During design, the architect will ___ the database schema to reduce redundancy.
To normalise the schema means to structure tables to minimise duplication and dependency. Normalise is the precise term, behind "third normal form." Tidy, clean up, and order lack the technical meaning. Designers "normalise the schema to 3NF," so normalise the schema is the correct collocation.
2 / 5
For read-heavy analytics, the team chose to ___ the data, accepting some redundancy for speed.
To denormalise data means to deliberately introduce redundancy to speed up reads (fewer joins). Denormalise is the precise term, the opposite of normalise. Flatten up, spread out, and unfold are informal or imprecise. Engineers "denormalise for the reporting tables," so denormalise the data is the correct collocation.
3 / 5
Each order record will ___ the customer table via a foreign key.
To reference a table means to relate to it, typically via a foreign key. Reference is the precise relational term, behind "foreign key reference." Point to up, link over, and connect out are informal. Designers say "the order references the customer," so reference the table is the correct collocation.
4 / 5
The team will ___ an entity-relationship diagram before writing any migrations.
To model the data (or "model an ERD") means to design the conceptual structure of entities and relationships. Model is the precise term, behind "data modelling." Draw up, sketch out, and design over are informal or imprecise. Architects "model the domain before implementing," so model the data is the correct collocation.
5 / 5
To capture the many-to-many relationship, the team will ___ a join table between products and orders.
To introduce a join table means to add an associative table that resolves a many-to-many relationship. Introduce (or "add") is the standard term. Put in up, add over, and bring out are informal or grammatically odd. Designers "introduce a join table for the many-to-many," so introduce a join table is the correct collocation.