**6.3.2 Knowledge acquisition phase: Learning**

Taking the problem and its modular decomposition, followed by the conceptualization of relevant knowledge, must define its scope. When knowledge has been collected, must be analyzed, coded and documented so that these activities take place according to the chosen acquisition technique. It also should be prioritized to represent in the KB. Later must be formalized to determine the method of acquisition. The acquisition stage can be difficult to the extent that knowledge is extracted directly from human experts. Once the execution is carried out, it should be the implementation, which includes programming and coding of knowledge in the computer, designing a prototype that allows refining and contrasting the results. Finally, the knowledge engineer test at the KB by means of examples (cases) and compares the results with experts to examine the validity of knowledge.

#### **6.3.2.1 Phase outcomes**

Knowledge Recovery Knowledge Bases Repositories Conceptual Map of the Knowledge or Syllabus Knowledge Structures.

#### **6.3.3 Knowledge representation phase: Ontology**

The representation of knowledge involves the consistent use of mathematical logic and information structures. The KE theories fall into two categories: mechanisms and content. Ontologies are content theories about classes of objects, their properties and their interrelationships, which allow the specification in the domain of knowledge. The representation has several roles:


#### **6.3.3.1 Phase outcomes**

56 New Research on Knowledge Management Models and Methods

The analysis of the results of the application of the requirements engineering is included. Also, for purposes of choosing the corresponding subsystem, in order to recompose the

Once chosen the subsystem, is analyzed functionally and, according to the priority of the modules and the requirements of users choose one of them for adding layers of knowledge.

The identification of the layers of the knowledge of the company (knowledge architecture).

Taking the problem and its modular decomposition, followed by the conceptualization of relevant knowledge, must define its scope. When knowledge has been collected, must be analyzed, coded and documented so that these activities take place according to the chosen acquisition technique. It also should be prioritized to represent in the KB. Later must be formalized to determine the method of acquisition. The acquisition stage can be difficult to the extent that knowledge is extracted directly from human experts. Once the execution is carried out, it should be the implementation, which includes programming and coding of knowledge in the computer, designing a prototype that allows refining and contrasting the results. Finally, the knowledge engineer test at the KB by means of examples (cases) and

The representation of knowledge involves the consistent use of mathematical logic and information structures. The KE theories fall into two categories: mechanisms and content. Ontologies are content theories about classes of objects, their properties and their interrelationships, which allow the specification in the domain of knowledge. The

compares the results with experts to examine the validity of knowledge.



**6.3.1.2 Determination of the subsystem to incorporate knowledge** 

The specification of the methodology, tools and techniques.

architecture of the involved knowledge.

Determination of intellectual assets.

A portfolio of the intelligent subsystems. The Intellectual capital resources. Scheduling and Time table. Quality Assurance Plan.

**6.3.2 Knowledge acquisition phase: Learning** 

Conceptual Map of the Knowledge or Syllabus

**6.3.3 Knowledge representation phase: Ontology** 

**6.3.1.4 Phase outcomes**  Classes of knowledge.

**6.3.2.1 Phase outcomes**  Knowledge Recovery Knowledge Bases Repositories

Knowledge Structures.

representation has several roles:

**6.3.1.3 Determination of the specific module** 

Knowledge Bases with fully reorganization Internal and External Networks Portfolio Capabilities: Assimilation, Strategic Technology and Innovation

#### **6.3.4 Knowledge acquisition methods**

Working with KE involves a trans-disciplinary team, in which the engineer is the intermediary between the KB and the Experts, refining and representing the KB. There are at least three types of methods for the acquisition of knowledge such as: manual, semicomputer based and automatic: artificial or machine learning.

#### **SELECTING THE APPROPRIATE METHOD**


#### **6.3.4.1 Phase outcomes**

Assessed and appropriate method Determination of appropriate computational resources Tutorial Online Help Determination of alternative sources of knowledge

#### **6.3.5 Problem solving solutions design**

The design of responses to problems or needs raised to be knowledge based is presented in several steps which are described as follows.

