Business Intelligence and Knowledge Management
This week we expand upon previous topics. We've already
discussed complex business systems that can gather a lot of data, and we've
also discussed databases and data warehouses. Once you have all that data
being collected or already stored, the next step is to take advantage of it
by analyzing it for potential opportunities. That's what this week's discussion
is about. The outline presented here is intended to help point out important
topics and terms, but is not intended to replace the lecture (or the text).
Objectives
- explain what data mining is
- explain what online analytical processing is
- describe how business intelligence can provide benefits to an organization
- describe how knowledge management can provide benefits to an organization
- explain why it is important to know about BI and KM applications
Business Intelligence
- BI: (Business Intelligence) software tools designed to find patterns and
extract useful information out of databases (primarily data warehouses)
- The two primary uses for data warehouses are data mining and online analytical processing.
- data mining: the process of selecting, exploring, and modeling large
amounts of data to discover previously unknown relationships that can
support decision making
- data mining has four objectives
- sequence or path analysis: finding out what events lead to other events
- classification: finding out what categories facts belong to
- clustering: finding previously unknown groups of related facts
- forecasting: finding patterns in data that can lead to future predictions
- a sampling of data mining applications (listed in the text):
- predicting customer behavior
- identifying common characteristics of customers who buy particular products
- finding out why customers switch to competitors
- identify patterns associated with fraud
- selecting potential customers for direct marketing campaigns
- choosing what items to display to an online shopper based on their previous
history and the history of similar customers
- discovering what items sell well together
- discovering trends
- OLAP: (OnLine Analytical Processing) provides an interface to the user
where large amounts of data can be filtered, categorized, summarized and displayed
quickly
- drilling down: a process by which one starts with tables containing broad
information and gradually retrieves tables containing more and more detailed information
- check out www.fedscope.opm.gov/
for a great example of an OLAP application which allows you to try out "drilling down"
- dimensional database: database organized so tables show summaries and ratios of
raw data; this speeds up processing time for OLAP applications
- multidimensional database: see dimensional database
- The text contains many examples of how various companies have used OLAP to detect
problems and opportunities. Read them.
- Other systems, such as CRM (Customer Relationship Management), often have data mining
and OLAP applications integrated into them
- dashboard: a name given to the user interface of BI tools when they present a
handful of results in a easily interpreted graphical format; gauges that look like
speedometers may have vertical indicators when everything is fine, but be off vertical
for non-optimal conditions; maps may have sections that show up in red when data
indicates a problem in an area; this way the user isn't always trying to analyze
the numbers
Knowledge Management
- KM: (Knowledge Management)
- KM is the combination of activities involved in gathering, organizing, sharing, analyzing,
and dissemination knowledge to improve an organization's performance. [This is directly
from the text]
- Organizations often have someone who has knowledge in an area, but they don't know
about it. Therefore, time and effort and expense are often wasted while another employee
seeks that knowledge elsewhere.
- Employee knowledge networks can be used to disseminate information about who has
expertise in a particular field. Often, the most important thing to know is where to find
the information that you need.
- The Web is one of best ways of disseminating (and collecting) knowledge. The biggest
problems with the Web are that the information may be totally unreliable and the
coverage of different areas of expertise is often inconsistent.
- autocategorization: great for classifying data being stored so it can be later
retrieved; in many ways, this is what Google already does with the Web
- automatic taxonomy: synonym for autocategorization
- knowledge worker: someone who works at developing or using knowledge