Application of knowledge engineering methods conclusion. Knowledge engineering. The history of the term
The system is an intermediary, the conclusion of a contract for the supply.
Knowledge engineering is a field of informatics, within the framework of which research is carried out on the representation of knowledge in a computer, keeping it up to date and manipulating it.
Knowledge system - a system based on knowledge.
SOZ SBZ DBMS ES IS SII - artificial intelligence system.
The structure of a knowledge-based system.
KB solution mechanism
INTERFACE
A knowledge base is a model that represents in a computer the knowledge accumulated in a particular subject area. This knowledge must be formalized.
Knowledge is formed using a model and then represented using a specific language.
Knowledge about specific objects and rules are usually allocated in knowledge base. These rules are executed as a mechanism for obtaining decisions in order to derive new ones from the original facts.
The interface provides a dialogue in a language close to the user.
Methods based on the use of inference are often used in knowledge engineering.
The concept of the subject area.
An object is something that exists or is perceived as a separate entity.
Main properties: discreteness; difference.
When presenting knowledge, a pragmatic approach is used, i.e. those properties of the object that are important for solving the problems that the created system will solve are highlighted. Therefore, a knowledge-based system deals with things that are abstract objects. The subject acts as a carrier of some properties of the object. The state of the subject area can change over time. At each point in time, the state of the subject area is characterized by a set of objects and relationships. The state of the subject area is characterized by a situation.
Conceptual means of describing the subject area.
The conceptual model reflects the most general properties. In order to detail the description, languages are needed. The characteristic features of the conceptual means of describing the subject area are abstractness and universality. They can be used to describe any subject area.
The concept of an object class.
The concept of an object is the concept of sets. Similar objects are grouped into classes. At different times, different sets of objects can correspond to the same class.
K is the class of the object.
Kt is a set of objects of class K at time t.
Group (1999) = (IA-1-99, IA-1-98, … , IA-1-94, IB-1-99,…)
Group (1998) = (IA-1-98, IA-1-97, … , IA-1-93, IB-1-98,…)
(t Кt = ( … )
Teaching position = (Professor, Associate Professor, Senior Lecturer, Lecturer, Assistant)
1 4 Geometric figure, square shape, color blue.
(K: A1 K1, A2K2, ... , AnKn) name attribute name of classes of object classes attribute pair
Identification of objects can be direct and indirect. In the case of a direct line, the names of objects, ordinal numbers of objects are used; indirect is based on using the properties of objects.
The attribute can be a component. An attribute is understood as a property, characteristic, name of the components.
(Geometric figure: Shape Geometric Shape Color Color)
Attribute name and attribute value pairs are often the same.
Situation example:
(Lecture: Lecturer Lecturer's Last Name, Place #Audience_Number, Subject Topic_Name, Listener Group_Code, Day Week_Day, Time Start_Time)
Situation - shows the relationship between "teacher" and "listener", other characteristics of this situation.
The roles of the participants in the situation:
Listener
Characteristics of the situation:
(К: А1К1,А2К2, … , АnКn) – representation of knowledge in the form of some structure.
(date, day, day of month)
(date, month, month_name)
(date, year, year)
(geometric_figure, shape, geometric_shape)
(geometric_shape, color, color)
This representation of knowledge corresponds to the representation of knowledge in the form of individual facts.
(K: A1K1, A2K2, ..., AnKn)
Representations of knowledge about objects are divided into:
object classes (data structure)
knowledge about specific objects (about data)
Object classes.
1. (K: A1K1, A2K2, ..., AnKn)
AI - attribute name
Ki - object classes, are the value of the attribute
K - class name
(teachers:
full name surname_with_initials,
Position teaching_position)
(teacher, full name surname_with_initials, teacher, position teaching_position)
3. K (K1, K2, ... , Kn)
4. K (A1, A2, ... , An)
(teacher (surname_with_initials, teaching_position), teacher (full name, position))
Knowledge representation for the first form:
(K: A1K1, A2K2, ..., AnKn) ki (Ki
Attribute representation of knowledge:
(teacher: - represents
Full name Semenov - some structure
Position Associate Professor) - data
Knowledge representation for the second form:
(K: AiKi) to (K, ki (Ki
Attributive representation of knowledge in the form of individual facts:
(teacher1, full name, Semenov) - 1, 2 are links between
(teacher1, position, associate professor) - facts
(teacher2, full name, Petrov)
(teacher2, position, assistant)
Knowledge representation for the third form:
K (K1, K2, ... , Kn)
(teacher (Semenov, associate professor) - positional representation of knowledge
If there are no attribute names, and the attributes themselves are written at certain positions, then this is a positional representation of knowledge.
Representation of knowledge in the form of "triples" - (object, attribute, value).
Confidence coefficients are used to represent inaccurate values - (object, attribute, value, confidence coefficient).
0 - corresponds to uncertainty. a negative value is the degree of confidence in the impossibility of the attribute value.
(patient1, diagnosis, gastritis, K740)
* (patient, full name, Antonov, diagnosis of colitis K760, gastritis K740)
The representation of knowledge about the class of an object is called minimal if the removal of one of the attributes leads to the fact that the remaining set of attributes ceases to be a representation of this object class.
Lease(rent_object, tenant, landlord, lease_term, fee).
If you remove "rent_term", you get a sale, and if you remove
“term_rent” and “fee”, then you get a gift.
Representation of knowledge in a relational database.
Relational database − Data is stored in positional format.
The data is stored in the form of a table, where the table name is the name of the class.
Each class corresponds to a table or database file. The class name is the name of the corresponding table. Attribute Names - Corresponding Table Fields
(column). Table rows are database records. A record corresponds to a record in positional format.
| A1 | A2 | . . .|An |
| | |. . | |
| K1 | K2 | . . .|Kn |
| | |. . | |
teachers
| full name | position |
| Semenov | Associate Professor |
|Petrov |assistant|
The concept of an attribute in a positional database is preserved.
The record K (A1, A2, ..., An) is called the relationship between attributes. This terminology is used in a relational database. The idea of data in a relational database is based on the concept of "key".
A key is a set of relationship attributes whose value uniquely identifies a record in the file.
Flat
| city | street | house | building | apartment | area | number of rooms |
| Moscow | Tverskaya | 2 | 1 | 47 | 60 | 2 |
| Moscow | Tverskaya | 2 | 1 | 54 | 50 | 1 |
In this case, the key will consist of several fields.
Ki sup Kj is a subclass of the class sup subclass; subclass sup class.
Ki is a subclass of Kj if (t Ki t (Kj t
(If at any time t the class Ki is a subclass of Kj)
Npr - network classification.
The network classification is represented as a hierarchical structure.
Student sup learner.
Ki part of Kj - is part of Ki part Kj
Ki is part of Kj if a particular object of class Ki is part of a uniquely defined object Kj.
Ownership relationship. k isa K - is an element
Ki ius K - is a component
Means that an object of class K consists of objects of class K1, K2, ...,
Kn, and an object of class K may include several objects of class Ki.
Lecture number 4.
Relationship properties.
Partial order relations have the property of transitivity.
Ki sup Kj Kj sup Km
Ki part Kj Kj part Km
If the element is part of a block, and the block is composed of...
There are no cycles in the membership graph.
K1 ins K2, K2 ins K3,…,Kn-1 ins Kj
It is not true that Knins K1
Moscow isa city
City sup Locality
Moscow isa Locality
Operations on classes of objects.
Using operations on object classes, you can define a new object class
Ki set of blocks, for example, TVs
Material objects are divided into three classes
Condition (Premises (Equipment = Material object
Person (Room = Person (Equipment = Room (
Equipment =?
Placement of object classes
Person (Surname, First Name, Patronymic, Year_of_Birth, Gender)
Gender=(male, female)
male, female = human gender
K (K1, K2, K3, K4, K5)
KK5 - Partitioning a class by class K5.
The union of all these classes is man.
Man? Woman=Human
Male? Female=?
(Knowledge of foreign language
knowledgeable person,
Subject foreign_language)
As a result of partitioning, we get classes of people who know a foreign language.
A conceptual schema of a subject area is a set of classes of objects, relations and operations defined on it.
Template descriptions of the state of the subject area:
Classes K conducts classes in the discipline in a group in on in.
Ivanov I.I. conducts classes in the discipline of TOE in the IT-1-98 group on Monday for the 4th pair in G-301.
(classes: teacher Teacher discipline Discipline_name group Group_code day Week_day time Pair_number place Audience)
Conceptual Domain Models - A conceptual schema along with a set of propositions built on a finite set of templates.
Entity and Relationship Diagram (ER Diagram)
Entity Relationship Diagram
Essence
Entity and Relationship Attributes
N teachers work at 1 department. "*" - the sign of the teacher - you can find the department.
Communication verb or object
Attributes - adjective, numerators, dimensions, location
Load schedule
Logical systems (models), based on a single example of the delivery of goods to the store.
Logical models of knowledge representation.
The description of the subject area in one of the logical programming languages is based on the predicate calculus.
The language of multiple predicate calculus of the 1st order. Multiple logic of the 1st order.
To compose this language:
The concept of a sort corresponds to the concept of object classes.
Many sorts of S
On the set are given by functions. f-function name;
sorts of arguments;
B is the sort of function value.
Z-signature is the top level of knowledge representation in logical models.
