Expert systems are developed to solve complex problems in a specific domain, such as system configuration, trouble shooting, data interpretation, etc. The core of an expert system is a knowledge base, which is primarily a collection of decision rules that represent explicit knowledge about the target domain. The amount and the quality of the knowledge in the knowledge base determine how useful and powerful the system is to the user.
Expert systems in business are usually developed with an expert system development tool. There used to several dozens of these tools in the market, ranging from many thoudand dollars in price to free of charge or less than $100. A great shakeup and the consequent consolidation took place around 15 years ago. Now there are only a few out there. For an example, go to www.exsys.com.
When the first batch of expert systems appeared, people were amazed at the problem-solving power and create unrealistic expectations for something even more powerful. When the expectation "bubble" bursted, the pendulum swun to the opposite extreme almost overnight. Even until today, expert system technology is still viewed as either completely useless or at least not practical for business applications.
The fact is, expert system is a great technology for certain types of problems. Often , they are integrated with other front-end or back-end applications to deliver intelligent services at some points of user-system interaction experience. It's how the technology is used, rather than the technology itself, that determines its value to the user.
Subscribe to:
Post Comments (Atom)
13 comments:
Public School Disciplinary Action Advisor
This Expert system was designed to help faculty at public schools decide what disciplinary action should be used when a child needs to be punished. It asks a series of questions then generates a list of possible actions to take.
This was an interesting expert system that I found to be helpful...however it seemed to need some improvements.
There is an automated system for forecasting frost and fog on roadways and bridges at specific locations in Iowa. The System takes in current forecasted values and observations of specific weather variables and procedures forecasts of the indicated roadway condition. An accurate forecast of frost in Iowa is greatly needed dut to potential safety problems for motorists, since the roadways need suppression measures for frost. The system uses a backward-chaining system and consists of 32 parameter and variables and 33 rules. Bridge and roadway data from December, Jan, Feb of 4 frost seasons (1985-1989) were used to develop the rules for the system. Then these rules are used in combination to forecast separate values of temperature for the bridge and roadway, which are compared with the forecast of the dew-point temperature to determine the likelihood of frost.
ISHS Acta Horticulturae
Expert Systems (ES) are computer programs that have been developed to emulate the logic and problem solving characteristics of a human expert. They are most successful in agricultural applications when addressing specific problems frequently encountered by extension specialists requiring experience, judgement and interaction to arrive at a solution.
An example I found was a restaurant selection system. The system asks questions concerning food preference, ocassion, and atmosphere you desire to dine in. Based on your response, the system may ask more questions or give you suggestions that rely on a ranking system. For example, if you say you want a quiet place to dine, then the system places more weight on setting and not as much on actual food preference. If a quiet dining place is not desired, then type of food is given more consideration. The main use of such a system is to take information and use it to make better decisions about products or services in business or on a smaller note, your next dinner date. :)
Seth
The aim of this expert system is to give specific advice concerning the treatment and monitoring of drug poisoning. It simulates expert reasoning, taking into account for each toxicological class, delay, clinical symptoms and ingested dose. It generates accurate monitoring and treatment advice, addressing also drug interactions and drug exceptions.
The expert system I found dealt with Intelligent Tutoring System (ITS) that assist student in learing by using techniques that adapt and personalize the learning experience according to prior knowledge, environment and students learning ability. The systems continually adjust learning pace and monitors and records progress. It can possibly take the place of tutors and diagnose problems.
Easy Diagnosis is an online expert system that provides a list and clinical description of the most likely conditions based on an analysis of your particular symptoms.
Easy Diagnosis is staffed by a team of doctors who assist you in diagnosing your medical symptoms. The main downside to using this expert system is that you have to subscribe to it. One evaluation costs $4.95 and a year of unlimited access is $19.95. The website also provides a disclaimer that if a patient presents more than one major symptom, the expert system tends to be less accurate.
I am not sure how well this system works, but the website appears to be legitimate and it may be worth the time and money if it saves you from having to go to the doctor.
http://easydiagnosis.com/
An example of an expert system that I see frequently are financial management tools on Fidelity.com. A specific one is a retirement planning tool which allows you to answer a few questions, the system then gives appropriate investing options based on your answers. A few of the questions include age, risk adversity, etc.
The expert system I found was a whale watcher identification system. This is used for Canadian coastal waters to pinpoint the type of whale spotted. The company that provided the ES is Acquire.
http://www.aiinc.ca/demos/whale.shtml
Here is an example of a consulting firm that will help you implement Business Rules Mnagement Systems (BRMS).
http://www.haley.com/brmsoverview/brms.html
There are great examples on this web side of clients that have taken business rules and integrated them into their systems to help them get:
Faster time to market for both reactive and proactive changes,
Reduced application development time, Shorter change cycles,
Centralized management and maintenance of business logic,
Reduced dependence on IT for system change, Parallel development of business logic and application code.
The idea is to capture the business rules (Read this to say Knowldege) and build a decision system (read this to say Expert Systems).
GIDEON
GIDEON is a program for diagnosing and referencing tropical and infectious diseases, epidemiology, microbiology, and antimicrobial chemotherapy. It works in 4 ways: Diagnosis, epidemiological module, therapy, and microbology. In a field trial GIDEON displayes a correct diagnosis in over 94% of cases. A trial can be found at www.gideononline.com
The expert system that I located is a career consultant. The user may sign up for free and take a survey-type test that will assess the user's likes and dislikes in the form of a basic career profile. From this test, the system will determine what areas (such as food service, science, administration) that user seems to favor because of the answers given on the test. It first gives an overall summary of the user’s answers, and then breaks its down into interests, personality, knowledge, skills, and abilities, and much more.
After the assessment, an option is given to upgrade to a premium career report for a fee. This report will match the user’s characteristics with jobs and educational opportunities.
I took the test and found out that my main areas of interest were food service, administration, and science. This makes me sad because I have gone to school for five years now to be an accountant! However, accounting was not an option given so maybe it's ok!
www.livecareer.com
Mycin is an expert system that designed to diagnose infectious blood diseases. In addition to diagnose, the system recommended antibiotics with the dosage adjusted for patient's body weight.
Mycin operated by using a simple inference engine and a knowledge base of 500 rules. The system would ask a series of yes/no or textual questions. In the end, the system provided a list of possible bacteria ranked from high to low based on the probability of each diagnosis. Last, it would recommended drug treatment.
Post a Comment