CASE: Akershus University
Chapter 7: MANAGING KNOWLEDGE
AND ARTIFICIAL INTELLIGENCE 3) Problem: Information was in
unstructured, textual reports
4) Solution: By using artificial intelligence
technology from IBM Watson Explorer
DATA: Flows of events or transactions captured by an organization’s
which can analyze structured and
systems that are useful for transacting
unstructured data to uncover trends and
A firm must expend resources to organize patterns
data into categories of understanding 5) Purpose: To determine whether its CT
INFORMATION
scans fell within recommended
KNOWLEDGE guidelines
A firm must expend additional resources to
Apply where, discover patterns, rules, and contexts where
when, and how the knowledge works Decision-Making Process
WISDOM: The collective and individual experience of applying
knowledge to the solution of problems Discovering, identifying, and
understanding the problems
occurring in the organization
Identifying and exploring
various solutions and
coming up with alternatives
Choosing among the
solution alternatives
Making sure the chosen
alternative works and monitoring
how well the solution is working
High-Velocity Automated Decision Making
• Captured by computer algorithms
• Humans eliminated
• Predefined range of acceptable solutions
1) Tacit knowledge: Knowledge residing in the minds of • Decisions made faster than managers
employees that has not been documented can monitor and control
2) Explicit knowledge: Knowledge that has been • Goal-oriented direction
documented
Knowledge Management Value Chain
Knowledge
management:
The set of
business
processes
developed in
an
organization
to create,
store, transfer,
and apply
knowledge
, Major Types of Knowledge Management Systems
Artificial intelligence (AI)
Types of Decisions
Unstructured data:
• Decision maker must provide judgment
to solve problem
• Novel and nonroutine
• No well-understood or agreed-upon
procedure for making them
Semi-structured data:
• A part of the problem has a clear-cut
answer
• There is a method to follow
• Mix of certainty and uncertainty
Structured data:
• Repetitive and routine
• Involves a definite procedure
• All variables are clearly understood
Decision Support Systems (DSS): A business intelligence (BI) delivery platform
for super-users who want to create own reports and use sophisticated analytics
and models to make decisions
• Examples: What-if analysis, Sensitivity analysis, Backward sensitivity analysis,
Pivot analysis, Intensive modeling techniques
• Application:
Semi-structured data – Approving bank loans, Finger matching
Structured data – Predicting stock prices, Disaster relief management
, Artificial Intelligence (AI)
The attempt to build computer systems that think and act like humans
Major Types of AI
"!Expert systems: %$Genetic algorithms:
Capture tacit knowledge of experts as a set of rules Find the optimal solution for a specific problem by
that can be programmed so that a computer can examining a very large number of alternative solutions
assist human decision makers for that problem
• Rules are interconnected • Search among solution variables by changing and
• Number of outcomes is known and is limited reorganizing component parts using processes
• There are multiple paths to the same outcome based on evolution and then identifying the right
• Can consider multiple rules at the same time string
• Typically perform limited tasks • Used in optimization problems
(i.e. diagnosing a malfunctioning machine, (i.e. minimization of costs, efficient scheduling,
determining whether to get credit for a loan) optimal jet engine design)
• Knowledge base: Set of hundreds or thousands of
rules &$Natural language processing (NLP):
• Inference engine: The strategy to search through Understand and analyze human natural language
the knowledge base and formulate conclusions (i.e. Google Translate, spam filtering systems,
customer call center interactions, digital assistances,
!!Machine learning (ML): interacting with the car heating system)
Identify patterns in very large databases without
explicit programming although with significant human '$Computer vision systems:
training Emulate the human visual system to view and extract
• Learns to find patterns and relationships by information from real-world images
analyzing a large set of examples and making a • Create a digital map of an image and recognize this
statistical inference image in large databases of images in near real time
• Supervised learning: The system is “trained” by (i.e. Facebook’s DeepFace, self-driving cars, passport
providing specific examples of desired inputs and control at airports)
outputs identified by humans in advance
• Unsupervised learning: The system is asked to ( Robotics:
