Artificial Intelligence And Expert Systems

Artificial Intelligence And Expert Systems

Instructions/Descriptions

Artificial Intelligence And Expert Systems

ABSTRACT:

• What is artificial intelligence?
• Characteristics of AI
• The field of AI
• How the AI field evolved
• What is Expert System?

KEYWORDS:

• Artificial intelligence Technology
• International Technology Transfers
• Expert Systems
• Applied Artificial Intelligence Human Experts

INTRODUCTION:

• AI to Predict and Adapt
• Making Decisions
• Continuous Learning
• Capability of Motion and Perception

OVERVIEW OF APPLIED ARTIFICIAL INTELLIGENCE:

• Natural language processing
• Robotics
• Computer vision
• Speech Recognition
• Machine learning

EXPERIMENT:

APPLICATIONS OF EXPERT SYSTEMS:

• Troubleshooting and Diagnosis
• Financial decision making
• Knowledge Publishing
• Design and Manufacturing

[Order Now]

EXPERT SYSTEMS IN ORGANISATION:

BENEFITS AND LIMITATIONS OF AI ALONG WITH EXPERT SYSTEMS:

LIMITATIONS OF EXPERT SYSTEMS:

NEURAL NETWORKING:

• Artificial neural networking
• Training Data

Artificial Intelligence and Expert Systems

Sapana Dahal

CSCI 303

Texas A& M University Commerce

Artificial Intelligence (AI) in the current technological crazed world has been on the center stage. AI has commendable gained notable popularity and visibility that is very string in the public domain, business community, Educations and various other special fields. Now the most successful business in the world have not hesitated to fully take advantage of AI to better their products, through presented AI to the consumer in a cheaper state.

This is the example Amazon Alexa AI, Apples Siri AI and Microsoft’s Cortana (Kaplan & Haenlein 2019). These are aperfect example of the use of AI to make things easier for the consumer, all these considered to be “personal assistants. Expert Systems (ES) on the other hand are a dominant filed in AI, basically the largest field in AI currently, this is because it offers scientific, commercial and military application of AI.

This paper is aimed at looking at and explaining the AI concepts and ES applications that have been able to make the life of every individual in different field easier. Like ROSS the Ai attorney, or the Dendral expert system in medicine.

The Implementation of AI technology cannot be ignored in our daily lives cannot be ignored because of the impact that such a technology is bringing to our world. Never have machines been able to mimic the human brain and be able to tackles decision making or problem-solving instance as better or just as the human brain.

These aspects have caused a major quagmire of mixed feelings. Mainly because the AI and ESs will make our lives easier bhut at the same time it will step in the place of human experts that will mean they will be replaced. The paper clearly discusses on this fact on how the ESs and Ai are not here to replace the Human experts, especially the white-collar jobs but only to make their task much easier.

KEYWORDS

Artificial Intelligence Technology, International Technology Transfers, Experts Systems, Applied Artificial Intelligence Human Experts.

INTRODUCTION

Artificial Intelligence (AI) can be described as the creation or the simulation of human intelligence into machines with the aim of mimicking the actions and thinking like human beings. The term is broad as it can eb associated with any machine that is able to display various human thinking traits like problem-solving and learning.

In the current world or the world to come artificial intelligence will be the most used or preferred technology as it will make the work easier or rather replace human beings in tasks that are hard to be completed by human beings the ideal characteristic of artificial intelligence is taking actions and at the same rationalization with the aim of completing a specific goal (Upadhyay & Khandelwal, 2019).

The most perceived idea about artificial intelligence is robots, the immediate thing that comes to mind when someone mentions the term artificial intelligence. This is through the novel and the movies that weave stories depicting AI machines that bring havoc to the world. But it might not be far from the truth, the most important aspect to understand is that AI is continuously developing into new and better versions on a daily basis; with machines build withing various fields like; psychology, linguistics, mathematics, computer science and more with the aim of solving issues in these fields better than human beings.

There are many merits of AI, it is because as a simulation of the human thinking it is developing and growing every day. These merits clearly represent the AI and the include: AI can adapt and predict. The use of algorithms has enabled AI to predict and adapt, whereby patterns of past data are used for future prediction and decision making.

These patterns give AI the ability to learn like human beings and put through software systems that can be used to correct errors or predicting what is going to be typed and estimation shortest routes to take (Kietzmann & Pitt, 2019). These are some of the abilities under this characteristic.

AI simulates human intelligence, and therefore every aspect the human intelligence has is mimicked like for instance the ability to analyze data, make decisions and gain new insights to make the best decisions, faster than human counterparts. Which makes it preferred to human labor by most companies.