#### **6.3.5.1 Defining of potential solutions**

The first step, the knowledge domain organization related to the selected module, is to list all possible solutions, inputs, outputs, responses, alternatives and recommendations. It should identify the precise outputs to be presented to the user on the computer screen. The aim of this step is that the KB provides advice at the time required and not suddenly, a conclusion based on momentary must know every possible answer across time.

#### **6.3.5.2 Determining the knowledge related to the relevant parts of the module**

Knowledge Layers to be introduced should be raised at different levels of the organization. At this point is to be determined the operational knowledge, i.e. what is related to the application or specific system´s functionalities.

Analytical Models for Tertiary Education by Propaedeutic

**6.3.5.6 Identifying of meta-rules and frameworks** 

control over the other and the process of inference.

Selected Method for solving the problems identified.

also be important in determining the quality of the result. **6.3.6.1 Graphical representation of knowledge layers** 

taking into account the inheritance and relationships between them.

the model is represented which really needs.

adjustments necessary to refine the model.

**6.3.6.2 Inference proof** 

process was implemented.

Each conclusion should be in the THEN portion of the rule.

rules lead to the conclusion.

Define rules for each output.

the system.

**6.3.5.7 Phase outcome** 

Meta-cognition.

**6.3.6 Prototipyng** 

Problem solution domain.

Specifications accomplished.

Cycles Applying Knowledge Engineering and Knowledge Management 59

Need to identify questions to collect the necessary data from the IF portion to trigger

Should list all required data system called "facts" (or symptoms) the user will enter

 A decision tree is a good help if the elements of knowledge can be arranged quickly in a tree format. If so, proceed directly to the construction of a tree from them. Some of them are so large that KB may require a decision tree for the entire domain; however,

Once have determined the rules concerning the problem domain defined those who exercise

The first objective is to build a small prototype, for which selects a subset of the KB and carries to the KE tool, which must be done quickly. The result is a prototype which can quickly verify the implementation and testing and verifying ideas. In this sub-step is the representation of knowledge with the tool by means of a prototype, since this technique to identify weaknesses and strengths of the model developed, by which you can refine the results to get quality. A prototype is also a good way to test the concepts before investing in a larger program. With the use of a shell can quickly assemble a small prototype to determine if you are on the right track. It allows for demonstrations, as its assessment will

When rules and frameworks have led to the KB by using the tool must be plotted to assess the concatenation rules, relationships and inheritance between frames, i.e. to determine if

Elaboration of the rules diagram. There is a network diagram of the rules and actions for each of the layers of knowledge, which will have a broad overview of the KB and make the

Elaboration of the diagram of the frames. There are graphically represented all the objects,

From the KB, the inference process is made to assess the results based on the reasoning

By identifying, the frames: Select the objects on which the knowledge requires modeling. By determining meta-knowledge: Control over knowledge, or "knowledge of knowledge"

must be determined to have domain information and the inference process.

to simplify the work, you can build decision trees for subsets of the domain.

Domain knowledge module: determines the knowledge that the user needs to perform properly in the company.

Inferred knowledge: knowledge which is needed that can be inferred through the use of AI which supports the user.

Knowledge of processes and tasks of the module: identify the knowledge required for the development of system tasks, and relates to the basis of plans. These layers also serve as support for strategic planning and tactical levels, and are used in the plan generation and plan recognition.

#### **6.3.5.3 Determining of knowledge related for problem solving**

**Layers** of knowledge that are introduced should also support the management at the tactical and operational level, so should determine the knowledge that supports the resolution of problems.

