CHAPTER-3 KNOWLEDGE REPRESENTATION. Types of Knowledge Knowledge Pyramid Knowledge Representation Methods

Size: px

Start display at page:

Download "CHAPTER-3 KNOWLEDGE REPRESENTATION. Types of Knowledge Knowledge Pyramid Knowledge Representation Methods"

Charlotte Perkins
6 years ago
Views:

1 CHAPTER-3 1 KNOWLEDGE REPRESENTATION Types of Knowledge Knowledge Pyramid Knowledge Representation Methods! Production Rules! Semantic Nets! Schemata and Frames! Logic 2 1

2 a) Definitions of Knowledge (1) the fact or condition of knowing something with familiarity gained through experience or association (2) acquaintance with or understanding of a science, art, or technique b) (1) the fact or condition of being aware of something (2) the range of one's information or understanding c) the circumstance or condition of apprehending truth or fact through reasoning : cognition d) the fact or condition of having information or of being learned 3 Types of Knowledge a priori knowledge (theoretical knowledge) comes before knowledge perceived through senses considered to be universally true (e.g., coin flip will give 50% heads and 50% tails) a posteriori knowledge (empirical knowledge) knowledge verifiable through the senses may not always be reliable (e.g., 100 coin flips give only 39 heads - what can you conclude?) procedural knowledge knowing how to do something declarative knowledge knowing that something is true or false tacit knowledge unconscious knowledge not easily expressed by language 4 2

3 Knowledge in Expert Systems Conventional Programming Algorithms + Data Structures = Programs Knowledge-Based Systems Knowledge + Inference = Expert System N. Wirth 5 Knowledge Pyramid Meta Knowledge Information Data Noise 6 3

4 Knowledge Representation Methods 1. Production Rules 2. Semantic Nets 3. Schemata and Frames 4. Logic 7 1. Production Rules Frequently used to formulate the knowledge in expert systems. A formal variation is Backus-Naur form (BNF). metalanguage for the definition of language syntax a grammar is a complete, unambiguous set of production rules for a specific language a parse tree is a graphic representation of a sentence in that language provides only a syntactic description of the language not all sentences make sense 8 4

5 Example: Production Rules (for a subset of the English language) Grammar <sentence> -> <subject> <verb> <object> <modifier> <subject> -> <noun> <object> -> <noun> Lexicon <noun> -> man cat water <verb> -> drinks walks <modifier> -> always sometimes 9 Example: Parse Tree Example sentence: Man drinks water always. <sentence> <subject> <verb> <object> <modifier> <noun> <noun> man drinks water always 10 5

6 Example in CLIPS: This program produces all possible sentences according to the production rules (i.e. grammar), although most of them will not be meaningful. (defrule sentence-rule (subj?s) (verb?v) (obj?o) (adverb?d) (assert (sentence?s?v?o?d))) (defrule subject-rule (noun?n) (assert (subj?n))) (defrule object-rule (noun?n) (assert (obj?n))) (deffacts lexicon (noun man) (noun cat) (noun water) (verb drinks) (verb walks) (adverb always) (adverb sometimes)) 11 CLIPS> (dribble-on) CLIPS> (reset) CLIPS> (run) CLIPS> (facts) f-0 (initial-fact) f-1 (noun man) f-2 (noun cat) f-3 (noun water) f-4 (verb drinks) f-5 (verb walks) f-6 (adverb always) f-7 (adverb sometimes) f-8 (subj water) f-9 (obj water) f-10 (sentence water drinks water sometimes) f-11 (sentence water drinks water always) f-12 (sentence water walks water sometimes) f-13 (sentence water walks water always) f-14 (subj cat) f-15 (sentence cat walks water sometimes) f-16 (sentence cat walks water always) f-17 (sentence cat drinks water sometimes) f-18 (sentence cat drinks water always) f-19 (obj cat) f-20 (sentence water drinks cat sometimes) f-21 (sentence water drinks cat always) f-22 (sentence water walks cat sometimes) f-23 (sentence water walks cat always) f-24 (sentence cat drinks cat sometimes) f-25 (sentence cat drinks cat always) f-26 (sentence cat walks cat sometimes) f-27 (sentence cat walks cat always) f-28 (subj man) f-29 (sentence man walks cat sometimes) f-30 (sentence man walks cat always) f-31 (sentence man walks water sometimes) f-32 (sentence man walks water always) f-33 (sentence man drinks cat sometimes) f-34 (sentence man drinks cat always) f-35 (sentence man drinks water sometimes) f-36 (sentence man drinks water always) f-37 (obj man) f-38 (sentence water drinks man sometimes) f-39 (sentence water drinks man always) f-40 (sentence water walks man sometimes) f-41 (sentence water walks man always) f-42 (sentence cat drinks man sometimes) f-43 (sentence cat drinks man always) f-44 (sentence cat walks man sometimes) f-45 (sentence cat walks man always) f-46 (sentence man drinks man sometimes) f-47 (sentence man drinks man always) f-48 (sentence man walks man sometimes) f-49 (sentence man walks man always) For a total of 50 facts. CLIPS> (dribble-off) 12 6