AI is also known as machine learning, meaning it can continually learn through of the most notable method know as deep learning used by Alexa, Siri, Netflix and Google. Whereby analytical models are built to perform tasks through countless trial and error. AI is reactive meaning that atypical AI will perceive a problem and develop a better solution for it; like the application on our gadgets predict on our actions and provide recommendations that suit the user.

Finally, the rother major characteristic is the capability of motion and perception since the development of AI in the last 50 years ago, AI has evolved into this phenomenon that is able to have auditory, visual, speech perception and information learning and processing that is perceptual. Whereby the goof examples include: Amazon Alexa, Apple’s Siri and the Tesla’s Self driving Cars.

The first practical application of AI was developed by the British Mathematician Alan Turing, who was responsible for conceptualizing machines that had the ability to think in the 1950s. at this point the seeds of AI were set and it picked up from this moments, whereby followed by an AI lab that was setup in Massachusetts in the year 1959 by Marvin Minsky (Garnham 2017).

Through this lab their collaboration with Stanley Kubrick who received radical ideas from Marvin who developed various literature and movies relating to AI the AI excitement was further catapulted by the Invention of the computer in the 1980s (Garnham 2017). whereby this opened way for the tech corporations who have explored various fields of AI that are used to bring technology close to the consumer through affordability and through the various applications.

Today AI is every way, in smartphones, traffic routes, Netflix, Amazon and Apples and the other various corporations or fields embracing AI. AI has transformed and is continuing to transform how things work in the world, through the basics of machine learning that are instrumental in various instances: such as advertisement that are based on machine learning focusing on the data inputs that are received from different channels.

But this can be considered as the beginning of AI take over, the future is envisioned as an economy and industry that is robot driven. A lot of companies are looking into self-driven cars, which basically will change the basic characteristics of private and public transport.

Additionally, AI has played an important part in the management and security sector, through the analysis of anomalous data patterns. A good example is the chatbot which is an AI application that is increasingly used by companies replacing regular customer service agents, which are not as effective as human.

Although creates an all-round customer engagement through machine learning. Whereby a good example is the trials of AI genoming, proving that humans and machines will come to works side by side, and it will be considered normal as it is growing form the various efforts put in place at the moment.

The future of AI is very clear, looking at how it has evolved to this moment; although there are various challenges in the widespread adoption, the world has overlooked over the advantages that AI brings to the world and making work much easier, but also ensuring that AI is managed effectively.

There is an increased risk which will affect the development of AI as it majorly depends on Data, and it will be challenging as it will cause social unrest and ensure human jobs are redundant. Whereby it is the responsibility of people pushing AI to makes sure there is a good balance between the positives and the negatives of AI which will not destroy but bring a Positive AI revolution in the Future (Pauschunder 2019).

Experts systems can be described as programs that rely on AI, in other words these two are tied. There is no way to discuss AI and without mentioning Expert systems, and vice versa. Experts systems can use AI to simulate behavior and judgement of an organization or human that has expert knowledge of a field. Basically, an expert system brings together an interface of rules known as engine and accumulated experience known as knowledge base (Andikos et al. 2016).

Whereby the rules engine applies knowledge based to each situation that is spelled out in the program. The capabilities of the system can be improved with various addition to the set of rules or the knowledge base. Current expert systems are typically equipped with capabilities of machine learning which makes it easy for them to perform better as humans through learning from experience.

Developed in the 1970s the expert systems have been significant in many industries like telecommunications, financial, customer services, healthcare, video games, transportation, manufacturing, written communication and aviation. A good example of an expert system was the groundbreaking medical system which helped in the diagnosis.

Deandra Expert System assisted chemists to understand organic molecules that led to the easy identification of various bacteria such as the meningitis and bacteremia and recommend the nest dosage and antibiotics (Tan 2017).

The expert systems therefore are useful in the current technological world in the sense that it can eb used to resolve a lot of issues, that would generally require human expertise. Interestingly enough AI and expert systems can eb sued to reason and expression in different domains of knowledge.

Basically, it is safe to say that expert systems are the predecessors of the modern-day AI, machine learning and deep learning systems. There are different examples of Expert systems: MYCIN, tis was developed earlier in the medical field and it was designed around backward chaining, that gave it the ability to identify various bacteria that could cause infections and at the same time recommend drugs to patients.

Similar systems are DENDRAL in the chemical engineering field, PXDES in the medical systems specifically for lung cancer and CaDet, in the medical filed for cancer identification. These systems are in every filed even the legal system, which will revolutionize labor in the world as it was previously identified.