process the development database and report
Movable machines that can substitute for human
whatever patterns it finds
movements as well as computer systems
(i.e. Google search, recommender systems on
• Programmed to perform a specific series of actions
Amazon and Netflix)
automatically
• Often used in dangerous environments (i.e. bomb
#$Neural networks: detection), manufacturing processes, military
Find patterns and relationships in very large amounts operations (i.e. drones), and medical procedures
of data that’s too complicated and difficult for humans (i.e. surgical robots)
to analyze by using machine learning algorithms and
computational models loosely based on human )$Intelligent agents:
neurons
Work without direct human intervention to carry out
• Are pattern detection programs – learn patterns by
repetitive, predictable tasks
finding pathways, searching for relationships,
(i.e. chatbots, agent-based modeling applications that
building models, and correcting over and over
model the behavior of consumers, stock market,
again
supply chains, the spread of epidemics)
• Learning Rule – Identifies the optimal pathway
• Use built-in or learned knowledge base with some
• Trained by feeding it data inputs for which outputs
capable of self-adjustment (i.e. Siri)
are known
• Constructs a hidden layer of logic which processes
inputs and classifies them based on the experience * Major forces driving the evolution of AI:
of the model • Development of Big Data databases
(i.e. distinguishing between valid and fraudulent credit • Reduction in the cost of computer processing
card purchases) and growth in the power of processors
• Refinement of algorithms
“Deep learning” neural networks: • Significant investment from business and
Uses multiple layers of neural networks working in a governments
hierarchical fashion to detect patterns
• Expected to be self-taught
, Security Vulnerabilities
Chapter 9: SECURING
INFORMATION SYSTEMS
Malware
Computer virus:
A rogue software program that attaches itself to
other software programs or data files to be
executed
Reasons systems are vulnerable:
• Spread from computer to computer when
1) Information systems are integrated
humans take action
2) Devices are interconnected
3) The Internet is open to anyone
Worms:
4) Use of fixed Internet addresses with cable/DSL modems
Independent computer programs that copy
5) Hardware is not configured properly
themselves from one computer to other
6) Hardware is damaged by improper use or criminal acts
computers over a network
7) Errors in programming
8) Improper installation
Trojan horse: 9) Unauthorized changes
A software program that appears to be benign 10) Power failures and natural disasters
but then does something other than expected 11) Partnering with another company
• Not itself a virus because it does not replicate 12) Portability
• A way for viruses or other malicious code to be 13) Email, instant messaging, and peer-to-peer file-sharing
introduced into a computer system programs
14) Bluetooth and Wi-Fi networks
SQL injection attacks: 15) Unencrypted VOIP
Exploit vulnerabilities in poorly coded web
application software to introduce malicious
Wireless security challenges:
program code into a company’s systems and
• Radio frequency bands are easy to scan
networks
• Service set identifiers (SSIDs) can be picked up easily by
intruders’ sniffer programs
Ransomware:
• War driving – eavesdroppers drive by buildings or park
Tries to extort money from users by taking
outside and try to intercept wireless network traffic
control of their computers, blocking access to
files, or displaying annoying pop-up messages
Software Vulnerabilities
Spyware:
Install themselves surreptitiously on computers to Bugs:
monitor user web-surfing activity and serve up Software program code defects
advertising • Virtually impossible to eliminate all bugs
• Main source is the complexity of decision-making code
Keyloggers: • Can open networks to intruders
Record every keystroke made on a computer to
steal serial numbers for software to launch Zero-day vulnerabilities:
Internet attacks, gain access to email accounts, Holes in the software unknown to its creator which are
obtain passwords to protected computer exploited by hackers before the vendor becomes aware and
systems, or pick up personal information fixes it
Internal Threats: Employees
Patches:
Social engineering: Small pieces of software to repair flaws without disturbing
Seeking system access by tricking employees into the proper operation of the software
revealing their passwords by pretending to be
legitimate members of the company in need of Patch management:
information Tracking vulnerabilities, testing, and applying all patches
