dr. R. Weber Copyright R. Weber Characteristics of frames support inheritance (subclasses and instances) support methods –when needed –after changed –before changed easy to implement in different programming paradigms, logic-based or not
21
INFO 629 Dr. R. Weber Copyright R. Weber Cases, similarity functions A description of an experience can be used as a knowledge representation formalism Cases can be represented in formlike representations (e.g., job applications) Cases can also be represented through networks or nodes in a graph A case typically describes a pair wrt the task: problem-solution; text-interpretation; task-lesson Similarity functions represent how to assess similarity between two cases in a give problem
22
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge containers in case-based reasoning systems
23
INFO 629 Dr. R. Weber Copyright R. Weber Semantic Networks commonly used in logic-based expert systems directed graphs where (Quillian, 1968): nodes represent objects and concepts arcs represent relationships between objects and attributes used to represent static elements of a representation such as the class, the instances and its features
24
INFO 629 Dr. R. Weber Copyright R. Weber Conceptual graphs variety of semantic networks represent meaning (Sowa, 1984) (Cyre, 1997), “A conceptual graph is a finite, connected, bipartitioned graph consisting of a set of labeled concept nodes, a set of labeled conceptual relation nodes, and a set of (directed) arcs linking concept and relation nodes”.
25
INFO 629 Dr. R. Weber Copyright R. Weber Neural Networks inputs and outputs are represented numerically a matrix of weights learns the input/output behavior weights in the matrix are information the learned matrix (for facts in the same category as the inputs) represents knowledge
26
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge bases, case bases, databases Databases Case bases Knowledge bases: – rule-based ES – ontologies
27
INFO 629 Dr. R. Weber Copyright R. Weber Databases Characteristics: characters, fields, records, files Types: relational, hierarchical, network Advantages: – Fast, reduces redundancy, easy to update and maintain Disadvantages: – brittle, not amenable to inference, do not contain knowledge Potential applications and uses – data warehouses (lose ease of update & maintain)
28
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge bases Rule-based Case-based Ontologies
29
INFO 629 Dr. R. Weber Copyright R. Weber Knowledge base Definition from Durkin (1994): Part of an expert system that contains the domain knowledge. AI definition: Part of a knowledge-based system that contains knowledge to be used in reasoning. General definition: Repositories of any knowledge representation formalism that have the ability to perform AI tasks
30
INFO 629 Dr. R. Weber Copyright R. Weber Case bases Types of case bases – textual, numeric features, discrimination networks case bases vs. databases – where do they meet? is the case base the only knowledge base in a CBR system?
31
INFO 629 Dr. R. Weber Copyright R. Weber Ontologies
32
INFO 629 Dr. R. Weber Copyright R. Weber What are ontologies (in AI)? general view –a formalism that represents shared conceptualizations and their interrelations in a domain (or subdomain) using a common vocabulary “Ontologies are explicit specifications of conceptualizations.” most cited definition from Gruber (1993) Conceptualizations represent interpreted concepts
33
INFO 629 Dr. R. Weber Copyright R. Weber What are ontologies (in AI)? specific view –an ontology is an explicit description of: concepts (or classes) in a domain properties of each concept describing various features and attributes and restrictions on the attributes (facets)
34
INFO 629 Dr. R. Weber Copyright R. Weber shared, explicit, and conceptual consensual knowledge –not private to one individual, accepted by a group types and constraints are explicitly defined conceptual (abstract) model of a domain through its relevant concepts
35
INFO 629 Dr. R. Weber Copyright R. Weber Types of Ontologies Domain –Additional specializations are possible applications, tasks Linguistic –Account for grammar and meanings in a natural language e.g., WordNet for American English
36
INFO 629 Dr. R. Weber Copyright R. Weber Types of Information concepts, atomic types cardinality of constraints is-a hierarchy among concepts relationships between concepts taxonomies of relations reified statements axioms semantic entailments
37
INFO 629 Dr. R. Weber Copyright R. Weber Uses of domain ontologies interoperability among information systems semantic web: link, coordinate software agents sharing knowledge bases among KBS intelligent retrieval, search
38
INFO 629 Dr. R. Weber Copyright R. Weber Uses of domain ontologies Further reading: Weber, R. & Kaplan, R. (2003). Knowledge-based knowledge management. In Innovations in Knowledge Engineering, Editors: Ravi Jain, Ajith Abraham, Colette Faucher and Berend Jan van der Zwaag. International Series on Advanced Intelligence, Volume 4. July 2003. Advanced Knowledge International Pty Ltd. http://www.pages.drexel.edu/~rw37/weberkaplan.pdf
39
INFO 629 Dr. R. Weber Copyright R. Weber Ontology Editors (development environments) ONTOLINGUA http://ontolingua.nici.kun.nl WEBONTO* http://kmi.open.ac.uk/projects/webonto/ PROTEGEWIN http://smi-web.stanford.edu/projects/prot-nt/ ONTOSAURUS* http://www.isi.edu/isd/ontosaurus.html ODE KADS22
40
INFO 629 Dr. R. Weber Copyright R. Weber Further reading on ontology editors Duineveld, A.J., Stoter, R., Weiden, M.R., Kenepa, B. and Benjamins, V.R. (2000). WonderTools? A comparative study of ontological engineering tools. International Journal of Human-Computer Studies 52(6): 1111-1133.
41
INFO 629 Dr. R. Weber Copyright R. Weber looking at some ontologies kmi.open.ac.uk/projects/webonto Open University http://www.isi.edu/isd/ontosaurus.html USC/Information Sciences Institute
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.