APPLICATIONS

Overview of Applied Artificial Intelligence

AI is rapidly growing and soon enough it will be covering almost all areas of the world, whereby it is quickly moving from the lad to the consumer and business applications (Bravo et al. 2016). The result creates a significant on how software is built now as compared to previous versions.

Now applied AI is one of the powerful technologies powering the most successful business currently, Apple, Google, Facebook and Amazon. There are a few areas where AI has been applied which are: Natural Language Processing (NLP) a branch of AI that mainly deals with interactions between humans and computers using the natural language, with the aim of reading, deciphering, understanding and making sense of human language in a way that is valuable.

Robotics, AI in this field helps in saving certain motions in the systems of robotics, these motions are constantly refined that makes moving and installing robotic systems easy to achieve; the result is robots being used in the customer service sector globally. Machine Learning, as mentioned before it is the way in machines learn new concepts on its own without explicit programming, which is an applied AI that enables automatic improvement and learning form experience. These are some of the Application of AI in addition to more others like Speech recognition and computer vision (Hengstler et al. 2016).

Capabilities of Expert Systems

In AI, expert systems emulate the decision-making ability of humans to solve various issues which is just a simple description. Experts systems are built to have knowledge to one domain like, engineering, science and medicine. This knowledge as mentioned before is known as a knowledge base and has gathered experiences that are tested after being loaded in the system. The rules and adds-on to the knowledge base maybe add to give the expert systems to create a better system.

The system is described as reliable, highly responsive, understandable and high performance. Expert system typically will come into the world to replace the individuals in the white collar and analytical jobs. Because expert systems are excellent in pattern matching, configuration, classification, reasoning, planning and diagnosis (Bogdanova et al. 2016).

The Inference engine is the basis for the capabilities of ESs. To have a solution the Inference engine manipulates the knowledge that as acquired.

For instance; when an Expert system is rule-based in AI, there is the need to add knowledge if require, the rules obtained from earlier application need to be applied to the facts and this will resolve rules conflict, especially when multiple rules are presented in an instance. The following strategies are used to recommend a solution, the forward chaining and the backwards chaining.

Forward chaining: this is a strategy used by an Experts System. This strategy helps to answer the question what happens next. Basically, the chain of derivations and conditions are followed in this strategy and at the same time reduce the outcome. All the rules and facts are considered and are sorted before the conclusion is made to recommend a solution. A good example is predicting the share market status, based on the change in interest rates.

Table 1. Prediction of share market status as an effect of changes in interest rates.

Backward Chaining: This strategy is sued by the Expert System to answer the question “why this happened?” normally in this case what had already happened will be very significant matter to the organization or the individual. Therefore, this strategy will focus on finding out the conditions that led to the particular outcome in place in the pat to have this result. Therefore, is a strategy that finds reason or cause. For example, human blood cancer.

Table 2. Diagnosis of blood cancer in humans.

Applications of Expert Systems

There are various application of Expert Systems, but the most important things to consider when choosing and Expert System ES is that the tool that has been chosen needs to match the qualification the project team possess, and the tool selected for the projected has to go hand in hand with the sophistication and capability of the projected ES.

The first application is the troubleshooting and diagnosis of Devices of all kinds, basically discovering faults and suggest remedies for these faults for a device or process that is malfunctioning. A good example is medical diagnosis the first ever application of ES.

Planning and Scheduling, systems in this case analyses interacting and potentially complex goals and determine set of actions that can be sued to achieve the same goals. At the same temporal ordering of the actions can be initiated with the ES considering material, account personnel and various constraints. Financial Decisions Making, the financial service industry has a common use of ES techniques some programs have been created for the sole reason of advising bankers to whether give loans to individuals and businesses (Wagner 2017).

Knowledge Publishing, this is a new application that has not been there for some years but potentially an area that is rapidly growing. The main function of the ES in this case is deliver relevant knowledge to the problems the user presented. A good example of ES in this sector is the advisor that guides the user on the use of grammar in a text and an ES that guides the user on individual tax policy and tax strategy, whereby the ES is known as the tax advisor.

Process Monitoring and Control, ES in this application analyze real-time data obtained from physical devices with the objective of predicting trends, establishing anomalies, controlling failure correction and optimality. Design and Manufacturing, Expert Systems in falling field assist in the design of physical processes and devices. Which can be factory floor configuration of processes in manufacturing or can be level conceptual design of entities considered to be abstract.