Chapter 7: MANAGING KNOWLEDGE
AND ARTIFICIAL INTELLIGENCE 3) Problem: Information was in
unstructured, textual reports
4) Solution: By using artificial intelligence
technology from IBM Watson Explorer
DATA: Flows of events or transactions captured by an organization’s
which can analyze structured and
systems that are useful for transacting
unstructured data to uncover trends and
A firm must expend resources to organize patterns
data into categories of understanding 5) Purpose: To determine whether its CT
INFORMATION
scans fell within recommended
KNOWLEDGE guidelines
A firm must expend additional resources to
Apply where, discover patterns, rules, and contexts where
when, and how the knowledge works Decision-Making Process
WISDOM: The collective and individual experience of applying
knowledge to the solution of problems Discovering, identifying, and
understanding the problems
occurring in the organization
Identifying and exploring
various solutions and
coming up with alternatives
Choosing among the
solution alternatives
Making sure the chosen
alternative works and monitoring
how well the solution is working
High-Velocity Automated Decision Making
• Captured by computer algorithms
• Humans eliminated
• Predefined range of acceptable solutions
1) Tacit knowledge: Knowledge residing in the minds of • Decisions made faster than managers
employees that has not been documented can monitor and control
2) Explicit knowledge: Knowledge that has been • Goal-oriented direction
documented
Knowledge Management Value Chain
Knowledge
management:
The set of
business
processes
developed in
an
organization
to create,
store, transfer,
and apply
knowledge
, Major Types of Knowledge Management Systems
Artificial intelligence (AI)
Types of Decisions
Unstructured data:
• Decision maker must provide judgment
to solve problem
• Novel and nonroutine
• No well-understood or agreed-upon
procedure for making them
Semi-structured data:
• A part of the problem has a clear-cut
answer
• There is a method to follow
• Mix of certainty and uncertainty
Structured data:
• Repetitive and routine
• Involves a definite procedure
• All variables are clearly understood
Decision Support Systems (DSS): A business intelligence (BI) delivery platform
for super-users who want to create own reports and use sophisticated analytics
and models to make decisions
• Examples: What-if analysis, Sensitivity analysis, Backward sensitivity analysis,
Pivot analysis, Intensive modeling techniques
• Application:
Semi-structured data – Approving bank loans, Finger matching
Structured data – Predicting stock prices, Disaster relief management
, Artificial Intelligence (AI)
The attempt to build computer systems that think and act like humans
Major Types of AI
"!Expert systems: %$Genetic algorithms:
Capture tacit knowledge of experts as a set of rules Find the optimal solution for a specific problem by
that can be programmed so that a computer can examining a very large number of alternative solutions
assist human decision makers for that problem
• Rules are interconnected • Search among solution variables by changing and
• Number of outcomes is known and is limited reorganizing component parts using processes
• There are multiple paths to the same outcome based on evolution and then identifying the right
• Can consider multiple rules at the same time string
• Typically perform limited tasks • Used in optimization problems
(i.e. diagnosing a malfunctioning machine, (i.e. minimization of costs, efficient scheduling,
determining whether to get credit for a loan) optimal jet engine design)
• Knowledge base: Set of hundreds or thousands of
rules &$Natural language processing (NLP):
• Inference engine: The strategy to search through Understand and analyze human natural language
the knowledge base and formulate conclusions (i.e. Google Translate, spam filtering systems,
customer call center interactions, digital assistances,
!!Machine learning (ML): interacting with the car heating system)
Identify patterns in very large databases without
explicit programming although with significant human '$Computer vision systems:
training Emulate the human visual system to view and extract
• Learns to find patterns and relationships by information from real-world images
analyzing a large set of examples and making a • Create a digital map of an image and recognize this
statistical inference image in large databases of images in near real time
• Supervised learning: The system is “trained” by (i.e. Facebook’s DeepFace, self-driving cars, passport
providing specific examples of desired inputs and control at airports)
outputs identified by humans in advance
• Unsupervised learning: The system is asked to ( Robotics:
