Lecture 1

Question 1

decision support systems

Answer 1

any computerized system that processes and analyzes data and supports decision-making in an organization

Question 2

business intelligence (BI)

Answer 2

• an umbrella term that combines architectures, tools, databases, analytical tools, applications, and methodologies
• information and knowledge that enables decision making
• relates to understanding preferences, coping with competition, identifying opportunities, enhancing efficiency
• uses tools such as data warehousing, knowledge management, queries, analysis, data mining, visualization

Question 3

business analytics (BA)

Answer 3

transforming data into meaningful information or knowledge to support business decision-making

Question 4

data

Answer 4

• internal or external
• structured or unstructured (more than 80%)
• items that are the most elementary descriptions of things, events and transactions

Question 5

DSS types

Passive DSS

Answer 5

Supports decision-making processes, but it does not offer
explicit suggestions on decisions or solutions

Question 6

DSS types

Active DSS

Answer 6

Offers suggestions and solutions

Question 7

DSS types

Collaborative DSS

Answer 7

Operates interactively and allows decision-makers to
modify, integrate, or refine suggestions given by the system.
Suggestions are sent back to the system for validation

Question 8

DSS types

Model-driven DSS

Answer 8

Enhances management of statistical, financial, optimization,
and simulation models.

Question 9

DSS types

Communication-driven DSS

Answer 9

Supports a group of people working on a common task.

Question 10

DSS types

Data-driven DSS

Answer 10

Enhances the access and management of time series of
corporate and external data.

Question 11

DSS types

Document-driven DSS

Answer 11

Manages and processes nonstructured data in many formats

Question 12

DSS types

Knowledge-driven DSS

Answer 12

Provides problem-solving features in the form of facts, rules,
and procedures

Question 13

information

Answer 13

organized data that has meaning and vakue

Question 14

knowledge

Answer 14

processed data or information that is applicable to a business decision problem

Question 15

BA method types

descriptive analytics

Answer 15

• use data to understand past & present
• results in well defined problems and opportunities
• “what happened and what is happening?”
• business reporting, dashboards, scoreboards, data warehousing, OLAP, performance dashboard

Question 16

BA method types

predictive analytics

Answer 16

• predict future behavior (states and conditions) based on past performance
• “what will happen and why will it happen?”
• data mining, text mining, web/media mining, forecasting

Question 17

BA method types

prescriptive analytics

Answer 17

• make decisions or recommendations to achieve the best performance
• “what should I do and why should I do it?”
• optimization, simulation, decision modeling, expert systems/knowledge based systems/rule based systems, (AI)

Question 18

descriptive analytics

performance dashboard

Answer 18

provides a comprehensive visual view of corporate performance measures, trends, and exceptions

Question 19

descriptive analytics

OLAP

Answer 19

online analytical processing

Question 20

OLTP

Answer 20

online transactional processing

Question 21

ETL staging

Answer 21

• extraction (getting the data)
• transformation (cleaning the data)
• loading (storing the data in a relevant environment)

getting the raw data to workable data is the most time consuming part of BI/BA/DSS; other difficulties with ETL are:
* temporary storage of data
* very complex data
* development and maintenance are time consuming
* extract & load frequencies

Question 22

data warehousing

Answer 22

a collection of methods, techniques, and tools used to support people to conduct data analyses that help with performing decision-making processes and improving information resources

Question 23

metrics

Answer 23

Perspectives (time, location, product, etc) that are relevant for management, and are used in multi-dimensional models.
We try to prevent information overload (about 6-7 metrics because of cognitive limitations)

Question 24

data warehouse process

Answer 24

a set of tasks that turn operational data into decision-making support information
* accessibility; to users not very familiar with IT and data structures
* integration of data; on the basis of a standard enterprise model
* query flexibility; to maximize the advantages obtained from the existing information
* information conciseness; allowing for target-oriented and effective analyses
* multidimensional representation; giving users an intuitive and manageable view of information
* correctness and completeness; of integrated data

Question 25

data warehouse

Answer 25

a collection of data that supports decision-making processes * subject-oriented (for client's analytic needs, NOT daily operations/transaction processing, excludes data that is not useful for decision making) * integrated (uses multiple data sources to provide a unified view of all data -> cleans and integrates data -converts it to data warehouse- to ensure consistency) * time-variant: displays snapshots of history (evolution over a long ass time) * is not volatile (read-only database; no operational updates of data - only loading and accessing) a data warehouse exists of tables and relationships between tables ("key relationships"), with a new database next to existing ones purely for decision making

Question 26

DBMS

Answer 26

database management system

Question 27

operational database

Answer 27

software that is designed to allow users to easily define, modify, retrieve, and manage data in real-time

Question 28

operational database properties

Answer 28

* thousands of users * workload consists of preset transactions * can access and write to hundreds of records * goal depends on applications * data is detailed, and both numeric and text * data integration is application based * quality graded on integrity * only current data * continuous updates * normalized model * optimized for OLTP access to subset of database

Question 29

data warehouse properties

Answer 29

* hundreds of users * workload consists of specific analysis queries * can access millions of records, mostly read-only * goal is decision-making support * RDBMS * data is summed up, mainly numeric * data integration is subject based (on the wishes of the decision makers) * high quality, graded on consistency * current and historical data * periodical updates (no up-to-date data) * denormalized and multidimensional models (+star/snowflake models) * pre-aggregation/pre-clustering/unstructured * optimized for OLTP access to most of database

Question 30

data warehouse needs to be a seperate database because?

Answer 30

* historic data: decision systems requires historical data which operational databases do not typically maintain * data integration & consolidation: data warehousing requires aggregation (clustering) & summarization of data from different heterogenous sources * data quality: different sources use inconsistent data representations, codes and formats which have to be reconciled for decision-making and also queries on ERPs are too taxing

Question 31

ERPs consist of

Answer 31

everyday operational/transactional processes of the company; one of the data sources in a data warehouse * different platforms (every side and vendor has their own dialect of SQL) * different databases * process/product oriented (not decision oriented) * internal/external data * structured/unstructured data * no integrated data * inconsistent data/low quality * limited historic data * transaction processing performance is key (no space for analytical/BI queries!!)

Question 32

RDBMS

Answer 32

relational database management system

Question 33

# data mart dependent/independent

Answer 33

If the data mart feed is dependent on the central data warehouse, the data is synced. If not, the data mart is often connected to some other component (XML) on the back-end.

Question 34

# data warehouse framework data mart

Answer 34

* focused on a single subject or line of business * subset of data warehouse (less dimensions, less history, less detail, less sources) * user group/application oriented * dependent & independent variants: ROLAP database & MOLAP multi-dimensional cube

Question 35

# data warehouse framework BI front-end applications

Answer 35

* query & reporting * OLAP (ROLAP, MOLAP, HOLAP, etc) * data mining (statistics, decision trees, neural networks, etc) * data visualization (graphs, animation, etc) trends are integrated BI-suits and enterprise information portals

Question 36

# data warehouse framework meta data/master data

Answer 36

very important * data about data: location, meaning, applied transformations, update frequency, access rights, data models, view definitions, etc * different heterogenous sources (terrible for maintenance) alternatives: shared repository or XML (interchange using standards)

Question 37

churning

Question 38

propensity to buy

Question 39

fraud detection

Question 40

customer segmentation

Question 41

single-layer architecture

Question 42

two-layer architecture