Predicate -
T=(0;1)
false truth
-varietal constant B
Consider, as an example, the processing of parts in production
2-turning;
1-milling;
S=(Part, Machine, Operation, Part_Type, Machine_Type, Time)
1) det: Operation Detail; f A1 B
2) st: Operation (Machine;
3) Start: Operation (Time
4) con: operation (time
5) part_type: Part (part_type
6) st_type: Machine (Machine_type
7) 0: (Time
T: (Time
8) st_shaft:(Part_type shaft_seats: (Part_type
9) cutters: (machine_type current: (machine_type
10) cutter_end: operation T current_reverse: operation T
11) +: Time*Time Time
12): Time*Time T
Knowledge about specific objects
(lower level of knowledge representation) in the language of multiple predicate calculus is called the structure of the integrated signature
1) signature
2) Structure integr. Signatures.
3) For each sort name, a set of objects of this sort is created.
Detail = (det.1, det.2, det.3, det.4)
Machine = (st.1, st.2, st3)
Operation =(op1, op2, op3, op4, op5, op6, op7, op8)
part_type = (st_shaft, shaft_seats)
Machine_type = (current, cutters)
Time = (1,2,…,t)
The union of all sets is the universe.
Each function and predicate from the structure in the system corresponds to many factors.
1) det.(op.1)=det1 det.(op.2)=det1 det.(op.3)=det2
…………………..
2) st.(oper.1)= st.3 st.(oper.2)= st.1 st.(oper.3)= st.3
…………………
3) start(oper.1)=0 start(oper.2)=5 start(oper.3)=5
…………………..
4) end(op.1)=5 end(op.2)=12 end(op.3)=0
…………………
5) type_det(det.1)=st_shaft type_det(det.2)=shaft_seat type_det(det.3)=st_shaft type_det(det.4)=shaft_seat
………………….
6) type_st. (st.1)=current type_st. (st.2)=current type_st. (Art. 3) = cutters
………………….
10) cutter_end(oper1) current_return (oper2) cutter_end(oper3)
|operation|detail |machine |beginning |end |milling cutters |
| Oper1 | Det.1 | Article 3 | 0 | 5 | 1 | 0 |
| Oper2 | Det.1 | St.1 | 5 | 12 | 0 | 1 |
| Oper3 | Det.2 | Article 3 | 5 | 10 | 1 | 0 |
| Oper4 | Det.2 | Art.2 | 10 | 17 | 0 | 1 |
| Oper5 | Det.3 | Art.3 | 10 | 16 | 1 | 0 |
| Oper6 | Det.3 | Article 1 | 16 | 26 | 0 | 1 |
| Oper7 | Det.4 | Article 3 | 16 | 22 | 1 | 0 |
| Oper8 | Det.4 | Article 2 | 22 | 32 | 0 | 1 |
|Detail|DetailType |
| Det.1 | St_shaft |
| Det.2 | St_shaft |
| Det.3 | Shaft_places |
|Det.4 |Shaft_places|
|Machine|Type_st |
| Article 1 | Current. |
| Article 2 | Current. |
| Article 3 | Milling cutters. |
3) Component: Logic formulas
Rules for constructing formulas: a) a constant of type A, there is a term of type A b) a variable taking a value from type A, there is a term of type A c) if the signature contains a function - the constructed terms of varieties, respectively, then
-there is a term of sort B d)if the signature contains a predicate-
,terms of constructed varieties
, that is, an atom. e) if - terms of the same sort, then the expression , that is, the atom e) Atom is a well-formed formula (PPF) The variable included in the atom is free in this atom. g) if the constructed formula which freely includes variables x of sort A, then the expressions:
It is also a BPF, the variable “x” is linked (in new files) h) if the formulas are already built, then , is also a BPF
Examples:
1) Knowledge representation b=> oper2 performed on a lathe type_st(st(oper2))=ncurrent
2) Oper2 completed at stop 1 at station 1 start 5 end 12
3)
Lecture 8 12.11.99.
Resolution Method
The resolution method proves impossibility.
To use this method, it is necessary to reduce the original formula to DNF.
DNF:
- disjunction of letters pii - atom or negation of an atom.
Then the DNF is represented as a set of clauses
In the resolution method - there is one inference rule
As a result, from 2 clauses we get a new one, called a ruovent
- we get an empty clause, which is always false.
If a set contains an empty clause, then it is not satisfiable.
It turns out an empty clause, which proves that the given set is unsatisfiable.
The resolution method is applied until an empty disjunct is obtained.
m,n - const
substitution instead of a constant variable is unification.
In this case, we perform the substitution (n / y):
From (1) and (2) => a(x)c(x,n) (5)
From (3) and (5) , substituting (m/n)=> c(m,n) (6)
From (4) and (6) without substitutions => 0
The principle of resolutions in Prolog
Prolog uses Chordian clauses, i.e. clauses containing one letter without negation.
For example
=>
conjunction without negation
Clauses that do not contain characters at all can be used. is the target statement in the prologue: ? – a a: - b, c, d. b: - e,f. c. e. f.
?-a a(1) a(2) a(3)
| Step number | Target | Initial | resolution |
| | disjunct | disjunct | |
|1 |?-a. |a:-b,c,d. |-b,c,d. |
|2 |?-b,c,d |b:-e,f |-e,f,c,d |
|3 |?-e,f,c,d |e |-f,c,d |
|4 |?-f,c,d |f |-c,d |
|5 |?-c,d |c |-d |
|6 |?-d |d |0 |
Representation of the program in the form of a graph a: - b; c b: - d, e c: - g, f. e: - i, h g: - h, j d. f. h.
?-a
"," - And
";" - or
The construction of the graph begins with the target clause.
The graph shows which and how many solutions the problem under consideration has.
Two solutions to the problem
Production model of knowledge representation.
The basis for this model is the production rules, which have the following form
- production rule >:=
If then [KD=]
Examples:
Rule 5
If gender=female
And addition=small
And weight=65 years_or_more
Then relative_weight=variable
The confidence factor is determined by the number 0-100
Rule 27
IF perspective=excellent
And risk = high
TO factor=0 CD=10
In general, the premise can be a logical expression.
If the premise is true, then the conclusion is also true, i.e. in conclusion, any action that is performed if the premise is true can be indicated
::[AI…I]
::== object, attribute, value, confidence coefficient - representation of knowledge in the form of a quadruple
::==
:==KD=
The same object can have different values.
Multivalued objects are objects that can have multiple valid values.
If an object is not declared as multi-valued, then it can have several values, then they must not be valid, i.e. CD= 100
For objects, the value that is requested from the user.
What addition?
1. Small
2. Average allowed values
3. Large
What is the age
1. less than 25
2. from 25 to 55
3. more than 55
Sending confidence coefficient=min(Kdusl)
The fact obtained as a result of the rule execution prospect=excellent AC=50 risk=high AC=70 factor=0
The basic structure of the production model of knowledge representation
Initial data
Result
Lecture 9 (End)
|№ |Conflicting |Execution|Inferred|
|Step|Set | | |
| | rules | rules | fact |
|1 | | | |
|2 | | | |
|3 | | | |
|4 | | | |
|5 | | | |
The conclusion ends when the target vertex is reached, or there are no applicable rules left, and the target is not reached.
Reverse conclusions - performed from top to bottom (with conclusions orienting to the goal)
P 1 P2 P3 P4
P5
C 4 C5 C6 C7 C8
|№ |Goal|Conflict |Implementation|Subgoals|Fact|
|Step| | many | | | | |
| | | rules | rules | | | |
|1 |S1 |P6,P7 |P6 |S2,S3 | |
|2 |S2 |P1,P2 |P1 |S1,S5,S| |
|3 |С3 | | |3 |F1 |
| 4 | C4 | | | |F2 |
|5 |S5 |P3 |P3 | | |
|6 |С6 | | |С6,С7,С|F3 |
| 7 | C7 | | |8 |F4 |
| 8 | C8 | | | |F5 |
| | | | | | |
Goal - "duration" - the goal is given by the name of the object.
It is compared with the conclusion of the rules and the rule with the conclusion is selected.
, which contain the name of the object. We select the rule that contains the target object, we form a hypothesis
In the process, the hypothesis is either confirmed or refuted. The conclusions continue until either one is confirmed, or all possible hypotheses are exhausted.
Fewer checks are used because There are several conditions and one conclusion in a rule.
bidirectional conclusions.
First, direct conclusions are made, based on a small amount of data, as a result, a hypothesis is formed to confirm or refute other conclusions.
To check the conditions of the rules, the rule activation apparatus is used, which selects at each step those rules in which the conditions are checked.
Conditions must also be used. In the conditions of the rules, individual, and then general, are distinguished.
General rules - rules of conditions of applicability. Scope of applicability.
Generalized structure of the production rule.
(i); Q; P; A; =B; N
(i) - rule name:
Q is the scope of the rule;
P – rule applicability condition (logical condition)
A=>B is the core of the rule, where A is the premise and B is the conclusion;
N - the condition set, determines the actions that are performed if the kernel is executed.
R - if true, the core of the rule is activated.
Frame - a data structure for representing a stereotyped situation
(to: А1К1, A2K2, ….,AnKn)
(k: A1k1, A2k2,….,An kn)
(file name: slot1 name (slot1 value) slot2 name (slot2 value)
……………………………….. slot name n (slot value n))
Protoframe - knowledge about a class of objects.
A frame instance is obtained from a protoframe by filling slots with specific values.
The frame structure usually includes system slots. The slot system includes:
Slots define a parent frame, a slot that points to direct child frames.
As a system of slots: slots containing information about the creator of the program, about its modification.
The structure includes:
- inheritance pointer;
- data type pointer;
- demons, etc.
LANGUAGE FMS (FMS).
Inheritance pointers can be:
U - unique - unique
S - same - some
R - range - border indicator;
0 -override - ignore
U - in frames of different levels with the same names will be different.
