There are three main types: conceptual, logical, and physical data models. Conceptual models define what the system contains. Logical models represent how the system should be implemented. Physical models describe how the system will be implemented using a specific database management system.

Normalization is the process of organizing data to minimize redundancy. It is crucial for reducing the potential for anomalies and ensuring data integrity. Normalization typically involves dividing a database into two or more tables and defining relationships between them.

Many-to-many relationships are usually handled by creating a junction table that includes the primary keys of both tables as foreign keys. This way, a many-to-many relationship is broken down into two one-to-many relationships.

The steps typically include gathering requirements, creating an entity-relationship diagram, defining entities and relationships, normalizing the data model, and finally, reviewing the model with stakeholders to ensure it meets business requirements.

A star schema is a type of data warehouse schema that organizes data into fact tables and dimension tables in a simple, star-like structure. Each dimension is directly linked to the fact table. A snowflake schema is a more complex version where dimensions are normalized into multiple related tables, resembling a snowflake shape.

Optimizing a database involves several strategies, such as indexing, query optimization, partitioning, and understanding the workload patterns. Proper normalization, denormalization when necessary, and regularly updating statistics and indexes also contribute to optimization.