It is modelled as a step up from data, the raw information and information which is the data that has been analysed. Knowledge is information that has been analysed and contextualized that can be used to solve problems. Over a long period of time, one builds up experience which leads to wisdom; the combination of past experiences and knowledge combined to solve problems. (Akerkar & Sajja, 2010)
There are many ways to gain knowledge. Some of these channels to gain knowledge include experience, books, the internet, tv, professionals and online resources. (Randles, 2019)
Facts, knowledge that can be proven true, such as the fact that goldfishes have gills, Heuristics, knowledge gained through general assumptions, such as if you drop your phone in water, it may no longer work and rules, a statement that states what the consequences of an action may b, such as, if you cheat in an exam, you will be disqualified from the exam. These 3 are the different components of knowledge.
Knowledge is a complex subject and not all knowledge is the seen as the same which is why there are different categories of knowledge (Akerkar & Sajja, 2010). Some of these different types of knowledge include –
• Meta knowledge – this is knowledge about knowledge. An example of this is how in HTML, there is the meta tag. This tag can be written by web developers and read by web browsers and contains the metadata of a document.
• Explicit knowledge – this is knowledge is less subjective and easily expressed through words and numbers and is easy to transfer and explain (Randles, 2019). It can be found in data containing mediums like databases and spreadsheets
• Common sense knowledge – this type of knowledge is something that builds up with experience. Over time an individual learns from their mistakes and knows how to approach certain problems. For example, if an individual touches a hot pan and burn themselves, they gain the “common sense not to touch a hot pan again”
• Domain knowledge – this is the type of knowledge in a certain area/domain. It can be knowledge that is developed over time, but it is also possible to have domain knowledge and not continue to develop it. For example, a zoo keeper has a large domain knowledge about zoo animals that they build over time through learning and experience. At the same time, the zoo keeper has a small domain knowledge about toasters. They would know how to use a toaster but not how the toaster works. It is not a domain they will develop but it is enough to able them to use a toaster
As knowledge can be grouped into different classifications, knowledge can also be split into 3 different types of consistencies
• Dynamic – frequently changing knowledge such as my phone’s charging
• Static – this is knowledge that is constant over a long period of time such as government regulations
• Permanent- knowledge that doesn’t change, such as the value of pi
Knowledge based systems architectures
Considered a major branch of artificial intelligence, a knowledge base system is a programming language paradigm that uses reasons and its knowledge base to solve complex problems. It consists of a knowledge base containing the knowledge of more than one expert. The main goals accomplished from utilizing a KBS is to provide a high level of intelligence, such that which is obtained from experts. It is advantageous as it can be used to assist people in discovering and developing in new fields (i.e. helping in the educational training process), offering a vast amount of knowledge, giving a new perspective of how to approach and solve a problem and offer improvement in software productivity. (Randles, 2019)
These systems also benefits society by assisting them when they seek knowledge. (Akerkar & Sajja, 2010) This is advantageous to the user for several reasons –
• There isn’t enough time and money being spend seeking an expert
• The expert saves time by having less people to give knowledge too
• The knowledge is available on demand to the use, regardless of the location or the time
Properties of a good knowledge-based system include:
• Being able to clearly represent knowledge from their knowledge base
• Being able to efficiently manipulate knowledge from their knowledge base to solve problems
• Being able to use knowledge from the knowledge base to create new knowledge. For example, if a knowledge base system knows from its base that fire is hot and fire burns., therefore if anything is hot, it will burn as well.
A knowledge-base system can be classified into different categories. Some example of these categories includes –
• CASE-Based Systems
• Expert Systems – The main use it to replace an expert which is beneficial for numerous reasons such as, an expert may not be available when required. Such as an appointment with a doctor may not be available for a few days, the expert system can stand in their place. Another reason can be that the expertise needs to be cloned or multiplied.
There are many benefits to this type of KBS. This has more flexibility than a real expert, increases output and productivity, has educational benefits which may be less costly and time consuming then acquiring a real expert and can be used to capture scarce expertise.
• Neural Networks (NNs)
• Intelligent tutoring systems(/agents)
• Genetic algorithms
• Database in conjunction with an intelligent user interface
• Linked Systems
• Data Mining
A KBS is made up of the following:
• Knowledge base
• Interface (IE)
The knowledge base contains all the knowledge the system requires to solve real-life problems, organized into categories. The type of knowledge in knowledge bases usually consists of meta knowledge and domain knowledge (Randles, 2019). This knowledge is accessed through the IE, which combs through the knowledge base searching for the required knowledge to solve the problem. In some cases, there may even some memory allocation to store the results and pieces of knowledge temporarily. (Akerkar & Sajja, 2010)