The Network Model is a data model that allows the representation of many-to-many relationships using a graph structure, consisting of nodes (entities) and edges (relationships) to organize data, enabling more efficient and direct access paths.
- This model is a generalization of the hierarchical model.
- It supports 1:1, 1:M (one-to-many), and M: N (many-to-many) data relationships.
- It can have multiple parent segments, which are grouped as levels.
Structure of a Network Model
Here, a relationship is called a set, and each set is made up of at least 2 types of records, which are given below:
- An owner record that is the same as of parent in the hierarchical model.
- A member record that is the same as of child in the hierarchical model.
In the above figure, member TWO has only one owner 'ONE' whereas member FIVE has two owners i.e, TWO and THREE. Here, Network model implements many-to-many relationships using multiple 1:M sets. This structure helps avoid data redundancy because multiple paths can access the same record.
- This model consists of both lateral and top-down connections between the nodes.
- There are various examples such as TOTAL by Cincom Systems Inc., EDMS by Xerox Corp., etc.
Example 1 : Network model for a Finance Department.
Below we have designed the network model for a Finance Department:
A one-to-many (1: N) relationship has a link between two record types. The above figure, SALES-MAN, CUSTOMER, PRODUCT, INVOICE, PAYMENT, INVOICE-LINE are the types of records for the sales of a company. The figure represents INVOICE-LINE is owned by PRODUCT and INVOICE. INVOICE has also two owners SALES-MAN and CUSTOMER.
Example 2: We have two segments, Faculty and Student. Say that student John takes courses both in CS and EE departments. Find how many instances will be there?
A student's instance can have at least 2 parent instances. There exist relations between the instances of students and faculty segment.
- The model can be very complex as if we use other segments say Courses and logical associations like Student-Enroll and Faculty-course.
- A student can be logically associated with various instances of Faculties and Courses.
Features
- Data Relationship Representation: It allows many-to-many relationships, providing greater flexibility in how data is connected.
- Records and Sets: Records are similar to rows in a relational table, and sets are used to define relationship between records.
- Owner-Member Relationships: An owner record can be linked to multiple member records, and a member record can belong to multiple owner records, facilitating complex relationships.
- Navigational Access: This is different from relational models, which use declarative query languages like SQL.
- Hierarchical and Non-Hierarchical Structures: It represents both hierarchical (tree-like) and non-hierarchical (graph-like) structures, providing flexibility in data modeling.
Advantages
- This model is more flexible than the hierarchical model but more complex to design.
- It is capable of handling multiple types of relationships which can help in modeling real-life applications.
- This network does not allow a member to exist without an owner which leads to the concept of Data integrity.
- Like a hierarchical model, this model also does not have any database standard.
Disadvantages
- The design or the structure of this model is not user-friendly and does not have any scope of automated query optimization.
- Once the the database of the model is formed it is difficult to change.
- There's an existence of operational anomalies as there is a use of pointers for navigation which further leads to complex implementation.
- It fails in achieving structural independence even though the network database model is capable of achieving data independence.
Operations
- Insertion: Adding new records and establishing owner-member relationships.
- Deletion: Removing records and maintaining data integrity by handling related records and relationships.
- Update: Modifying existing records and relationships between records.
- Traversal: Navigating through the network structure to access related records using predefined paths.
- Search: Retrieving specific records based on criteria by navigating the network structure.
Network Model vs Hierarchical Model
| Feature | Hierarchical Model | Network Model |
|---|---|---|
| Structure | Tree-like structure | Graph structure |
| Relationships | One-to-many (single parent, multiple children) | Many-to-many (multiple parents and children) |
| Flexibility | Less flexible | More flexible |
| Data Access | Single access path | Multiple access paths |
| Redundancy | Higher redundancy due to rigid hierarchy | Lower redundancy due to shared relationships |
| Complexity | Simpler to design and implement | More complex to design and manage |
| Usage Scenario | Suitable for simple, hierarchical data structures | Suitable for complex, interconnected data structures |
| Efficiency | Efficient for hierarchical traversal | Efficient for complex queries and data retrieval |
| Example | Organizational chart | Telecommunications network |