Introduction
Knowledge
-the data are uninterpreted signals; they carry information.
-information is an interpretation of data;
-knowledge uses information within the framework of actions, for a specific purpose. Actions can be making decisions, creating new information, etc.
Example:We are conducting an experiment to compare the reaction of three students at different school levels (e.g., a primary school student, a middle school student, and a high school student). We ask students to read the recorded display on the thermometer to take a baby's body temperature.
-1. the primary school student reads the number 40 and gives no reaction.
-2. the middle school student reads the temperature 40° but gives no reaction.
-3. the high school student reads the temperature 40° and asks for medical assistance for the baby.
-the primary school student is able to read the symbol 40 that it is the number forty but not the other symbol ° for the temperature;
-the middle school student is able to interpret the symbols 40° in a temperature.
the high school student in addition to being able to interpret the symbols 40° in a temperature, he requests medical assistance for the baby. His reaction can only be justified if he already has other knowledge which allows him to know that the baby's condition is in danger and therefore the baby requires medical assistance.
In summary, knowledge is a set of notions and principles that a person acquires through study, observation or experience and that they can integrate into skills. (definition on the web)
Accomplishments during the decades of the 1950s and 1960s:
● Importance and possibility of automated reasoning.
● The problem of processing large search spaces.
● Need to understand natural language and other knowledge representations
● Potential of semantic networks (and graphical representations in general) as layers of abstraction
● Relevance of systems and high-level languages to manage data.
Limitations of contemporary techniques:
● Physical, technical and material cost limitations
● Gap between graphical representation and linear implementation
● Gap between the logic of human language and the data processed by computer systems
Achievements:
● The need for independence of representation, to have the relational model as a first example. This approach could also be implemented in practical systems.
● The need to formalize semantic networks using formal logic tools.
● The possibilities of combining logic and data using networks.
Contemporary limitations:
● On the DATA side, more flexible data structures were needed to represent new forms of data giving rise to object and graph oriented data structures.
● On the KNOWLEDGE side, a better understanding was necessary on the formalization of knowledge in logic giving rise to description logics.
Achievements:
● The combination of logic and data should be tightly coupled (not just a prologue/expert system layer on top of a database)
● Trade-off between the expressive power of logical languages and the computational complexity of reasoning tasks
Contemporary limitations:
● Negation was a killer. This was not well understood at the time.
● Large-scale reasoning was still difficult. The equipment was not going to be up to the task.
● Realization of what might be called the knowledge acquisition bottleneck
Achievements:
● The Web was rapidly beginning to change the world of data, information and knowledge
● New types of data were propagated (especially media: images, video, voice)
● Data must be (and now can be) connected to achieve value
Contemporary limitations:
● Computing power to handle new levels of data produced by the Web
● Pure logic techniques have complexity limits that make scalability impossible
Achievements:
● We learned to think about data and knowledge much more broadly (web-scale)
● Enter the era of neural networks with new hardware and intelligent learning techniques
Contemporary limitations:
● I don't know how to integrate logical and statistical views
● Statistical methods (especially in neural networks) do not provide information about the process of "reasoning" or "deduction", which generates problems in areas where explanation is needed.
Conclusion :
Throughout this story, we have observed two important threads:
1) Represent and manage data and knowledge at scale
2) Integrate the most diverse, disparate and almost unlimited amount of data and knowledge sources (structured data text, rules, images, voice, videos, etc.).
Knowledge Engineering
In computer science, knowledge engineering, the English equivalent of Knowledge engineering, would refer to techniques for manipulating knowledge on a computer.
-Integration of artificial intelligence techniques and software engineering to design and build knowledge systems.
-Discipline studying the extraction and formalization of knowledge from a third-party source with a view to its integration into knowledge systems.
Knowledge systems
Knowledge system or knowledge-based system (KBS) is a computer system whose knowledge base is explicitly represented. Two central components of KBS are:
•KnowledgeBase: Consists of a set of facts and a set of rules, frames or procedures
•Inference engine: Responsible for applying the knowledge base to the problem at hand.
Engineering Process
Knowledge Management
The terms "knowledge management" and "knowledge engineering" seem to be used as interchangeably as the terms data and information. Knowledge engineering is primarily concerned with building knowledge systems, while knowledge management is primarily concerned with identifying and exploiting knowledge for the benefit of the organization.
Process roles
It is important to identify a number of roles that humans play in knowledge management and engineering processes. We distinguish six different roles:
Knowledge provider/specialist.
This is traditionally an “expert” in the application area, but can also be other people in the organization who do not have “expert” status.
Knowledge engineer/analyst.
While, strictly speaking, the term "knowledge engineer" refers to workers in all phases of the development process, this term is generally reserved for system analysis work. Therefore, “knowledge analyst” might actually be the better term.
Knowledge manager.
The knowledge manager is not directly involved in knowledge systems development projects. Knowledge manager develops enterprise-level knowledge strategy.
Bibliographic Notes
This chapter introduced the concept of knowledge and highlighted the relationship between knowledge, information and data.
Kendel and Creen [An Introduction to Knowledge Engineering. In: An Introduction to Knowledge Engineering. (2007). Springer, London, https://doi.org/10.1007/978-1-84628-667-4_1] provide a relevant discussion to differentiate between these notions. Knowledge-based systems or simply knowledge systems are computer systems that include a knowledge base. Knowledge engineering is the discipline put in place to develop knowledge systems. While knowledge management mainly deals with identifying and exploiting knowledge for the benefit of the organization. Schreiber et al [Schreiber, A. T., Schreiber, G., Akkermans, H., Anjewierden, A., Shadbolt, N., de Hoog, R., ... & Wielinga, B. (2000). Knowledge engineering and management: the CommonKADS methodology. MIT press. ] give a detailed comparison of these two disciplines and discuss the different roles involved in the engineering process and thus the skills necessary to exercise the role of knowledge engineer