A Brief History of Non-Relational Databases - DATAVERSITY

• XML tags are used to record metadata in a file
• Metadata is used to identify values, can
capture the affinity among values of the same
entity, and can collect together a group of
entity instances
• Database tables have names and fields that
describe the attributes of the entity contained
in the table
• The data that quantitatively records each
property has a specific data type and is atomic

• There are 5 fundamental operations on
tables: select, project, union, difference,
product; these operations are the only ones
you need to create new tables from other
database tables
• Join is an especially useful operation that
associates information from separate tables
in new ways, based on matching fields
• Relationships are the key to associating
fields of the physical database

• The physical database resides on the disk
drive; it avoids storing data redundantly and
is optimized for speed
• The main approach for creating logical
views from physical data is the join-and-trim
• There is a direct connection between the
theoretical ideas of database tables and
the software of database systems.