Business Intelligence (Midterm)

What is DATA MINING?

Data miss management and data removal

End result of successful DATA MINING PROCESS will have:

Data integration, data consolidation, data aggregation, decision management support

______ is a way of “mining” knowledge from large amounts of data

Data Mining

_____ draws information from many sources

Data Warehouse

By identifying interesting patterns, you perform the following: 

Pattern Evaluation 

Graphical User Interface, pattern evaluation, data mining engine, and data warehouse server are all directly using knowledge based services in typical data mining systems

False

Knowledge Base

a storage of helpful information for mining. decides what patterns matter and how to mine them.

Who uses the knowledge base?

Data mining engine and pattern evaluation module

Graphical User Interface

Uses user input.Talks to people, not to the knowledge base.

What does the data warehouse server use? 

Raw or cleaned data stored in the database 

Data Mart can be classified as large units in organizations such as offices or divisions

False

Database uniquely identifies each record through UNIQUE KEY, represents each record as transaction, holds highest level of details

Traditional and relational database

Traditional Database

simple tables, simple queries, old style

Relational Database

Stores data in tables (rows + columns) and links them with relationships. Excel sheets that can talk to each other

Object Oriented Database (OODB)

Stores objects like in programming (Data + methods together). Stores full “objects” like in java or python

Object Relational Database (ORDB)

Mix of relational tables plus object oriented features. Tables that can store complex data types and objects.

Specialized Database

Designed for specific types of data (maps, images, text, scientific data) custom database for special jobs

Heterogeneous Database

Combines different types of databases so they work together.

Predictive patterns perform simple calculations on current data and can classify new databases on already learned pattern

False

Predictive Pattern

a model used to predict unknown values or classify new data

Classification

putting new items into categories 

Clustering

grouping similar data points together based on similarity

Transactional Database

Stores daily business transactions like sales, orders, or payments

Euclidean Distance

A way to measure how far apart two points are in space (or data)

Database Management System (DBMS)

Software that stores, organizes, and manages databases