Development of AI and Expert Systems

Experts Systems ES have been identified to be systems that utilize AI to provide decisions making and problem techniques that as close as to human intelligence working. Which means that ES really needs Ai to function but at the same time it is an independent entity on its own.

As mentioned before expert systems are hand in hand, AI as one grows the other follows the pattern. As AI is taking over the world, ES follows the same path are they are being deployed and developed in a various part of the world in applications considered to be myriad. This is because of the explanation capabilities and symbolic reasoning of the ES.

A good example is the recently developed Expert System known as ROSS, which is an attorney AI that is built under the principles of pattern recognition, deep learning, self-learning systems and natural language processing that utilizes data mining and at the same copy the working of the human brain (Semmler & Rose, 2017). Whereby this qualifies to be an exciting development yet controversial. It may not be in every enterprise, but the development of AI and Experts Systems has taken over the world.

EXPERT SYSTEMS (ESs) IN ORGANIZATION

ES provide both important and tangible value to various organizations. As mentioned before, the most successful business in the world have benefitted from the use AI and Experts systems majorly through introducing them to their consumers. The benefits should be measured against the exploitation and development costs of an ES, that are basically very expensive especially for ESs that are considered to organizationally important and large (Masuch 2018).

At this moment it seems that Experts systems are the substitution for an individual’s work but that is not the case as the ES can substitute overall performance for a knowledge worker or the white-collar jobs, mainly in analytics and problem-solving tasks. But this is not true, yes, the systems can significantly reduce the individual’s work dramatically required for the knowledge worker to solve the problem but leave the innovative and creative aspect of problem solving to the people.

Basically, ESs are tools that are developed to make work easier. The issue is that it will reduce the work that two people would have performed to be performed by one person, which is good for the company, but it will create societal unrest of unemployment. Therefore, looking at the benefits and limitation will help in establishing where the ESs stand in organizations.

Benefits of AI along with Expert Systems

• The benefits of ESs are straight forward and are used to reduce the tine taken, workload or resources used to complete a task. This is only possible when AI is incorporated to the ESs. The benefits include:
• Es are designed to be able to complete task presented to them faster than how long a human expert would take in the same task.
• The systems that are successful presents a low error rate, whereby at times the error rate is much lower than when a human expert is performing the same task.
• At the same when the ESs are given task, or problems to solve they will come up with various recommendations, which is only possible to a human expert who has experience in the field.
• At some point there is the need for difficult-to-use sources of knowledge in an application, ESs are used and turn out to be convenient to make sure bring9ng the application to this point.
• ESs are instrumental because through machine they can eb able to capture scarce expertise, maybe there is only a few experts in a field, let’s say finance who are the only ones who can perform the specific task. Through Experience the uniquely qualified expert can be accurately mimicked by ESs.
• Currently ES are known used differently, basically organizations know use ESs to build up organizational knowledge, as compared to the organization’s knowledge of individuals.
• For novices, wen training vehicles are used ESs produce higher and faster learning curves.
• Finally, a point that will change the current narrative that taints ESs existence is that, organizations can utilize AI and ESs in environments that are hazardous for regular Human Experts. A good example can be the exploration of Mars; the planet is not safe for human habitation but AI and ESs can be used to come up a better environment that will pave way for human experts.

Limitation of Expert Systems

Given the various advantages or benefits that the ESs present, the human expert remains the most intelligent and efficient system on the planet. This is because there is no total and easy solution that technology can offer. ESs are large, take a lot of time to develop, need a lot of computer resources and overall, very costly. This makes ESs have their own limitations which are:

• Technological limitations, such as the dependence on data.
• There are also issues with the acquisition of knowledge.
• The principle area of ES application is operational domains.
• Affect the maintenance of Human Expertise in an organization, whereby it is important as ESs are not 100% reliable at some point they will need upgrading or break down.
• Expert systems also do not have common sense to apply when making a decision. Takin for an example; an expert system in the Judiciary or any legal enterprise, in such a field most decisions are based on common sense something that will not be possible.
• The expert systems will require upgrades just any other computer systems, meaning it lacks the ability to adapt to changing environments.
• The Expert systems cannot explain the reasoning and logic behind a decision made.

NEURAL NETWORKING

Neural Networks are described as computer systems that are modelled on the brain of a human being. These systems look like interconnected mesh network of processing elements known as neurons. The neural network has around 100 billion neuron brain cells making them simpler that the brain (Shakya et al. 2018).

