Specialised Business Information Systems - Business Intelligence and Data Analytics

Business Intelligence Strategy and Big Data Analytics: A General Management Perspective

Steve Williams
Morgan Kaufmann, 08-Apr-2016 - Computers - 240 pages


Business Intelligence Strategy and Big Data Analytics is written for business leaders, managers, and analysts - people who are involved with advancing the use of BI at their companies or who need to better understand what BI is and how it can be used to improve profitability. It is written from a general management perspective, and it draws on observations at 12 companies whose annual revenues range between $500 million and $20 billion. Over the past 15 years, my company has formulated vendor-neutral business-focused BI strategies and program execution plans in collaboration with manufacturers, distributors, retailers, logistics companies, insurers, investment companies, credit unions, and utilities, among others. It is through these experiences that we have validated business-driven BI strategy formulation methods and identified common enterprise BI program execution challenges.

In recent years, terms like “big data” and “big data analytics” have been introduced into the business and technical lexicon. Upon close examination, the newer terminology is about the same thing that BI has always been about: analyzing the vast amounts of data that companies generate and/or purchase in the course of business as a means of improving profitability and competitiveness. Accordingly, we will use the terms BI and business intelligence throughout the book, and we will discuss the newer concepts like big data as appropriate. More broadly, the goal of this book is to share methods and observations that will help companies achieve BI success and thereby increase revenues, reduce costs, or both.

Provides ideas for improving the business performance of one’s company or business functions
Emphasizes proven, practical, step-by-step methods that readers can readily apply in their companies
Includes exercises and case studies with road-tested advice about formulating BI strategies and program plans

https://books.google.co.in/books?id=xTpUCwAAQBAJ

Ch 11 - Specialised business information systems: AI, expert & virtual reality

Artificial Intelligence

An overview of artificial intelligence

Ask Jeeves demonstrates characteristics of AI: It lets users seek info by asking questions instead of search using key words


Artificial intelligence in perspective

Objective of AI systems include:

* Achieving organisational objectives
* Assisting of medical diagnosis
* Duplication of the functions of the human brain

The objective is NOT to replace human decision making completely.


The nature of intelligence

Some characteristics of intelligent behaviour:
* Learn from experience and apply the knowledge acquired from experience
* Handle complex situations
* Solve problems when important information is missing
* Determine what is important
* React quickly and correctly to a new situation
* Understand visual images
   (A perceptive system approximates the way a human sees, hears and feels objects).
* Process and manipulate symbols
* Be creative and imaginative
* Use heuristics


The major branches of AI
Expert systems
They give the computer the ability to make suggestions and act like an expert in a certain field

Robotics
Mechanical devices perform tasks that require precision / are tedious a hazard for humans.

Vision systems
Hardware & software permits computers to capture, store, and manipulate visual images.
E.g. Performing fingerprint analysis / identifying people based on facial features.
Generally, robots with vision systems recognise only black, white, and grey shades.

Natural language processing
A computer can understand and react to statements and commands made in English.
Three  levels of voice recognition:
* Command (recognises dozens to hundreds of words)
* Discrete (recognises dictated speech with pauses between words)
* Continuous (recognises natural speech)

Dragon Naturally Speaking uses continuous voice recognition / natural speech.

Learning systems
The computer changes how it functions / reacts to situations based on feedback it receives.
E.g. If the computer doesn’t win a game, it remembers not to make those same moves.

Neural networks
A computer system simulates the functioning of a human brain.
Neural networks can process many pieces of data at once and learn to recognise patterns.
Special features:
* Ability to retrieve information even if some of the neural nodes fail
* Fast modification of stored data as a result of new information
* Ability to discover relationships and trends in large databases
* Ability to solve complex problems for which all the information is not present

Expert Systems

An overview of expert systems

Expert systems use heuristics (rules of thumb) to arrive at conclusions and make suggestions


Characteristics of an expert system
* Can explain their reasoning / suggested decisions
* Can display ‘intelligent’ behaviour (The computer makes suggestions, acting like an expert)
* Can draw conclusions from complex relationships
* Can provide portable knowledge (i.e. capture human expertise that otherwise might be lost)
* Can deal with uncertainty (uses probability, statistics, heuristics)


Limiting characteristics:
* Not yet widely used / tested
* Difficult to use
* Limited to relatively narrow problems
* Cannot readily deal with ‘mixed’ knowledge
* Possibility of error
* Cannot refine own knowledge base (Can’t eliminate redundant / contradictory rules)
* Difficult to maintain (adding new knowledge requires sophisticated programming)
* May have high development costs (Expert system shells can reduce costs = a collection of
software tools used to develop expert systems)
* Raise legal & ethical concerns

Modern developments
New Computers like IBM Watson have been humans in extremely complex events like Quizzes.


Capabilities of an expert system

Intelligent agents (software robots / bots) = special purpose knowledge-based information
systems that accomplish specific tasks for the user.
Strategic goal setting
Expert systems can suggest strategic goals and explore the impact of adopting them.