S - value inheritance slots from higher-level slots with the same names
The value of the lower equation must lie within the limits defined in the upper equation.
R
Human
If the value is not set, then it is inherited from the slot of the upper equation, and if it is set, then the inheritance is ignored.
Lecture 11 3.12.99
Combination of Network and Frame Models in the OPS-5 Knowledge Representation System
This language has production rules and databases
::=({| }+)
()+ - May repeat several times
::=((value))
::= |
(substance class acid
Name
color colorless)
(Order - Tasks: Source, Leak Fencing)
What are the rules:
::=(P )
::={}+
::= | -
::= | |
::=((value>)+) |
# (Order of tasks)
([{ }+])
# (substance)
The pattern does not necessarily specify all the attributes of a given class, i.e. we can write
(substance class acid
Name) i.e. variable acid -thing will get value
::= ({ >}+)
The value of the corresponding element attribute of the memory operation must match one of the elements specified in this sheet, at least one.
These values are given by specific words.
# (Substance class acid
Color)
::= ({{{}+}}+)
The list of values can also be specified in the form of restrictions
# (Engine power ( 100 200))
(engine power 160)
:={}+
::=(make | remove | (modif
{} +)
# (P coordinate _a
(target state active
name coordinate)
If the target is able to coordinate and the order of tasks is not defined, then create
(Order of tasks) –>
(make target state active
name to order tasks)
(modif1 wait state))
The problem solving strategy is based on an explicit goal setting
Performance
1. comparison with memory elements as a result, a conflicting set of rules is formed
2. Selection of rules from the conflict set
3. Performing the actions specified in the conclusion of the rules
Runs until the goal is reached.
Acquisition of knowledge
Extracting knowledge from the source, converting it into the desired form, as well as transferring it to the knowledge base of an intelligent system.
Knowledge is divided into:
- objectified;
- subjective
Objectified - knowledge presented in external sources - books, magazines, research.
- formatted, i.e. presented in the form of laws, formulas, models, algorithms.
Subjective - knowledge that is expert and empirical is not presented in an external form.
The knowledge of an expert is informal, it is a set of heuristic techniques and rules, allows you to find approaches to solving problems and put forward hypotheses that can be confirmed or refuted.
Knowledge can be obtained in the process of observing any object.
Modes of work of a knowledge engineer, a consultant in the process of acquiring knowledge.
1. protocol analysis
- write down the reasoning aloud in the process of solving problems.
O.s. protocols are drawn up and reviewed
2. Interview - a dialogue is conducted with the experiment, aimed at acquiring knowledge.
3. Game imitation of professional activity.
Interviewing methods.
1. Chopping at the steps, links are highlighted that allow you to build hierarchical structures
2. Repertual lattice 3 concepts are offered and it is required to name the difference between the 2nd concept of the 3rd one. The expert is offered a couple of concepts and needs to name common properties => form classes.
The method of work of a conitologist on the formation of a field of knowledge
Includes 2 stages
1. preparatory
1.1. Clear preparation of the task that the system must solve
2. Acquaintance with litas
3. Selection of experts
4. Acquaintance of experts with a copy
5. Acquaintance of an expert with a popular technique for artificial intelligence
6. Formation from a copy of the field of knowledge
2. Main stage
1. pumping the field of knowledge in the mode
2. teamwork of a cosmetologist - analysis of the protocol, determination of relationships between concepts, preparing questions for an expert
3. Pumping up the field of knowledge - the task of questions to the expert
4. Formalization of the conceptual problem.
5. Checking the completeness of the model
If the model is incomplete, then the 2nd approximation is used.
Lecture 12 10.12. 99.
fuzzy sets
– product thickness small medium large
degree belongs
10 15 40 product thickness
- fuzzy set x - universal set
x - form a set of pairs A
- is called the membership function of a fuzzy set.
The values of the membership function for a particular element X is called
Membership degree
Carrier of fuzzy set
A normal fuzzy set is a set for which
fuzzy set
X - universal set
X - form a set of pairs A
: - is called the membership function of the fuzzy set.
The value of the membership function for a particular element X is called the degree of membership
- carrier of fuzzy set
&
A normal fuzzy set is a set for each
If you bring to the normal form => you need to divide all its values by
.
Let the membership function be given by an integer from 10 to 40
Define the concept of small thickness of the product.
| | | | | | | | x x
10 11 12 13 14 15 16 17 18
18
Operations on fuzzy sets
1. Union of fuzzy sets
2. Intersection of fuzzy sets
3. Complementing the fuzzy set
Beginning of lectures 12 and 13.
(A1,(A2,….,(An x1,x2,…,xn x1(X1 x2(X2 … xn(Xn
(A1 x(A2 x … x(An = ()
(x (x1,x2,…,xn) = min((A1 (x1), (A2 (x2)…(An (xn) )
(Ax(B = (,
, }
5. Raising a fuzzy set to a power.
(A2 = con((A) - concentration
(A0.5 = dil((A) – stretching
Methods for determining the membership function.
A little more than 2. From 0 to 5.
|x |0 |1 |2 |3 |4 |5 |
|n1 |- |- |- |10|8 |4 |
|n2 |10|10|10|- |2 |6 |
(A = n1 / (n1 + n2)
Ranking method.
Fuzzy variable.
(- the name of the fuzzy variable x is the area of its definition
(A is the meaning, the fuzzy set determines the semantics of the fuzzy variable
linguistic variable.
(- linguistic variable name
T - basic term set - forms the names of fuzzy variables
(rarely, sometimes, often), which are linguistic variables
X - carrier of linguistic meanings - domain of definition
G - syntactic procedure
M - semantic procedure
Syntactic procedure in the form of grammatical terms whose symbols form terms from terms of sets (and, or, not), type modifiers
(very, slightly, not, etc.)
(- frequency
T = (rarely, sometimes, often)
Often
Such terms, together with the original ones, form a derivative of the terms of the set.
Semantic procedures allow to rewrite thermo-fuzzy semantics.
M((1 or (2) = (A1 ((A2
((1, x1, (A1)
((2, x2, (A2)
M((1 and (2) = (A1 ((A2
M(very () = con ((A)
M(slightly () = dil ((A)
Scenario.
It is a class of frame models of knowledge representation, where knowledge about the sequence of actions and events typical of the subject area is presented in a generalized and structural form. Consider the stereotype of a causal scenario - it determines the sequence of actions necessary to achieve goals, this is a frame model.
(kcus name: slot name 1(slot value 1); slot name 2(slot value 2);
…slot name n(slot value n))
(kcus doer target doer premise key consequence system name)
The premise defines the actions that must be performed before the key action, necessary for its action. Consequence is the final action. The system name is script.
(kcus "putting out the fire": doer (S:) goal of the doer (C: "stopping the fire")
P11, P12 parcels (cus: "search for extinguishing agents" R1, "extinguishing vehicles")
K1 key (f: "use of extinguishing agents for a complete ceasefire") consequence (P: "ceasefire") system name (sys: cus*1))
R1 - to be earlier
(kcus "search for extinguishing agents": agent (S:) agent's goal (C: "finding extinguishing agents")
P121, P22 parcels (cus: “determining the coordinates of the location of extinguishing agents” R1, “moving to the location of extinguishing agents”)
K2 key (f: "seizure of extinguishing agents") consequence (P: "located at the location of extinguishing agents") system name (sys: cus*2))
(kcus "transportation of extinguishing agents to the place of fire": actor (S:) goal of the agent (C: "delivery of extinguishing agents to the place of fire")
P31, P32 parcels (cus: “presence of extinguishing agents” R1, “determining the coordinates of the fire site”)
K3 key (f: "moving to the fire site") consequence (P: "presence of extinguishing agents at the fire site") system name (sys: cus*3))
Replenishment of knowledge based on the scenario.
Sequencing:
D = cus: P11 R1 cus: P12 R1 K1 =
P21R1P22R1K2 P31R1P32R1K3
P21R1P22R1K2 R1 P31R1P32R1K3 R1 K1
The premises define the actions that must be performed before the key action, necessary for its action. Consequence final action. The system name of the script.
Replenishment of knowledge based on pseudophysical logics.
P1 - aircraft landing
P2 - ladder supply
P3 - exit of passengers from the aircraft
P4 - bus supply
P5 - arrival at the airport terminal
The structure of the text at the linguistic level is represented by the following formula:
TS = PR4dt&P1R3 10,(P2&P2R1P3&P4R3 2,(P5 t = 15 hours 20 minutes
PR4dt , P1R3 10,(P2 (P2R4 dt + 10
P1R3 10,(P2 (P1R1P2
P4R3 2,(P5 (P4R1P5
TS* = P1R1P2& P1R1P3& P2R1P3& P4R1P5
Models and methods of knowledge generalizations.
Generalization is understood as the process of obtaining knowledge that explains the existing facts, as well as being able to classify, explain and predict new facts. The initial data is represented by a training sample. Objects can be divided into classes. Depending on whether a priori divisions of objects into classes are specified or not, generalization models are divided into generalization models by samples and by classes.
(+ = (01+, 02+…0nj+) is a positive sample.
A negative sample can be set (- = (01-, 02-…0bj-)
It is required to find such a rule that allows to establish whether or not an object belongs to the class Kj.
In data generalization models, a sample is represented by a set of class objects. Generalization methods are divided into methods of generalization by features and structural-logical methods of generalization.
Z = (z1, z2, …, zr)
Zi = (zi1, zi2, …, zini)
The object is characterized by a set of feature values Qi = (z1j1, z2j2, …, zrjr).
