Notes > Foundations of Computing > Data Modelling and Mining
Data modelling is a technique which is used to visually describe the structure of a database's information model. A data model allows users to understand the system and how objects within it relate to each other. There are 3 main types of data model, each showing different amounts of detail regarding the objects within the system and their attributes and relationships.
A conceptual data model involves the laying out of the objects on a diagram then lines being drawn between them to show how each of the objects relates to others.
A logical data model specifies the structure of the database system by having a set of tables which are linked to each other with arrows. Each table represents an object and it contains the attributes (or fields) associated with that object.
A physical data model describes how the logical data will be represented in storage i.e. the data type of each field (e.g. string, integer etc...)
The relationships that exist between entities can be represented within data models. The different types of relationships are described below.
- One-to-one - Both of the entities that are linked can only be in a single relationship and therefore may only have one related entity each.
- One-to-many - One entity will have multiple relationships with the other entities involved in the relationship. All these other entities may only have a single relationship with another entity. For example, one mother may have many children. Each child can only have a single relationship with their natural mother but the mother can have multiple relationships with her children.
- Many-to-many - Each entity may participate in any number of relationships. For example, a customer at a library may rent out different books and each book may be rented by different customers.
Within data models, keys are used to identify and reference records. A key is an attribute that has a unique value for each entity. The primary key is the main key used to identify single entities. A foreign key is the primary key of another type of entity. These foreign keys are used to identify relationships between entities.
Data mining is the process of extracting information from a collection of data. There are two important classes of problems within data mining and these are classification rules and association rules.
Classification rules define the groupings that exist within a collection of data i.e. which pieces of data relate to other pieces of data. Association rules involve the identifying of correlations among sets of items. An example of an association rule is as follows: "if a customer bought object A, then in X% of transactions, they also bought object B".
Search for "Data Modelling" on:
eBay (UK) |
Search for "Data Modelling" on the rest of Computing Students: Data Modelling