7 Advantages and Disadvantages of Production Rules Advantages: simple and easy to understand straightforward implementation formal foundations for some variants Disadvantages: simple implementations are very inefficient some types of knowledge are not easily expressed in such rules large sets of rules become difficult to understand and maintain Semantic Nets! graphical representation for propositional information (shallow knowledge structure)! originally developed by M. R. Quillian as a model for human memory! labeled, directed graph! nodes represent objects, concepts, or situations! labels indicate the name! nodes can be instances (individual objects) or classes (generic nodes)! links represent relationships! the relationships contain the structural information of the knowledge to be represented! the label indicates the type of the relationship 14 7

8 Semantic Net Example Cétautomatix buys-from Panoramix sells-to fights-with Ordralfabetix Abraracourcix is-boss-of is-a is-a is-a is-a Gaul AKO Human is-boss-of barks-at is-a Astérix liveswith is-a Dog is-a Idéfix is-friend-of takes-care-of Obélix 15 Example in CLIPS (deffacts initial-facts-of-semantic-net (is-a Asterix Gaul) (is-a Obelix Gaul) (is-a Abraracourcix Gaul) (is-a Cetautomatix Gaul) (is-a Panoramix Gaul) (is-a Ordralfabetix Gaul) (is-a Idefix Dog) (lives-with Dog Human) (AKO Gaul Human) (is-boss-of Abraracourcix Asterix) (is-boss-of Abraracourcix Cetautomatix) (is-friend-of Asterix Obelix) (buys-from Panoramix Cetautomatix) (sells-to Ordralfabetix Panoramix) (fights-with Cetautomatix Ordralfabetix) (takes-care-of Obelix Idefix) (barks-at Idefix Ordralfabetix) ) 16 8

9 (defrule humans (is-a?name Gaul) (assert (AKO?name Human))) (defrule mutual-friends (is-friend-of?name1?name2) (assert (is-friend-of?name2?name1))) (defrule mutual-fighting (fights-with?name1?name2) (assert (fights-with?name2?name1))) (defrule shopping1 (buys-from?name1?name2) (assert (sells-to?name2?name1))) (defrule shopping2 (sells-to?name1?name2) (assert (buys-from?name2?name1))) 17 Relationships without relationships, knowledge is an unrelated collection of facts reasoning about these facts is not very interesting inductive reasoning is possible relationships express structure in the collection of facts this allows the generation of meaningful new knowledge generation of new facts generation of new relationships 18 9

10 Types of Relationships relationships can be arbitrarily defined by the knowledge engineer allows great flexibility for reasoning, the inference mechanism must know how relationships can be used to generate new knowledge inference methods may have to be specified for every relationship frequently used relationships IS-A relates an instance (individual node) to a class (generic node) AKO (a-kind-of) relates one class (subclass) to another class (superclass) 19 Objects and Attributes attributes provide more detailed information on nodes in a semantic network often expressed as properties combination of attribute and value attributes can be expressed as relationships e.g. has-attribute 20 10

11 Implementation Questions simple and efficient representation schemes for semantic nets tables that list all objects and their properties tables or linked lists for relationships conversion into different representation methods predicate logic nodes correspond variables or constants links correspond to predicates propositional logic nodes and links have to be translated into propositional variables and properly combined with logical connectives 21 OAV-Triplets (Object-Attribute-Value triplets) can be used to characterize the knowledge in a semantic net quickly leads to huge tables Object Attribute Value Astérix profession warrior Obélix size extra large Idéfix size petite Panoramix wisdom infinite 22 11

12 (deftemplate OAV (slot Object) (slot Attribute) (multislot Value)) Method-1: (deffacts objects (OAV (Object Asterix) (Attribute profession) (Value warrior)) (OAV (Object Obelix) (Attribute size) (Value extra large)) (OAV (Object Idefix) (Attribute size) (Value petite)) (OAV (Object Panoramix) (Attribute wisdom) (Value infinite))) Method-2: (deffacts objects (Asterix profession warrior) (Obelix size "extra large") (Idefix size petite) (Panoramix wisdom infinite)) 23 Disadvantages of Semantic Nets expressiveness no internal structure of nodes relationships between multiple nodes no easy way to represent heuristic information extensions are possible, but cumbersome best suited for binary relationships efficiency may result in large sets of nodes and links search may lead to combinatorial explosion especially for queries with negative results usability lack of standards for link types naming of nodes classes, instances 24 12