Structural-logical methods of generalization are used to represent knowledge about objects that have an internal structure among structural-logical methods. Two directions can be put forward: inductive methods of normal calculus and methods of generalization on semantic networks.
Algorithm for generalization of concepts by features.
The rules for determining whether objects belong to a certain class are represented in a number of logical formulas whose elements are hij and functions ((((((((
Z = (z1, z2) (gender, age)
Z1 = (z11, z12) (m, f)
Z2 = (z21, z22, z23) (young, middle, old)
(j+ = (01+, 02+) (j- = (01-, 02-, 03-)
01+ = (z11, z21) 02+ = (z11, z22)
01- = (z11, z23) 02- = (z12, z21) 03- = (z12, z22)
&i hij - generalized conjunctive concept
0 = max(xij – 1/(i), where 0 is a criterion, xij is the frequency of occurrence of a certain feature value, (i is the number of features.
0 = 3/5 – 1/2 = 0.1
(j+ = (01+, 02+) (j- = (01-)
(-1+ = 0 (-1- = {02-, 03-}
-----------------------
Situation
static
Dynamic
Constant properties and relationships
states
sustainable
Temporary
Processes
(patient1, diagnosis, colitis, K760)
(patient1, diagnosis, gastritis, K740)
student
undefined object
Defined object
material object
intangible object
situation
space
room
equipment
students
administrative staff
teachers
service staff
graduate student
head of department
Methodist
Professor
assistant
laboratory assistant
space
province
locality
working
teacher
Department name
substitution
discipline
Group code
Knowledge Engineering
After designing the knowledge base and developing the appropriate software or choosing a ready-made solution that satisfies the requirements, it is necessary to form a knowledge base, filling it with knowledge about the subject area, and possibly necessary facts. The stage of forming the knowledge base of an intelligent system is critical for the functioning of the entire system as a whole, since if the implementation affects the possible limits of the system's capabilities, then the knowledge of the system determines its capabilities in solving problems of the subject area.
This stage of IIS development is called the stage of knowledge acquisition. The area of activity that studies the problems of forming and filling knowledge bases is called knowledge engineering.
knowledge engineering- a fairly young direction of artificial intelligence, which appeared when practical developers faced very non-trivial problems of the difficulty of "mining" and formalizing knowledge. In the first books on AI, these facts were usually only postulated, then serious research began to identify optimal knowledge discovery strategies. The research led to the emergence of the discipline of knowledge engineering, and the emergence of a new specialty - a knowledge engineer, or a specialist who has theoretical knowledge and practical methods for building intelligent systems, and most importantly, trained in the methods of forming knowledge bases of intelligent systems.
The subject knowledge of an intelligent system can be taken from many sources, but the main source of knowledge of modern IIS is a specialist expert in the subject area. A knowledge engineer usually acquires knowledge by interacting directly with an expert. This process consists of a long series of intense, systematic interviews, usually lasting for many months. During conversations, the knowledge engineer gives the expert to solve problems that are close to real and of the same type, the solution of which is guided by the created expert system. The knowledge engineer must work with an expert, observing how he solves specific problems. Rarely is an effective approach in which the expert is directly asked questions about his rules or methods for solving a particular class of problems in his area of competence. Experts usually have great difficulty in formulating such rules.
It seems that "experts" tend to formulate their conclusions and how they arrived at them in general terms that are too broad and vague for effective computer analysis. It is preferable that the computer operate at a more specific level, dealing with well-defined portions of input information that can be embedded in more complex judgments. But an expert rarely works at such a specific level. He quickly draws difficult conclusions without bothering to carefully analyze and formulate each step of his reasoning process. Portions of initial knowledge are implicit assumptions, and they are combined with each other so quickly that it is difficult for an expert to describe this process. When he analyzes a problem, he cannot easily describe each step, and he may not even have an idea of the individual steps that were taken to find the solution. He may attribute to intuition or call a premonition what turns out to be the result of a very complex logical process based on a huge amount of data held in his memory and rich practice. Subsequently, when explaining his conclusion or premonition, he will repeat only the main steps, often omitting most of the small steps that may have seemed obvious to him at the time. Knowing what to consider basic and relevant and not requiring further reassessment is what makes a specialist an "expert". This aspect of the expert's nature is somewhat unusual. In practice, it is called the paradox of expertise. In the context of the development of intelligent systems, we will call it the knowledge engineering paradox:
The more competent experts become, the less able they are to describe the knowledge they use to solve problems!
Even worse, research has shown that when experts try to explain how they came to a conclusion, they often build plausible lines of reasoning that bear little resemblance to their actual problem-solving behavior. This feature has two important consequences. Here is the first one:
Experts need outside help to clarify and make explicit their way of thinking and solving problems. This can be summarized as the following rule of thumb - don't be your own expert!
The second heuristic for the knowledge engineer is:
Don't believe everything the experts say!
Experienced knowledge engineers develop working hypotheses based on what the experts say, and test those hypotheses for correctness and consistency by challenging the experts to solve new problems that require putative (according to the hypothesis) knowledge. The knowledge engineer believes that he or she has obtained a legitimate expert rule, not because the expert vouches for the accuracy of the rule, but because the expert demonstrates the application of that rule in solving the problem.
Sometimes the behavior of experts looks more intuitive than rational. The fact is that a significant amount of knowledge is the most important prerequisite for the skill of an expert. However, an expert's knowledge is not just a random collection of facts. A large number of situation templates serve as pointers, helping the expert to access the right parts of his stock of knowledge in a fraction of a second. This ability to use templates to guide the process of interpretation and decision is perhaps a significant part of what we call physical intuition.
Moreover, an expert can apply completely different methods of work and techniques in different non-standard situations, and it is extremely difficult to consciously identify the criteria for using different techniques.
Knowledge engineering terminology
Field of knowledge- this is a conditional informal description of the main concepts and relationships between the concepts of the subject area, identified from the expert's knowledge system, in the form of a graph, diagram, table or text.
The field of knowledge, as the first step towards formalization, represents a model of knowledge about the subject area, in the form in which the expert was able to express it in some “own” language. We will not consider the problems of adequate language formation for describing the subject area in this course. A good source of specialized information on this topic is.
The knowledge field serves as a key tool for the subsequent construction of a domain model in the knowledge base.
For the name of the process of obtaining knowledge (the type of strategy chosen in this case), several terms have become widespread in the literature: acquisition, extraction, extraction, acquisition, identification, formation of knowledge. In the English-language specialized literature, two are mainly used: acquisition (acquisition) and elicitation (identification, extraction, establishment).
The term "acquisition" is interpreted either very broadly - then it includes the entire process of transferring knowledge from an expert to the knowledge base of the ES, or already as a way of automated building a knowledge base through a dialogue between an expert and a special program (in this case, the structure of the knowledge field is pre-laid into the program). In both cases, the term "acquisition" does not refer to the very mystery of extracting the structure of knowledge from the flow of information about the subject area. This process is described by the concept of "extraction".
Knowledge Extraction(knowledge elicitation) is a procedure for the interaction of an expert with a source of knowledge, as a result of which the process of reasoning of specialists when making a decision and the structure of their ideas about the subject area become clear.
Acquisition of knowledge(knowledge acquisition) - the process of filling the knowledge base with an expert using specialized software (direct interaction between the IIS and the expert).
The term knowledge formation has traditionally been assigned to an extremely promising and actively developing field of knowledge engineering, which deals with the development of models, methods and learning algorithms. It includes inductive models of knowledge formation and automatic generation of hypotheses based on training samples, learning by analogy, and other methods. These models make it possible to identify causal empirical dependencies in databases with incomplete information containing structured numerical and symbolic objects (often in conditions of incomplete information).
Formation of knowledge(machine learning) - the process of data analysis and the identification of hidden patterns using a special mathematical apparatus and software.
Methods for obtaining knowledge
Knowledge engineering offers certain methods (techniques, methods) of working with experts.
Classification of methods of working with experts
Communicative methods are understood as all types of contacts of a knowledge engineer with a living source of knowledge - an expert. Among these methods, there are two large groups: active and passive.
Passive methods imply that the leading role in the knowledge extraction procedure belongs to the expert. In this case, the knowledge engineer mainly records the reasoning and actions of the expert.
In active methods, the initiative is entirely in the hands of the knowledge engineer. He conducts a conversation with an expert, offers various "Games", organizes a "round table", etc.
Passive methods are simple at first glance. At the same time, they require the knowledge engineer to be able to analyze the expert's "stream of consciousness" and highlight valuable fragments of knowledge in it.
Active methods are divided into two groups depending on the number of experts involved in the knowledge extraction procedure - In group methods, discussion between experts is of great importance, in which non-trivial aspects of knowledge are often revealed. At the same time, individual methods play a leading role today. To a large extent, this is due to the delicacy of the “knowledge taking” procedure.
Rice. 1.31. Classification of methods of working with experts
Passive Methods
Observations
The observation method is the only "pure" method, where the knowledge engineer does not interfere in the process of the expert's work and does not impose any of his own ideas on him. There are two types of observations:
· Observation of the real process.
· Monitoring process simulation.
First, the first type is usually used and the real process is observed at the expert's workplace. This helps to gain a deeper understanding of the subject area and note all the external features of the decision-making procedure necessary for the design of the user interface.
At the second stage, the expert imitates the process. In this mode, he is less stressed and works on "two fronts" - he conducts professional activities and at the same time demonstrates it.
Observation sessions impose the following requirements on the knowledge engineer:
· Proficiency in stenography.
· Acquaintance with timing techniques for a clear structuring of the production process in time.