process the development database and report
Movable machines that can substitute for human
whatever patterns it finds
movements as well as computer systems
(i.e. Google search, recommender systems on
• Programmed to perform a specific series of actions
Amazon and Netflix)
automatically
• Often used in dangerous environments (i.e. bomb
#$Neural networks: detection), manufacturing processes, military
Find patterns and relationships in very large amounts operations (i.e. drones), and medical procedures
of data that’s too complicated and difficult for humans (i.e. surgical robots)
to analyze by using machine learning algorithms and
computational models loosely based on human )$Intelligent agents:
neurons
Work without direct human intervention to carry out
• Are pattern detection programs – learn patterns by
repetitive, predictable tasks
finding pathways, searching for relationships,
(i.e. chatbots, agent-based modeling applications that
building models, and correcting over and over
model the behavior of consumers, stock market,
again
supply chains, the spread of epidemics)
• Learning Rule – Identifies the optimal pathway
• Use built-in or learned knowledge base with some
• Trained by feeding it data inputs for which outputs
capable of self-adjustment (i.e. Siri)
are known
• Constructs a hidden layer of logic which processes
inputs and classifies them based on the experience * Major forces driving the evolution of AI:
of the model • Development of Big Data databases
(i.e. distinguishing between valid and fraudulent credit • Reduction in the cost of computer processing
card purchases) and growth in the power of processors
• Refinement of algorithms
“Deep learning” neural networks: • Significant investment from business and
Uses multiple layers of neural networks working in a governments
hierarchical fashion to detect patterns
• Expected to be self-taught
, Security Vulnerabilities
Chapter 9: SECURING
INFORMATION SYSTEMS
Malware
Computer virus:
A rogue software program that attaches itself to
other software programs or data files to be
executed
Reasons systems are vulnerable:
• Spread from computer to computer when
1) Information systems are integrated
humans take action
2) Devices are interconnected
3) The Internet is open to anyone
Worms:
4) Use of fixed Internet addresses with cable/DSL modems
Independent computer programs that copy
5) Hardware is not configured properly
themselves from one computer to other
6) Hardware is damaged by improper use or criminal acts
computers over a network
7) Errors in programming
8) Improper installation
Trojan horse: 9) Unauthorized changes
A software program that appears to be benign 10) Power failures and natural disasters
but then does something other than expected 11) Partnering with another company
• Not itself a virus because it does not replicate 12) Portability
• A way for viruses or other malicious code to be 13) Email, instant messaging, and peer-to-peer file-sharing
introduced into a computer system programs
14) Bluetooth and Wi-Fi networks
SQL injection attacks: 15) Unencrypted VOIP
Exploit vulnerabilities in poorly coded web
application software to introduce malicious
Wireless security challenges:
program code into a company’s systems and
• Radio frequency bands are easy to scan
networks
• Service set identifiers (SSIDs) can be picked up easily by
intruders’ sniffer programs
Ransomware:
• War driving – eavesdroppers drive by buildings or park
Tries to extort money from users by taking
outside and try to intercept wireless network traffic
control of their computers, blocking access to
files, or displaying annoying pop-up messages
Software Vulnerabilities
Spyware:
Install themselves surreptitiously on computers to Bugs:
monitor user web-surfing activity and serve up Software program code defects
advertising • Virtually impossible to eliminate all bugs
• Main source is the complexity of decision-making code
Keyloggers: • Can open networks to intruders
Record every keystroke made on a computer to
steal serial numbers for software to launch Zero-day vulnerabilities:
Internet attacks, gain access to email accounts, Holes in the software unknown to its creator which are
obtain passwords to protected computer exploited by hackers before the vendor becomes aware and
systems, or pick up personal information fixes it
Internal Threats: Employees
Patches:
Social engineering: Small pieces of software to repair flaws without disturbing
Seeking system access by tricking employees into the proper operation of the software
revealing their passwords by pretending to be
legitimate members of the company in need of Patch management:
information Tracking vulnerabilities, testing, and applying all patches