13 Decision Trees A decision tree is a semantic net in which: Each node is connected to a set of possible answers. Each nonleaf node is connected to a test that splits its set of possible answers into subsets corresponding to different test results. Each branch carries a particular test result s subset to another node. Each leaf represents a possible answer. 25 Example-1 (Binary decision tree) Is the animal warm blooded? Yes No Does the animal purr? The animal is a snake Yes No The animal is a cat Yes Does the animal fly? No The animal is a bird The animal is a dog 26 13

14 SOLUTION-1 (Simple but inefficient) (deffunction ask-user (?question) (printout t?question " (y/n) " ) (bind?answer (read)) (while (and (neq?answer y) (neq?answer n)) do (printout t "(y/n)? " ) (bind?answer (read))) (return?answer)) (defrule start (assert (node1)) (defrule node1?n <- (node1) (retract?n) (if (eq (ask-user "Is the animal warm blooded?") n) then (printout t "The animal is a snake" crlf) else (assert (node2)) ) ) 27 (defrule node2?n <- (node2) (retract?n) (if (eq (ask-user "Does the animal purr?") y) then (printout t "The animal is a cat" crlf) else (assert (node3)) ) ) (defrule node3?n <- (node3) (retract?n) (if (eq (ask-user "Does the animal fly?") y) then (printout t "The animal is a bird" crlf) else (printout t "The animal is a dog" crlf) ) ) 28 14

15 SOLUTION-2 (Generalized and better) (deffunction ask-user (?question) (printout t?question " (y/n) " ) (bind?answer (read)) (while (and (neq?answer y) (neq?answer n)) do (printout t "(y/n)? " ) (bind?answer (read))) (return?answer)) ; "answer" nodes : ; (node <name> answer <value>) ; "decision" nodes : ; (node <name> decision <question> <yes-node> <no-node>) (deffacts decision-tree (node root decision "Is the animal warm blooded?" node1 node3) (node node1 decision "Does the animal purr?" node4 node2) (node node2 decision "Does the animal fly?" node5 node6) (node node3 answer "The animal is a snake") (node node4 answer "The animal is a cat") (node node5 answer "The animal is a bird") (node node6 answer "The animal is a dog")) 29 (defrule start (assert (current-node root))) (defrule do-decision-node?n <- (current-node?name) (node?name decision?question?yes-branch?no-branch) (retract?n) (if (eq (ask-user?question) y) then (assert (current-node?yes-branch)) else (assert (current-node?no-branch)) ) ) (defrule do-answer-node?n <- (current-node?name) (node?name answer?value) (retract?n) (printout t?value crlf)) 30 15

16 Example-2 Outlook Sunny Overcast Rain Humidity High Normal Play Outside True Windy False Don t Play Outside Play Outside Don t Play Outside Play Outside Schemata! suitable for the representation of more complex knowledge (deep knowledge structure)! causal relationships between a percept or action and its outcome! deeper knowledge than semantic networks! nodes can have an internal structure! related to the notion of records in databases 32 16

17 Concept Schema! abstraction that captures general/typical properties of objects! has the most important properties that one usually associates with an object of that type! may be dependent on task, context, background and capabilities of the user,! similar to stereotypes! makes reasoning simpler by concentrating on the essential aspects! may still require relationship-specific inference methods 33 Schema Examples the most frequently used instances of schemata are frames [Minsky 1975] scripts [Schank 1977] frames consist of a group of slots and fillers to define stereotypical objects scripts are time-ordered sequences of frames 34 17

18 Frames represents related knowledge about a subject provides default values for most slots frames are organized hierarchically allows the use of inheritance knowledge is usually organized according to cause and effect relationships slots can contain all kinds of items rules, facts, images, video, comments, debugging info, questions, hypotheses, other frames slots can also have procedural attachments procedures that are invoked in specific situations involving a particular slot on creation, modification, removal of the slot value 35 Simple Frame Example Slot Name name height weight profession armor intelligence marital status Filler Astérix small low warrior helmet very high presumed single 36 18