· Developed skills of "reading the eyes", that is, observation of gestures, facial expressions and other non-verbal components of communication.
· Preliminary knowledge of the subject area.
The protocols of observations after the sessions are carefully deciphered, and then discussed with the expert.
Analysis of "thinking out loud" protocols
When recording “thoughts out loud”, the expert is asked to reveal the entire chain of reasoning that explains his actions and decisions. With such recording, it is considered important to record not only the entire "stream of consciousness" of the expert, but even pauses and interjections in the expert's speech. This method is sometimes referred to as "verbal reporting".
When recording "thoughts aloud", the expert can express himself as brightly as possible. He is not constrained by anything, no one interferes with him, he seems to freely float in the stream of his own reasoning and conclusions, he can show off his erudition and demonstrate the depth of knowledge. For a large number of experts, this is the most pleasant and flattering way to extract knowledge.
At the same time, as noted above, not every specialist, even those who can deliver impressive monologues about their work, is able to formalize and structure reasoning. However, there are people who are prone to reflection, capable of constructive presentation of thoughts. Such people are a godsend for a knowledge engineer.
Lecturer's gift is rare. An experienced lecturer structures his knowledge and reasoning well. But it happens that some people have a lecture
for free, but are unaware of its presence. In any case, a knowledge engineer should try to puzzle an expert by preparing a lecture on a topic of interest. If an expert is able to overcome a specific psychological barrier and enter the image of a teacher, this can be very effective for solving the problem of extracting knowledge.
A good question from a knowledge engineer during a lecture is essential. Serious, deep and interesting questions, on the one hand, stimulate the creative imagination of the lecturer, and on the other hand, increase the authority of the knowledge engineer.
The method of extracting knowledge in the form of lectures, like all passive methods, is used at the beginning of a multi-stage procedure for extracting knowledge from the memory of an expert. It contributes to the rapid immersion of a knowledge engineer in the subject area.
Active individual methods
Questionnaire
The questionnaire is the most standardized method. Drawing up a questionnaire is a rather delicate and responsible moment. Here are some recommendations:
· The questionnaire should not be monotonous and monotonous, so as not to cause boredom and fatigue. To do this, the questions should vary, the topics should change. In addition, special joke questions and game questions are often inserted into the questionnaire;
· the questionnaire should be adapted to the language of the experts;
It should be taken into account that the questions influence each other. Therefore, the sequence of questions should be well thought out;
The questionnaire must have “good manners”. It must be expressed in clear, understandable and extremely polite language. The methodical skill of compiling a questionnaire can only be mastered through practice.
The questionnaire procedure can be carried out in two ways. In the first, the analyst asks questions aloud and fills out the questionnaire himself based on the expert's answers. In the second, the expert fills out the questionnaire independently after preliminary instruction.
The choice of method depends on a number of conditions (in particular, on the design of the questionnaire, its clarity, and the readiness of the expert). At the same time, the second method seems to be preferable, since the expert has unlimited time to think over questions and the so-called presence effect is reduced.
Interview
Before conducting an interview, it’s a good idea to ask yourself: “Are we good at asking questions?” Consider the classification of questions.
Fig.1.32. Question classification
Open the question designates a topic or subject, leaving the expert free in the form and content of the response.
At closed question, the expert is offered a set of answers, among which he must make a choice.
Closed questions are easier to process, but to a certain extent they "program" the expert's answer and "close" the course of his reasoning. Therefore, when writing an interview script, they usually alternate between open and closed questions and especially carefully think over the “menu” and the content of closed questions.
Private the question appeals to the individual experience of the expert. Personal questions usually activate the expert's thinking, "play" on his ego, decorate the interview.
Impersonal the question is aimed at identifying the most common and generally accepted patterns of the subject area.
When preparing questions, it is taken into account that the language capabilities of an expert are, as a rule, limited. In addition, they keep in mind that due to isolation, stiffness and timidity, individual experts cannot immediately express their opinion and provide the required knowledge. Therefore, not direct questions are often used, which directly indicate the subject or topic, but indirect ones, indirectly directing attention to the actual problem. Sometimes, in the interests of the case, you have to ask several indirect questions instead of one direct one.
Verbal questions are traditional oral questions.
Questions using visual material diversify the interview and reduce the fatigue of the expert. Photographs, drawings and cards are used as visual material.
Dividing questions by function into basic, probing and control due to the fact that often the expert, for some reason, goes away from the question and the main questions of the interview turn out to be unproductive. Then the analyst uses probing questions that focus the expert's attention in the right direction. Control questions are used to check the reliability and objectivity of the information received.
Neutral questions are non-partisan. In the same time, suggestive questions force the expert to listen or even take into account the position of the interviewer.
In addition to those listed in the classification, it is useful to distinguish and include in the interview the following types of questions:
contact ("breaking the ice" between the analyst and the expert); o Buffer (to distinguish between different interview topics);
· revitalizing the memory of experts (for the reconstruction of individual cases from practice);
Provocative (to get spontaneous, unprepared answers).
Free dialogue
With a free dialogue between a knowledge engineer and an expert, there is no regulated plan. However, this form of knowledge extraction requires the most serious preliminary preparation.
Rice. 1.33. Scheme of preparation for an interview and free dialogue
Qualified preparation for dialogue is genuine dramaturgy. Her scenario provides for a smooth development of the procedure for extracting knowledge from a pleasant impression at the beginning of a conversation to a professional contact through awakening interest and gaining the trust of an expert.
To ensure the expert’s desire to continue the conversation, “stroking” is usually performed such as: “I understand you ...”, “... This is very interesting”, etc. At the same time, the analyst’s behavior must be sincere, because it has long been known that the best trick - avoid any tricks and treat the interlocutor with true respect and real interest.
There is a catalog of properties of the ideal interviewer: "He should look healthy, calm, confident, inspire confidence, be sincere, cheerful, show interest in the conversation, be neatly dressed, well-groomed."
Games with an expert
In games with an expert, the knowledge engineer takes on a role in the simulated situation. For example, it can be the role of the Student, who, in front of the expert (Teacher), who corrects the Student, performs work on a given topic. This game is a good way to get a shy expert to talk.
Another example is the game of Specialist (knowledge engineer) and Consultant (expert). This game gives sometimes impressive results. Let's take an example from classical literature. The expert acted as a doctor who knows the patient well, and the consultant asked questions and made a prediction about the appropriateness of using one or another type of treatment. Such a game made it possible to establish that only 30 questions are required for a successful forecast, while the original version of the questionnaire compiled by physicians contained 170 questions.
To reveal hidden layers of knowledge, a game is used in which a specialist makes predictions in professional situations and gives them justifications. Then, after a certain time, the specialist is presented with his own justifications and asked to make forecasts on them. As it turns out, such a simple technique often makes it possible to detect missed steps in the expert's reasoning.
In the context focus game, the expert plays the role of the expert system, while the knowledge engineer plays the role of the user. The consultation situation is modeled. The expert's first reactions center around the most significant concepts and the most important aspects of the problem.
In general, for games with an expert, the following main tips are given to a knowledge engineer:
· Play bolder, invent games yourself.
· Do not impose the game on an expert if he is not located.
· Don't "push" the expert, don't forget the goals of the game.
· Do not forget about the time and that the game is tiring for an expert.
· Play fun, unconventional.
Active group methods
Active group methods alone cannot serve as a source of more or less complete knowledge. They act as additional and serve as a good addition to individual methods of extracting knowledge, activating the thinking and behavior of experts.
"Round table"
The round table method involves an equal discussion of the problem of interest by several experts. The task of the discussion is to collectively, from different points of view, from different angles, explore the controversial problems of the subject area. For spice, representatives of various scientific fields and generations are invited to the round table. The number of participants in the discussion usually ranges from three to five or seven.
Before starting the discussion, the facilitator (knowledge engineer) needs to make sure that all participants correctly understand the problem. Then you need to set the rules and clearly formulate the topic.
In the course of the discussion, it is important to ensure that overly emotional and talkative experts do not change the topic and that criticism of each other's positions is justified. Some effort must be made by the facilitator to reduce the "façade effect" when the participants' desire to impress others prevails and they say something completely different from what they would say in a normal setting.
"Brainstorm"
"Brainstorming" or "brainstorming" is one of the most popular methods of liberating and activating human thinking. This method was first used in 1939 by A. Osborne in the USA to generate new ideas.
The main position of the assault is the separation of the procedure for generating ideas in a closed group of specialists from the process of their analysis and evaluation. The usual duration of the assault is about 40 minutes. The number of participants is up to 10 people. These participants are invited to express any thoughts on a given topic, including humorous, fantastic and erroneous. Criticism is prohibited. Time limit - up to 2 minutes per performance.
It is known from experience that the number of ideas expressed often exceeds 50. The most significant moment of the assault is the onset of a peak (hype), when ideas begin to literally “gush out”. Subsequent analysis by a group of outside experts typically shows that only 10-15% of the ideas are reasonable, but some of them are very original.
The art of the brainstorming knowledge engineer lies in the ability to ask questions of the audience, "warming up" the audience. Questions serve as a kind of “hook” by which ideas are extracted.
Role playing in a group
In each group game, a script is prepared in advance, roles are distributed, portraits-descriptions of roles are prepared, and a system for evaluating players is developed.
There are various ways to conduct role-playing games. In some games, participants come up with new names for themselves and perform under them. In others, all players switch to "you". In the third, the players choose the roles; in the fourth, lots are drawn to distribute the roles.
Usually, three to six experts take part in a game designed to gain knowledge: In the case of a larger number of experts, they are divided into groups between which a competition is organized: whose diagnosis will be closer to the true one, whose plan uses resources more rationally, who will quickly determine the malfunction in technical block, etc.
