An expert system is one that consists of five technical components. Firstly a user interface that the user will interact with. Second there exists a working memory that contains known facts regarding the domain of knowledge that the system is meant to provide assistance with. Thirdly there is the Knowledge base that brings in the expertise of specific professionals in getting the job done; it can analyze the working memory and edit it in order to provide a more succinct answer. Fourthly there is a inference engine that brings in the programming expertise to provide an answer that the computer is actually looking for. The inference engine uses all of the working memory and knowledge base to fulfill antecedents of an argument until they have each been filled and the solution can be "fired" or output, generally the inference engine will consist of high level if-then statements to fulfill it's goal. And finally there is the Explanation engine that provides the complete reasoning to the user of why an answer was given.
In business an expert system provides knowledge and reasoning to employees as a learning tool and utility. They fulfill domain specific roles that bring in expertise that can usually be too costly when it comes to travel requirements. Sharing that system with others provides a learning tool and a great asset to a company. One example would be in many current healthcare facilities there exists software that monitors patients medications and determines if they will likely have bad reactions.
An expert system applied in business requires less expertise from the end user but will take longer to arrive at a conclusion.Expert systems provide a great resource to the user for simultaneous events in a situation where rules do not have to overlap. Because they provide their reasoning, this understanding can be passed to the end user, so are used in business applications in cases of business acquisitions, large loans, and the like. However, because they lack common sense, the full power of human knowledge is still not encapsulated within this system. There is also the challenge that experts are resistant to providing their full expertise as to avoid devaluing themselves as employees.
In a neural network it models the human brain by utilizing a mathematical model of individual neurons in the brain. Each neuron works similarly to that in the human neural network. In a human brain a neuron has many inputs and provides a singular output. It is connected to a plethora of other neurons with synapse that vary the input to the neuron. In an artificial neural network or ANN they use a mathematical model developed in the forties to pride a binary output of 1 or -1. When a certain threshold is met the neuron is fired with the release of the 1. These systems are developed much differently than an expert system because they don't use any knowledge base. Instead they are given a learning algorithm and a perception structure that begins with many input neurons a layer of switch neurons and finally the output neurons.
When it comes to their application within knowledge management, a neural network helps streamline a process into a more and more efficient use of resources and helps even more quick access to knowledge contained within the system. Advantages to an ANN in knowledge management are very dramatic after a training period. This type of system has a high tolerance for "bad" or inaccurate data. It starts all over with every new application. It must also be retrained for each problem that must undergo analysis. A system like this requires a higher expertise from the user, but is processed much quicker. Specific applications include financial forecasting, investment, and risk management.
These applications of artificial intelligence are real in today's society. In the future when the feeling of experience and emotion can be attributed to a computer in a safe atmosphere then, and probably only then will Siri be a much more compelling purchase. As of now, though she has some uses, she is still quite limited because she is an obvious emulation, that still can come off quite cold. There is a movie that comes to mind however, the 2009 Moon is about the mining of a rare element abundant on the moon for power generation. The main character lives alone the entire time, except when he is in an accident on the moon. All of the sudden a new him appears at the base station. He is overlooked by a computer AI, voiced by Kevin Spacey. Then the new main character goes to try and find the missing vehicle is earlier counterpart embarked in. When he finds it he sees he is a perfect clone of the man. The company has been using clones to mine the moon. But what is interesting is that the data provided to the character to find him is given by the AI. It is at this point the AI reveals that he cares about each clone, and that he has a genuine emotion for the "people" under his care. A Skynet like AI is the counterpoint to this, but in the future let us hope that if there is an AI it can be endeared with the same appreciation for non artificial life that this one had.
No comments:
Post a Comment