Knowledge of **management skills**: includes the reasoning and decision-making DSS (Decision Support System), taking into account the user's management style in order to identify the subspace and data-information required by the executive for making their decisions in each application that contains the specific module. These layers of knowledge must be determined according to the administrative level support, i.e., differentiating strategic as tactical. Knowledge of **user interface**: where should the user's knowledge representation and should be considered cognitive style and managerial. In this respect it pursues a natural language interface to work with the system databases

#### **6.3.5.4 Tasks decomposition**

Tasks are evaluated regardless of the user model. At this point should answer:

*¿What tasks can be decomposed?* 

*¿Is there any difference between the user's perspective and that of the expert?* 

The second question is useful for determining if it is necessary to analyze and develop a comprehensive plan to draw conclusions. The tasks that are discussed are related to the tasks performed by the experts to solve their problems and do their job. These tasks should be decomposed into subtasks or sub-functions according to the natural order in which they develop.

To analyze the order of tasks: It should be a state diagram of the tasks previously listed, considering their order. With this diagram, it should do an analysis to determine if the order is correct, any extra work, or can be rearranged for performance optimizing.

Construction of the model: The model of a KB is useful to the extent as it constitutes a pattern of knowledge, easily interpretable, which determines how it will solve the problem in the KB. This model must be designed by determining the knowledge islands and regions.

Designing the architecture of the KB: Should design the hardware architecture of the KB, in which represents the different components and their relationships to have a "map" (e.g. semantic network or cognitive map) of the KB and facilitate construction. The conceptual design of the map gives the main idea or structure of meta-knowledge on the KB and the way they are going to solve problems. Should exhibit the capabilities and the interfaces with other MIS.

#### **6.3.5.5 Identifying of production rules**

To determine the rules of the tasks and subtasks is important to consider the solutions, inputs, outputs, responses, alternatives and recommendations that were identified previously, considering the following aspects:


#### **6.3.5.6 Identifying of meta-rules and frameworks**

Once have determined the rules concerning the problem domain defined those who exercise control over the other and the process of inference.

By identifying, the frames: Select the objects on which the knowledge requires modeling.

By determining meta-knowledge: Control over knowledge, or "knowledge of knowledge" must be determined to have domain information and the inference process.

#### **6.3.5.7 Phase outcome**

58 New Research on Knowledge Management Models and Methods

Domain knowledge module: determines the knowledge that the user needs to perform

Inferred knowledge: knowledge which is needed that can be inferred through the use of AI

Knowledge of processes and tasks of the module: identify the knowledge required for the development of system tasks, and relates to the basis of plans. These layers also serve as support for strategic planning and tactical levels, and are used in the plan generation and

**Layers** of knowledge that are introduced should also support the management at the tactical and operational level, so should determine the knowledge that supports the resolution of

Knowledge of **management skills**: includes the reasoning and decision-making DSS (Decision Support System), taking into account the user's management style in order to identify the subspace and data-information required by the executive for making their decisions in each application that contains the specific module. These layers of knowledge must be determined according to the administrative level support, i.e., differentiating strategic as tactical. Knowledge of **user interface**: where should the user's knowledge representation and should be considered cognitive style and managerial. In this respect it

The second question is useful for determining if it is necessary to analyze and develop a comprehensive plan to draw conclusions. The tasks that are discussed are related to the tasks performed by the experts to solve their problems and do their job. These tasks should be decomposed into subtasks or sub-functions according to the natural order in

To analyze the order of tasks: It should be a state diagram of the tasks previously listed, considering their order. With this diagram, it should do an analysis to determine if the order

Construction of the model: The model of a KB is useful to the extent as it constitutes a pattern of knowledge, easily interpretable, which determines how it will solve the problem in the KB. This model must be designed by determining the knowledge islands and regions. Designing the architecture of the KB: Should design the hardware architecture of the KB, in which represents the different components and their relationships to have a "map" (e.g. semantic network or cognitive map) of the KB and facilitate construction. The conceptual design of the map gives the main idea or structure of meta-knowledge on the KB and the way they are going to solve problems. Should exhibit the capabilities and the interfaces with

To determine the rules of the tasks and subtasks is important to consider the solutions, inputs, outputs, responses, alternatives and recommendations that were identified

**6.3.5.3 Determining of knowledge related for problem solving** 

pursues a natural language interface to work with the system databases

*¿Is there any difference between the user's perspective and that of the expert?* 

Tasks are evaluated regardless of the user model. At this point should answer:

is correct, any extra work, or can be rearranged for performance optimizing.

properly in the company.

which supports the user.