Creating a game environment requires imagination and invention from a knowledge engineer. The main thing is that the experts in the game immerse themselves in the situation as much as possible, really “play”, loosen up and “reveal their cards”.
Requirements and knowledge engineer
Concluding a concise review of what is actually a vast field of knowledge engineering, we note a number of basic requirements for a knowledge engineer.
· The knowledge engineer must have a good theoretical knowledge in the field of knowledge representation models in order to optimally select and use the capabilities of knowledge representation models to solve the problem.
· Although it has not been explicitly stated before, it is clear from the presentation of the material that a knowledge engineer must have excellent communication skills and have some knowledge in the field of communication psychology in order to work productively with experts.
· A knowledge engineer must have systems thinking and master the methods of analysis of the subject area, the methods of cognitive psychology.
· Have a comprehensive general scientific training and be fluent in the methods of scientific research, formal methods of description and documentation.
· Be fluent in the field of information technology.
The role of the knowledge engineer in the development of intelligent information systems is often key to the success of a system design. As a rule, knowledge engineers become specialists - software developers and analysts with the necessary skills and inclinations. In conclusion, we note that the role of a knowledge engineer is in many ways similar to the functions of analysts and implementation specialists in the development of conventional information systems.
Search strategies in POPs
Similar information.
17.2. Knowledge Extraction Practices
17.3. Structuring knowledge
The central problem in the creation of intelligent information technologies is the adequate display of knowledge of a specialist in computer memory. This led to the development of a new direction in computer science - knowledge engineering, which determines the ratio of human knowledge and its formalized (information) display in a computer. Knowledge engineering studies and develops issues related to the acquisition of knowledge, its analysis and formalization for further implementation in an intellectual system.
Purpose of the Chapter– give an overview of the main theoretical aspects of knowledge engineering and introduce some practical methods of work of knowledge engineers.
After studying the chapter, you should know:
Approaches to obtaining knowledge in the development of expert systems
Theoretical problems arising in the extraction of knowledge
Features of psychological and linguistic factors that need to be taken into account by a knowledge engineer
The influence of the philosophy of knowledge on the work of a knowledge engineer
Knowledge Engineer Techniques When Working with a Knowledge Source
Knowledge Extraction Methods
The essence of expert games
Methods for extracting knowledge from texts
Structuring the acquired knowledge
Formation of the conceptual and functional structure of the subject area
How knowledge is formalized and the knowledge base is formed
17.1. Theoretical aspects of obtaining knowledge
Knowledge Strategies
Psychological aspect
Linguistic aspect
Gnoseological aspect
KNOWLEDGE STRATEGIES
There are several strategies for gaining knowledge. The most common:
acquisition;
extraction;
formation.
Under acquiring knowledge is understood as a method of automated construction of a knowledge base through a dialogue between an expert and a special program (in this case, the knowledge structure is pre-programmed into the program). This strategy requires significant pre-development of the subject area. Knowledge acquisition systems do acquire ready-made pieces of knowledge according to the structures laid down by the system designers. Most of these tools are specifically focused on specific expert systems with a strictly defined subject area and knowledge representation model, i.e. are not universal. For example, the TEIRESIAS system, which has become the progenitor of all knowledge acquisition tools, is designed to replenish the knowledge base of the MYCIN system or its child branches built on the EMYCIN "shell" in the field of medical diagnostics using a production model representation knowledge.
Term knowledge extraction concerns direct live contact between the knowledge engineer and the source of knowledge. The authors tend to use this term as a more capacious and more accurately expressing the meaning of the procedure for transferring the competence of an expert through a knowledge engineer to the knowledge base of an expert system.
Term formsupknowledge has traditionally been assigned to an extremely promising and actively developing field of knowledge engineering, which is engaged in the development of models, methods and algorithms for data analysis for obtaining knowledge and learning. This area includes inductive models for generating hypotheses based on training samples, learning by analogy, and other methods.
Thus, we can distinguish three strategies for conducting the knowledge acquisition stage in the development of expert systems (Fig. 17.1).
Rice. 17.1. Three strategies for gaining knowledge
At the present stage of development of expert systems in our country, the knowledge extraction strategy seems to be the most relevant, since there are practically no industrial systems for the acquisition and formation of knowledge in the domestic software market.
Knowledge Extraction- this is a procedure for the interaction of an expert with a source of knowledge, as a result of which the process of reasoning of specialists when making a decision and the structure of their ideas about the subject area become clear.
Currently, most developers of expert systems note that the process of knowledge extraction remains the bottleneck in the construction of industrial systems.
The knowledge extraction process is a long and laborious procedure in which a knowledge engineer, armed with special knowledge in cognitive psychology, systems analysis, mathematical logic, etc., needs to recreate the domain model that experts use to make a decision. Often novice developers of expert systems, wishing to avoid this painful procedure, ask the question: can an expert himself extract knowledge from himself? For many reasons this is not desirable.
First, most of the knowledge of an expert is the result of numerous layers, stages of experience. And often knowing that BUT should IN, the expert does not realize that the chain of his reasoning was much longer, for example FROM D, D A, A IN, or BUT Q, Q R, R B.
Secondly, as was known to the ancients (recall Plato's Dialogues), thinking is dialogical. And therefore, the dialogue between a knowledge engineer and an expert is the most natural form of "unwinding" the labyrinths of an expert's memory, in which knowledge is stored, partly of a non-verbal nature, i.e. expressed not in the form of words, in the form of visual images, for example. It is in the process of explaining to the knowledge engineer that the expert puts clear verbal labels on these blurry associative images, i.e. verbalizes knowledge.
Thirdly, it is much more difficult for an expert to create a domain model due to the depth and vastness of information that he possesses. Numerous cause-and-effect relationships of a real subject area form a complex system, from which the “skeleton”, or main structure, is sometimes more accessible to an analyst who also owns a system methodology: Any model is a simplification, and it is easier to simplify with less knowledge of the details.
To understand the nature of knowledge extraction, we single out three main aspects of this procedure (Fig. 17.2): psychological, linguistic, epistemological, which are described in detail in.
Rice. 17.2. Basic aspects of knowledge extraction
PSYCHOLOGICAL ASPECT
Communication Model for Knowledge Extraction
Of the three highlighted aspects of knowledge extraction psychological is, apparently, the main one, since it determines the success and effectiveness of the interaction of a knowledge engineer (analyst) with the main source of knowledge - a professional expert. We single out the psychological aspect also because the extraction of knowledge occurs most often in the process of direct communication between the developers of the system.
The desire and ability to communicate can characterize the degree of professionalism of a knowledge engineer.
It is known that the loss of information during conversational communication is large (Fig. 17.3). In this regard, let us consider the problem of increasing the information content of communication between an analyst and an expert through the use of psychological knowledge.
Rice. 17.3. Loss of information during communication
We can offer the following structural model of communication when extracting knowledge:
participants in communication (partners);
means of communication (procedure);
the subject of communication (knowledge).
In accordance with this structure, we single out three "layers" of psychological problems that arise in the extraction of knowledge (Fig. 17.4), and consider them sequentially.
Rice. 17.4. The structure of the psychological aspect of knowledge extraction
contact layer
Almost all psychologists note that any collective process is influenced by the atmosphere that arises in a group of participants. There are experiments, the results of which undeniably say that the friendly atmosphere in the team affects the result more than the individual abilities of individual members of the group. It is especially important that the development team develops cooperative rather than competitive relationships. Cooperation is characterized by an atmosphere of cooperation, mutual assistance, interest in each other's success, i.e. the level of moral communication, and for a competitive type of relationship - an atmosphere of individualism and interpersonal rivalry (lower level of communication).
Unfortunately, it is impossible to predict compatibility in communication with a 100% guarantee. However, a number of personality traits, character and other features of the participants in communication can be distinguished, undoubtedly affecting the effectiveness of the procedure. Knowledge of these psychological patterns is part of the baggage of psychological culture, which a knowledge engineer must have in order to successfully carry out the knowledge extraction stage:
goodwill and friendliness;
sense of humor;
good memory and attention;
observation;
imagination and impressionability;
great concentration and perseverance;
sociability and resourcefulness;
analyticity;
prepossessing appearance and manner of dressing;
self-confidence.
procedural layer
A knowledge engineer who has successfully mastered the science of trust and mutual understanding with an expert (contact layer) must still be able to take advantage of the beneficial effects of this atmosphere. The problems of the procedural layer relate to the conduct of the knowledge extraction procedure itself. There is little insight and charm useful for solving the problem of contact, professional knowledge is needed here.
Let us dwell on the general patterns of the procedure.
A conversation with an expert is best done in a small tête-à-tête room. Lighting, warmth, comfort directly affect the mood. Tea or coffee will create a friendly atmosphere. The American psychologist I. Atvater believes that the most favorable distance for business communication is from 1.2 to 3 m. The minimum "comfortable" distance can be considered 0.7 - 0.8 m.
Reconstruction of one's own reasoning is not easy work, and therefore the duration of one session usually does not exceed 1.5 - 2 hours. It is better to choose these two hours in the first half of the day (for example, from 10 to 12 hours). It is known that the mutual fatigue of partners during a conversation usually occurs after 20-25 minutes, so pauses are needed in the session.
Every knowledge engineer has their own unique way of speaking. Some speak quickly, others slowly; some are loud, others are quiet, etc. It is almost impossible to change the style of conversation - it is laid in a person in early childhood. However, knowledge extraction is a professional conversation, and the length of phrases spoken by the knowledge engineer also affects its success.