However, just as the brain, these networks can process various information pieces at the same time and learn how to understand patterns and program to resolve problems independently. A neural network “is an array of interconnected processing elements, each of which can accept inputs, process them, and produce a single output with the objective of imitating the operation of the human brain.”

Information or knowledge that is given out in a neural network in form of pattern connections that is within the elements of processing and through the adjustment of the weights of the same connections. Therefore, the neural networks get strength from applications that need sophisticated recognition of patterns. The weakness of neural networks is that they do not sugar coat an explanation to the conclusion or answers they give out.

But what has not yet been mentioned is that Deep Learning is the name of this new approach that involves neural networks. This approach in artificial intelligence has been there more than 70 years as they were first developed by Walter Pitts and Warren McCullough in 1944 by these two University of Chicago researcher who later transferred to MIT in 1952. These are the founding members of neural networking or alternatively identified as the first cognitive science department (Pospichal & Kvasnicka, 2015).

In both neuroscience and computer science neural network was a major filed of research. Therefore, the ability of google to have applications such as the speech recognizers, google translator on smartphones have only resulted from this technique known as the “deep learning,” that will be further discussed below.

In other words, it is seen that artificial neural networks use various layers of mathematical processing to make sense of what information is needed to be fed. Furthermore, an artificial neural network has millions of artificial neurons which are known as units. These are arranged in a series of layers (Mohapatra, Dandapat, & Thatoi, 2017). With the development of the brain, various cognitive neuroscientists have learned about the effective functioning of the human brain. It is seen that the human brain is responsible for processing different aspects of information and these aspects are arranged in an order that shows hierarchy.

This means that input comes from the brain and each neuron level provides and insight on the information which gets passed on to the next complex level. This is the mechanism that artificial neural systems try to develop and replicate. In addition, it is seen that artificial neural systems need a huge amount of data for them to learn.

This information needed is referred to as training data. In the event the training data has reached a significant amount, it will then classify future information based on what it sees through different units (Prakashkumar, Murugan, Thiagarajan, Krishnaveni, & Babby, 2019). There are different uses of artificial neural networks.

These are, the network classified information, predict outcomes, and clusters data. For example, it is seen that Google used a neural network (30 layered) to power Google Photo. On the other hand, Facebook also uses this aspect for DeepFace algorithm. This is important in that it can give 97% accuracy. Additionally, Skype’s ability to translate information in real time is associated to the effectiveness of artificial neural network.

REFERENCES

Andikos, A. F., Ali, G., & Purnomo, W. A. (2016). Expert System for Decision Support Division of Inheritance According to Islamic Law. IAES International Journal of Artificial Intelligent (IJAI), 5(3), 89-94.

Bravo Serrano, À., Li, T. S., Su, A. I., Good, B. M., & Furlong, L. I. (2016). Combining machine learning, crowdsourcing and expert knowledge to detect chemical-induced diseases in text.

Bogdanova, Y. A., Zaripova, G. R., Kataev, V. A., & Galimov, O. V. (2016). CAPABILITIES OF EXPERT SYSTEMS IN FORECASTING OPERATIONAL RISK FOR THE MOST COMMON INTERVENTIONS OF ABDOMINAL SURGERY. YAKUT MEDICAL JOURNAL, (4), 54-58.

[Order Now]

Bhosale, A., & Konnur, B. A. (2019). Use of Artificial Neural Network in Construction Management.

Garnham, A. (2017). Artificial intelligence: An introduction. Routledge.

Puaschunder, J. M. (2019). Artificial Intelligence evolution: On the virtue of killing in the artificial age. Scientia Moralitas-International Journal of Multidisciplinary Research, 4(1), 51-72.

Hengstler, M., Enkel, E., & Duelli, S. (2016). Applied artificial intelligence and trust—The case of autonomous vehicles and medical assistance devices. Technological Forecasting and Social Change, 105, 105-120.

Kaplan, A., & Haenlein, M. (2019). Siri, Siri, in my hand: Who’s the fairest in the land? On the interpretations, illustrations, and implications of artificial intelligence. Business Horizons, 62(1), 15-25.

Kietzmann, J., & Pitt, L. F. (2019). Artificial intelligence and machine learning: What managers need to know.

Li, H., Zhang, Z., & Liu, Z. (2017). Application of artificial neural networks for catalysis: a review. Catalysts, 7(10), 306.

Masuch, M. (Ed.). (2018). Organization, management, and expert systems: models of automated reasoning (Vol. 23). Walter de Gruyter GmbH & Co KG.

Artificial Intelligence And Expert Systems

PLACE THE ORDER WITH US TODAY AND GET A PERFECT SCORE!!!