19 Overview of Frame Structure two basic elements: slots and facets (fillers, values, etc.); typically have parent and offspring slots used to establish a property inheritance hierarchy (e.g., specialization-of) descriptive slots contain declarative information or data (static knowledge) procedural attachments contain functions which can direct the reasoning process (dynamic knowledge) (e.g., "activate a certain rule if a value exceeds a given level") data-driven, event-driven ( bottom-up reasoning) expectation-drive or top-down reasoning pointers to related frames/scripts - can be used to transfer 37 control to a more appropriate frame Slots each slot contains one or more facets facets may take the following forms: values default used if there is not other value present range what kind of information can appear in the slot if-added procedural attachment which specifies an action to be taken when a value in the slot is added or modified (data-driven, event-driven or bottom-up reasoning) if-needed procedural attachment which triggers a procedure which goes out to get information which the slot doesn't have (expectation-driven; top-down reasoning) other may contain frames, rules, semantic networks, or other types of knowledge 38 19

20 Usage of Frames filling slots in frames can inherit the value directly can get a default value (these two are relatively inexpensive) can derive information through the attached procedures (or methods) that also take advantage of current context (slot-specific heuristics) filling in slots also confirms that frame or script is appropriate for this particular situation 39 Restaurant Frame Example generic template for restaurants different types default values script for a typical sequence of activities at a restaurant 40 20

21 Generic RESTAURANT Frame Specialization-of: a_kind_of Business-Establishment Types: range: (Cafeteria,Fast-Food, Seat-Yourself, Wait-To-Be-Seated) default: Generic Seat-Yourself Restaurant Frame if-needed: IF plastic-orange-counterthen Fast-Food, IF stack-of-trays THEN Cafeteria, IF wait-for-waitress-sign or reservations-made THE N Wait-To-Be-Seated, OTHER WISE Seat-Yourself. Location: range: an ADDRESS if-needed: (Look atthe MENU) Name: if-needed: (Look atthe MENU) Food-Style: range: (Burgers, Chinese, A merican, Seafood, French) default: American if-added: (Update Alternatives of Restaurant) Times-of-Operation: range: a Time-of-Day default: open eveningsexcept Mondays Payment-Form: range: (Cash, CreditCard, Check, Washing-Dishes-Script) Event-Sequence: default: Eat-at-Restaurant Script Alternatives: range: allrestaurants with same Foodstyle if-needed: (Find all Restaurants withthe same Foodstyle) 41 EAT-AT-RESTAUR A NT Script Restaurant Script Props: (Restaurant, Money, Food, Menu, Tables,Chairs) Roles: (Hungry-Persons, Wait-Persons, Chef-Persons) Point-of-View: Hungry-Persons Time-of-Occurrence: (Times-of-Operation of Restaurant) Place-of-Occurrence: (Location of Restaurant) Event-Sequence: first: Enter-Restaurant Script then: if(wait-to-be-seated-sign or Reservations) then Get-Attention Script then: Please-Be-Seated Script then: Order-Food-Script then: Eat-Food-Scriptunless(Long-Wait) when Exit-Restaurant-Angry Script then: if(food-quality was betterthan Palatable) then Compliments-To-The-Chef Script then: Pay-For-It-Script finally: Leave-RestaurantScript 42 21

22 Advantages of Frames fairly intuitive for many applications similar to human knowledge organization suitable for causal knowledge easier to understand than logic or rules very flexible 43 Disadvantages of Frames it is tempting to use frames as definitions of concepts not appropriate because there may be valid instances of a concept that do not fit the stereotype exceptions can be used to overcome this can get very messy inheritance not all properties of a class stereotype should be propagated to subclasses alteration of slots can have unintended consequences in subclasses 44 22

23 KR Languages and Programming Languages how is a knowledge representation language different from a programming language (e.g. Java, C++)? programming languages can be used to express facts and states what about "there is a pit in [2,2] or [3,1] (but we don't know for sure)" or "there is a wumpus in some square" programming languages are not expressive enough for situations with incomplete information we only know some possibilities which exist 45 KR Languages and Natural Languages how is a knowledge representation language different from natural language e.g. English, Spanish, German, natural languages are expressive, but have evolved to meet the needs of communication, rather than representation the meaning of a sentence depends on the sentence itself and on the context in which the sentence was spoken e.g. Look! sharing of knowledge is done without explicit representation of the knowledge itself ambiguous (e.g. small dogs and cats) 46 23

24 Good Knowledge Representation Languages combines the best of natural and formal languages: expressive concise unambiguous independent of context what you say today will still be interpretable tomorrow efficient the knowledge can be represented in a format that is suitable for computers practical inference procedures exist for the chosen format effective there is an inference procedure which can act on it to make new sentences 47 24

Knowledge Representation

7 Knowledge Representation 7.0 Issues in Knowledge Representation 7.1 A Brief History of AI Representational Systems 7.2 Conceptual Graphs: A Network Language 7.3 Alternatives to Explicit Representation