This fact was established by American scientists - linguist Yngve and psychologist Miller. It turned out that a person best perceives sentences with a depth (or length) of 7 plus or minus 2 words. This number (7+2) is called the Yngve-Miller number. It can be considered as a measure of "colloquial" speech.
No one doubts the need to fix the procedure for extracting knowledge. The question arises: in what form should this be done? There are three ways to record results:
writing down on paper directly in the course of the conversation (disadvantages - this often interferes with the conversation, in addition, it is difficult to have time to write down everything, even if you have shorthand skills);
a tape recording that helps the analyst analyze the entire course of the session and his mistakes (the disadvantage is that it can fetter the expert);
memorization with subsequent recording after the conversation (disadvantage - suitable only for analysts with a brilliant memory).
cognitive layer
Cognitive psychology (English cognition - knowledge) studies the mechanisms by which a person learns the world around him.
Here are some tips for a knowledge engineer from the standpoint of cognitive psychology:
not to impose on the expert the representation model that is more understandable and natural to him (the analyst);
use different methods of working with an expert based on the condition that the method should approach the expert like a "key to a lock";
be clearly aware of the purpose of the extraction procedure or its main strategy, which can be defined as the identification of the main concepts of the subject area and the relationships connecting them;
often draw diagrams that reflect the expert's reasoning. This is due to the figurative representation of information in human memory.
The material presented above is closely related to the basics of psychological culture, which includes understanding and knowledge of oneself and other people; adequate self-esteem and evaluation of other people; self-regulation of the mental state. It is easier to master this culture with the help of specialists - psychologists, psychotherapists, but you can do it yourself with the help of books, at least popular ones, for example. In addition, successful overcoming of psychological failures is facilitated by mastering the basics of acting and participating in special classes in socio-psychological video training.
To conclude, here are some of the traditional psychological failures of the beginning analyst:
lack of contact between an expert and a knowledge engineer (due to the psychological characteristics of one or the other; errors in the procedure; the appearance of the "facade" effect, i.e. the expert's desire to "show himself");
lack of understanding (due to the "projection" effect, i.e. transferring the analyst's view to the views of the expert; or the "order" effect, i.e. focusing attention primarily on what is said at the beginning, etc.);
low effectiveness of interviews (weak motivation of the expert, i.e. lack of interest; or unsuccessful pace of the conversation; or inappropriate form of questions; or unsatisfactory answers of the expert).
LINGUISTIC ASPECT
The structure of the linguistic aspect
Since the process of communication between a knowledge engineer and an expert is a language communication, consider linguistic aspect knowledge engineering. We single out three layers of linguistic problems important for knowledge engineering (Fig. 17.5).
Rice. 17.5. The structure of the linguistic aspect of knowledge extraction
Shared code problem
Most psychologists and linguists believe that language is the main means of thinking along with other "internal use" sign systems. The languages spoken and thought by the analyst and the expert can be very different.
So, we are interested in two languages - analytic language, consisting of three components:
terms of the subject area, which he learned from the special literature during the preparation period;
general scientific terminology from his "theoretical baggage";
everyday colloquial language used by the analyst;
and language expert, consisting:
from the special terminology accepted in the subject area;
general scientific terminology; everyday language;
neologisms created by an expert during his work (his professional jargon).
If we assume that everyday and general scientific languages are approximately the same for two participants in communication, then some common language, or code that partners need to develop for successful interaction, will consist of the flows shown in Fig. 17.6. In the future, this general code is transformed into a certain conceptual (semantic) network, which is a prototype of the knowledge field of the subject area.
Rice. 17.6. Scheme for obtaining a common code
The development of a common code begins with the analyst writing down all the terms used by the expert and clarifying their meaning. In fact, this is the compilation of a dictionary of the subject area. Then follows the grouping of terms and the choice of synonyms (words that mean the same thing). The development of a common code ends with the compilation of a dictionary of domain terms with their preliminary grouping by meaning, i.e. by conceptual proximity (this is already the first step in structuring knowledge).
Rice. 17.7 gives an idea of the ambiguity in the interpretation of terms by two specialists. In semiotics, the science of sign systems, the problem of interpretation is one of the central ones. Interpretation connects the "sign" and the "signified object". It is only in interpretation that the sign gains meaning. So, in fig. 17.7 the words "instrument X" for an expert mean some specific scheme that corresponds to the scheme of the original instrument, and in the head of a novice analyst the words "instrument X" evoke an empty image or some kind of "black box" with handles.
Rice. 17.7. Ambiguity of interpretation problem
Conceptual structure
Most specialists in artificial intelligence and cognitive psychology believe that the main feature of natural intelligence and memory in particular is the connection of all concepts into a network. Therefore, to develop a knowledge base, you need not a dictionary, but an encyclopedia in which all terms are explained in dictionary entries with links to other terms.
Thus, the linguistic work of a knowledge engineer on a given layer of problems is to build such connected fragments by "stitching" terms. With the careful work of an analyst and an expert, a hierarchy of concepts begins to appear in conceptual structures, which is generally consistent with the results of cognitive psychology.
Hierarchy of concepts is a global scheme that can be the basis for a conceptual analysis of the knowledge structure of any subject area.
It should be emphasized that the work of compiling a dictionary and conceptual structure requires a linguistic "feel", the ease of manipulating terms and a rich vocabulary of a knowledge engineer, since often the analyst is forced to develop a vocabulary of features on his own. The richer and more expressive the common code, the more complete the knowledge base.
The analyst is forced to remember all the time the difficulty of conveying images and ideas in verbal form. Often a knowledge engineer has to suggest words and expressions to an expert.
User Dictionary
Linguistic results, correlated with the layers of the common code and conceptual structure, are aimed at creating an adequate knowledge base. However, we should not forget that the professional level of the end user may not allow him to use the special language of the subject area in full. To develop the user interface, additional refinement of the common code dictionary is required, adjusted for the accessibility and "transparency" of the system.
In conclusion, we list the characteristic linguistic failures that lie in wait for a novice knowledge engineer:
conversation in different languages (due to poor training of a knowledge engineer);
mismatch with context and inadequate interpretation of terms (due to lack of feedback, i.e. too independent work of a knowledge engineer);
lack of differences between the common code and the user's language (differences in the level of knowledge of the expert and the user are not taken into account).
GNOSEOLOGICAL ASPECT
The essence of the epistemological aspect
Epistemology- This is a branch of philosophy associated with the theory of knowledge, or the theory of the reflection of reality in the human mind.
Knowledge engineering as a science, so to speak, is doubly epistemological - reality (O) is first reflected in the mind of an expert (M 1), and then the activity and experience of the expert are interpreted by the mind of a knowledge engineer (M 2), which already serves as the basis for constructing the third interpretations (P z) - knowledge fields of the expert system (Fig. 17.8). The process of cognition is essentially aimed at creating an internal representation of the surrounding world in the human mind.
Rice. 17.8. The epistemological aspect of knowledge extraction
In the process of knowledge extraction, the analyst is mainly interested in the knowledge component associated with non-canonical individual knowledge of experts, since subject areas with this type of knowledge are considered the most susceptible to the introduction of expert systems. These areas are usually called empirical, since they have accumulated a large amount of individual empirical facts and observations, while their theoretical generalization is a matter of the future.
Cognition is always associated with the creation of new concepts and theories. It is interesting that often the expert, as it were, "on the go" generates new knowledge, right in the context of a conversation with an analyst. Such generation of knowledge can also be useful to the expert himself, who up to that moment might not have been aware of a number of relationships and patterns of the subject area. The analyst, who is the "midwife" at the birth of new knowledge, can be helped here by the tools of system methodology, which allows using the known principles of the logic of scientific research, the conceptual hierarchy of science. This methodology forces him to see the general behind the particular, i.e. build chains:
FACT GENERALIZED FACT EMPIRICAL LAW THEORETICAL LAW
The knowledge engineer does not always reach the last link in this chain, but the very desire for movement can be extremely fruitful. This approach is fully consistent with the structure of knowledge itself, which has two levels:
empirical (observations, phenomena);
theoretical (laws, abstractions, generalizations).
Criteria of scientific knowledge
Theory is not only a coherent system of generalization of scientific knowledge, it is also some way of producing new knowledge. The main methodological criteria of scientificity, which make it possible to consider both the new knowledge itself and the method of obtaining it scientific, are:
internal consistency and consistency;
consistency;
objectivity;
historicism.
Internal Consistency. At first glance, this criterion simply does not work in empirical fields: in them, the facts often do not agree with each other, the definitions are contradictory, diffuse, and so on. An analyst who knows the peculiarities of empirical knowledge, its modality, inconsistency and incompleteness, has to smooth out these "roughness" of empiricism.
Modality of knowledge means the possibility of its existence in various categories, i.e. in the constructions of existence and obligation. Thus, some regularities are possible, others are obligatory, and so on. In addition, one has to distinguish between such shades of modality as: the expert knows that...; the expert thinks that...; the expert wants...; The expert believes...
Possible inconsistency empirical knowledge is a natural consequence of the basic laws of dialectics, and these contradictions do not always have to be resolved in the field of knowledge, but on the contrary, it is the contradictions that most often serve as the starting point in the reasoning of experts.
incompleteness knowledge is associated with the impossibility of a complete description of the subject area. The analyst's task is to limit this incompleteness to certain limits of "completeness", i.e. narrow the boundaries of the subject area, or introduce a number of restrictions and assumptions that simplify the problem.