**6.3.5.4 Tasks decomposition** 

*¿What tasks can be decomposed?* 

which they develop.

other MIS.

**6.3.5.5 Identifying of production rules** 

previously, considering the following aspects:

plan recognition.

problems.

Selected Method for solving the problems identified. Problem solution domain. Meta-cognition. Specifications accomplished.

#### **6.3.6 Prototipyng**

The first objective is to build a small prototype, for which selects a subset of the KB and carries to the KE tool, which must be done quickly. The result is a prototype which can quickly verify the implementation and testing and verifying ideas. In this sub-step is the representation of knowledge with the tool by means of a prototype, since this technique to identify weaknesses and strengths of the model developed, by which you can refine the results to get quality. A prototype is also a good way to test the concepts before investing in a larger program. With the use of a shell can quickly assemble a small prototype to determine if you are on the right track. It allows for demonstrations, as its assessment will also be important in determining the quality of the result.

#### **6.3.6.1 Graphical representation of knowledge layers**

When rules and frameworks have led to the KB by using the tool must be plotted to assess the concatenation rules, relationships and inheritance between frames, i.e. to determine if the model is represented which really needs.

Elaboration of the rules diagram. There is a network diagram of the rules and actions for each of the layers of knowledge, which will have a broad overview of the KB and make the adjustments necessary to refine the model.

Elaboration of the diagram of the frames. There are graphically represented all the objects, taking into account the inheritance and relationships between them.

#### **6.3.6.2 Inference proof**

From the KB, the inference process is made to assess the results based on the reasoning process was implemented.

Analytical Models for Tertiary Education by Propaedeutic

tutoring.

to obtain. **6.3.7.4 Phase outcome** 

work the first time.

**6.3.8.1 Phase outcome** 

Intellectual Assets Institutional Memory

Populated Knowledge Bases Institutional Learning System

Integrated, Robust intelligent System

Tutorial Intelligent Engineering Knowledge

users.

Correct source of knowledge. Methodology used properly

Refined knowledge System.

man-machine.

Cycles Applying Knowledge Engineering and Knowledge Management 61

Systems). Without knowing the source, are asked to compare the solutions. With the results of the comparison determines how valid the results of IMIS are. To use this approach should consider potential disagreement between the criteria of the evaluators, since the problems

Should meet a recognized and important business needs or assistance, counseling,

One advantage of rule-based systems is that they are modular, so can build subdivisions of large systems and then test them step by step, which is made possible with prototypes. One approach is using the evolutionary prototyping, where a prototype is enhanced until obtaining the final system, which can be built gradually by adding pieces, parts or modular components. This is a cyclical process, which is an advantage that can improve the result

If each subsection is evaluated and approved separately, the final system will most likely

In this sub-step is performed integration of subsystems that is, if KB is created different or heterogeneous in terms of sources, we must take into account the need to create interfaces between them, allowing optimal operation and communication between the system and

and get the best approach to user requirements and needs of knowledge of IMIS.

can be so complex that occur disagreements over its interpretation and solution.

The system must be able to increase the skills of the user: dual learning,

Knowledge should be easily modified (allow add, delete and modify rules.)

 It must be reasonable degree of physical and mental effort of a novice user. Must be clear requirements in terms of data entry and should be simple

Other requirements to be considered in validating the final system are:

The processing speed of the system must be very high.

 The system must be able to answer simple questions. The system should ask questions for additional information.

Make the user feel the control system to encourage its use.

Error correction should be easy to perform.

The solutions checked and conclusions as expected.

**6.3.8 Obtaining the final system and evolution** 

#### **6.3.6.3 Phase outcome**

Verified operational prototype and with quality assurance Documentation of software testing and prototype Effective and proven outcomes Prototype of Dual-Use: as a Professional Tool and for Powered Training Tutorial