Planning
Can investigate the impact of plans, the ways they will help an organisation compete.

Design
Use design principles, understanding of procedures, and design rules to assist in design.

Decision making
Suggest alternatives, ways of looking at problems, and logical approaches to decision-making

Quality control and monitoring
Can monitor systems and propose solutions to system problems.

Diagnosis
Expert systems can analyse test results and patient symptoms.
Can provide the doctor with the probable cause of the problem and propose treatments.


When to use expert systems

* Provide high payoff / reduced downside risk
* Capture and preserve irreplaceable human expertise
* Develop a system more consistent than human experts
* Provide expertise needed at many locations at the same time / where it is dangerous
* Provide expertise that is expensive / rare
* Develop a solution faster than human experts can
* Provide expertise needed for training & development


Components of expert systems

Knowledge base
Stores all information, data, rules, cases, and relationships used by the expert system.
A knowledge base must be developed for each unique application.
If-then statements are rules that suggest certain conclusions.
Purpose of a knowledge base:
To hold the relevant facts and information for the specific expert system.
Assembling human experts:
It is a challenge to assemble the knowledge of multiple human experts.
Human experts can disagree on interpretations, presenting a dilemma for designers.


The use of fuzzy logic:
Fuzzy logic is  a research area that allows shades of grey (conditions are not true / false).
Used in embedded technology (auto-focus cameras, temperature sensors…)

The use of rules:
Rules relating data and conclusions can be developed for any knowledge base.

The use of cases:
Cases can be selected by comparing the parameters of the new problem with stored cases.

Inference engine

The component that delivers the expert advice.
The inference engine must find the right facts, interpretations and rules, and assemble them.

Some ways of accomplishing tasks:
1. Backward chaining:
 Starting with conclusions and working backwards to the supporting facts.
If the facts don’t support the conclusion, another conclusion is selected and tested.
2. Forward chaining:
 Starting with the facts and working forwards to the conclusions.
Comparing backward & forward chaining:

Forward chaining can reach conclusions and yield more info with fewer queries.
However, forward chaining requires more processing and sophistication.
Mixed chaining = a combination of backward & forward chaining.

Explanation facility
The expert system indicates all the facts and rules used in reaching the conclusion.

Knowledge acquisition facility
The part that provides the means of capturing & storing components of the knowledge base
Knowledge acquisition software can have easy-to-use menus, making maintenance easier.

User interface
There is specialised user interface software for designing, creating, updating, & using an ES.

Expert Systems Development

The development process
Determine requirements: Identify the system’s objectives and potential use.
Identify experts.
Participants in developing and using expert systems
* Domain expert
Domain expert is the person / group whose knowledge is captured.
* Knowledge engineer & knowledge users
Knowledge engineer is the  person with training in design & development of an expert system.
Knowledge user is person / group who use the expert system.


Expert systems development tools and techniques
* Expert system shells and products
Expert system shell is  a collection of software packages and tools used to design, develop,
implement, and maintain expert systems.

Advantages of expert system shells and products
* Easy to develop and modify
* Use of satisficing (giving good, but not necessarily optimal, solutions to complex problems)
* Use of heuristics (helps handle situations with imprecise relationships)
* Development by knowledge engineers and users

Expert systems development alternatives
* In-house development: develop from scratch
The most costly alternative, but you have more control over the features.
Disadvantage: Can result in a more complex system, with higher maintenance costs.

* In-house development: develop from a shell
The same shell can be used to develop many expert systems.
Easier to develop and less complex to maintain.
Disadvantage: The system may need to be modified and the features can be harder to control


* Off-the-shelf purchase: use existing packages
Least expensive approach.
Advantages: Cost, time, can also be easy to maintain and update.
Disadvantage: Might not satisfy unique organisation needs.


Applications of expert systems and AI
Credit granting (Expert systems at banks review credit applications…), games, information
management and retrieval, legal profession, AI and expert systems embedded in products,
plant layout, hospitals and medical facilities, help desks and assistance, employee
performance evaluation, loan analysis, virus detection, repair and maintenance, shipping,
marketing, warehouse optimisation.


Integrating expert systems
An expert system can be integrated with other systems through a common database.

Virtual Reality

Virtual reality systems enable users to move and react in a computer-simulated environment.
Users can sense and manipulate virtual objects like real objects.

Interface devices
Head mounted displays monitor the location of the head and the direction you’re looking.
BOOM: a head-coupled stereoscopic display device.
CAVE provides the illusion of immersion by projecting stereo images on walls.
Earphones update audio signals.
The haptic interface relays the sense of touch (e.g. by using a glove) - challenging to create.

Immersive virtual reality
The virtual world is presented in full scale.

Other forms of virtual reality

Applications that are not fully immersive:
E.g. Mouse-controlled navigation through a 3D environment on a screen.

Telepresence systems immerse you in a world captured by video cameras at a distance.


Useful applications

Medicine

Education

Real estate marketing (You can take a virtual walk through properties without wasting time). The virtual reality makes the individual feel like he is going around the property himself and seeing every feature.

Computer-generated images