Consistency. The system-structural approach to cognition (dating back to Hegel) orients the analyst to the consideration of any subject area from the standpoint of the laws of the systemic whole and the interaction of its constituent parts. Modern structuralism proceeds from a multi-level hierarchical organization of any object, i.e. all processes and phenomena can be considered as a set of smaller subsets (features, details) and, conversely, any objects can (and should) be considered as elements of higher classes of generalizations.
Objectivity. The process of cognition is deeply subjective; it essentially depends on the characteristics of the cognizing subject himself. Subjectivity begins already with the description of facts and increases as the idealization of objects deepens.
Therefore, it is more correct to speak about the depth of understanding than about the objectivity of knowledge. Understanding is co-creation, the process of interpreting an object from the point of view of the subject. This is a complex and ambiguous process that takes place in the depths of human consciousness and requires the mobilization of all the intellectual and emotional abilities of a person. The analyst must concentrate all his efforts on understanding the problem. Psychology confirms the fact that people who quickly and successfully solve intellectual problems spend most of their time understanding it, while those who quickly start looking for a solution most often cannot find it.
Historicism. This criterion is related to development. Knowledge of the present is knowledge of the past that gave birth to it. And although most expert systems give a "horizontal" slice of knowledge - without regard to time (in statics), a knowledge engineer must always consider processes taking into account temporary changes - both connection with the past and connection with the future. For example, the structure of the knowledge field and the knowledge base should allow adjustment and correction both during the development period and during the operation of the expert system.
The structure of knowledge
Having considered the main criteria for the scientific character of cognition, we will now try to describe its structure. The methodological structure of cognition can be represented as a sequence of stages (Fig. 17.9), which we will consider from the standpoint of a knowledge engineer.
Description and generalization of facts. This is like a "dry residue" of conversations between an analyst and an expert. Thoroughness and completeness of keeping records during the extraction process and punctual "homework" on them is the key to a productive first stage of cognition.
In practice, it turns out to be difficult to adhere to the principles of objectivity and consistency described above. Most often, at this stage, the facts are simply collected and, as it were, thrown into the "general bag"; an experienced knowledge engineer often immediately tries to find a “shelf” or “box” for each fact, thereby implicitly preparing for the conceptualization stage.
Rice. 17.9. The structure of knowledge
Establishing links and patterns. Links are established in the expert's head, though often implicitly; the task of the engineer is to reveal the framework of the expert's conclusions. Reconstructing the reasoning of an expert, a knowledge engineer can rely on the two most popular theories of thinking - logical and associative. At the same time, if the logical theory, thanks to ardent admirers in the person of mathematicians, is widely cited and exploited in every possible way in works on artificial intelligence, then the second, associative, is less known and popular, although it also has ancient roots. The beauty and harmony of the logical theory should not obscure the sad fact that a person rarely thinks in terms of mathematical logic.
The association theory presents thinking as a chain of ideas connected by common concepts. The main operations of such thinking are associations acquired on the basis of various connections; recalling past experiences; trial and error with occasional successes; habitual ("automatic") reactions, etc.
Building an idealized model. In order to build a model that reflects the subject's idea of the subject area, a specialized language is needed that can be used to describe and construct those idealized models of the world that arise in the process of thinking. This language is created gradually with the help of the categorical apparatus adopted in the relevant subject area, as well as the formal-sign means of mathematics and logic. For empirical subject areas, such a language has not yet been developed, and the field of knowledge that the analyst will describe in a semi-formalized way may be the first step towards creating such a language.
Explanation and prediction of models. This final stage of the structure of cognition is at the same time a partial criterion for the truth of the acquired knowledge. If the identified system of expert knowledge is complete and objective, then on its basis it is possible to make predictions and explain any phenomena from this subject area. Usually knowledge bases of expert systems suffer from fragmentation and modularity (disconnection) of components. All this does not allow creating truly intelligent systems that, being equal to a person, could predict new patterns and explain cases that are not explicitly indicated in the database. An exception here is the systems of knowledge formation, which are focused on the generation of new knowledge and "prediction".
In conclusion, we list the most common failures associated with the epistemological problems of knowledge engineering (partially from ):
fragmentation, fragmentation of knowledge (due to violations of the principle of consistency or errors in choosing the focus of attention);
inconsistency of knowledge (due to the natural inconsistency of nature and society, the incompleteness of the extracted knowledge, the incompetence of an expert);
misclassification (due to an incorrect definition of the number of classes or an inaccurate description of the class);
an erroneous level of generalization (due to excessive detail or generalization of object classes).
An engineering discipline that deals with the integration of knowledge with computer systems in order to solve complex problems, usually requiring a high level of human expertise:
- knowledge configuration management (accounting);
- change management (evolution);
- logistics (search and delivery on demand).
At a high level, the knowledge engineering process consists of two:
- Knowledge Extraction- transformation of "raw knowledge" into organized, the process of obtaining knowledge from its sources, which can be material carriers (files, documents, books) and experts (groups of experts). It is part of Knowledge Engineering.
- Knowledge Implementation- transformation of organized knowledge into realized, the process of transforming organized knowledge into realized.
Knowledge Management Technologies
There are the following knowledge management technologies:
- working with implicit knowledge(tacit knowledge) in the minds of experts(most often they are meant when talking about "knowledge management"). Cognitive scientist (role):
- helps the expert to identify and structure the knowledge necessary for the operation of the expert system, extracts non-formalized knowledge from the expert;
- selects the intellectual system that is most suitable for a given problem area, and determines the way to represent knowledge in this IS;
- allocates and programs standard functions that will be used in the rules introduced by the expert.
- working with written knowledge("knowledge management" applies to computers: corporate knowledge management, Knowledge Management) - emphasis on "full-text search", "semantic search", "automatic annotation".
- NLP as a datalogical discipline ("form work"), swing technique, perceptual modalities, submodalities, spatial marking, calibration
- using web 2.0 (blogs and wikis)
- working with written formal knowledge(knowledge engineering, which is also included in knowledge management, but not so confident) - emphasis on structural databases, engineering models, data integration. Most technologies in knowledge engineering have gone the way of implementing the so-called "semantic network", the Husserl-Wittgenstein-Bunge approach that knowledge is representable by facts (and facts are relations of concepts). A semantic network emerges from a set of facts (see the review by John F. Sowa), in which relation-edges connect concepts-vertices. The implementation of the idea of storing and using knowledge in the form of a semantic was taken up by many almost non-overlapping parties / schools (community of practice), which is why a huge number of implementations and standards have appeared in which not a single word matches, but which are ideologically and technologically compatible.
- Data Modeling + Data Integration. Used when you need to combine data from multiple CAD systems from different vendors when building a large industrial facility. Keywords: ISO 15926 , Gellish , ISO 10303 . Instead of the word "ontology" they say "data model". : practically none, all requests for data. Everyone fights hand-to-hand with knowledge. No graphics, solid XML, proprietary storage formats Data schemas in every single CAD system. Recently, there are other solutions aimed at integrating heterogeneous data, for example from CYC and (based on the standardized UMBEL ontology, RDF expressions and access to data via HTTP, see). ISO 15926-7 projects come down to the same thing: some kind of ontology + semantic web standards.
- concept map() Used for (often collaborative via the web) educational and creative work. Key formats (all in XML): XCT 3.0, but ready to eat and Topic Map, and more to edit and display. Knowledge management tools: graphical display, combining networks that were drawn by two participants in the creative process. A close relative is MindMap , where it is not a graph at all, but a beautifully drawn tree, and the links are not named.
- Conceptual Graphs They use artificial intelligence, expert systems, agent systems and other classics of the genre for academic studies. Based on the work of the philosopher and logician Pierce ("intelligent indexing"), the key person is John F.Sowa. Key Knowledge Storage Format: three syntaxes, the main one being CGIF (XML). Knowledge management tools: Common Logic (or ISO ISO/IEC IS 24707:2007, ).
- Topic Map They will be used for Knowledge Management initiatives - and they came from catalogs (bibliographers). Big fans of standardization (see), but lost focus (they are inexorably attracted to general data modeling, in which they lose to Semantic Web approaches). Key knowledge storage formats: ISO 13250, XTM 2.0, HyTM. Knowledge management tools: topic map engine is used (a dozen options), because TMAPI 2.0 is standardized. In addition, a special standard for specifying constraints for topic maps - ISO / IEC FCD 19756 (TMCL) has entered the finish line, and the Topic Map Query Language (draft ISO 18048) seems to have died out.
Knowledge Engineering (KI) was defined by Feigenbaum and McCordack in 1983 as:
"IS is a branch (discipline) of engineering aimed at introducing knowledge into computer systems to solve complex problems that usually require rich human experience."
Currently, this also involves the creation and maintenance of such systems (Kendal, 2007). It is also closely related to software development and is used in many information studies, such as artificial intelligence studies, including databases, data collection, expert systems, decision support systems, and geographic information systems. CI is related to mathematical logic, also used in various scientific disciplines, for example, in sociology where people are “experimental”, and the goals of research are understanding how human logic works on the example of relationships in society.
Examples
An example of the operation of a system based on IS:
- Consideration of the problem
- Query to databases by task
- Entering and structuring the information received (IPK model)
- Creating a database of structured information
- Testing the information received
- Make adjustments and improve the system.
IS has practical applications. In the US, up to 90% of lending decisions for retail banking customers are made using expert systems based on FICO knowledge bases. A subsection of IS is knowledge metaengineering suitable for AI development.
Principles
Since the mid-1980s, IS has developed several principles, methods, and tools that have made it easier to acquire and work with knowledge. Here are some of the key ones:
Knowledge engineering uses knowledge structuring methods to speed up the process of obtaining and working with knowledge.
