Efficient structure for querying membership in many-to-many relationship where tableA has ONLY the indicated members from tableB?

I’m working on an application with a PostgreSQL back end. There is a many-to-many relationship between items in one table and tags in another table. The specs require that the application be able to respond to frequent queries that provide a list of tags as parameters and expect in response a list of all items which contain only those tags but not necessarily all of those tags.

Is there an efficient (non-exponential complexity) way to structure my database (or application) to suit this task? At launch, I shall have on the order of 12K items and 1K tags.

E.g.
A user requests a list of all items for tags a, b, and c.
The response contains items with the following tag sets:

{}
{a}
{a,b,c}

The response must not contain items with the following tag sets:

{d}
{a,b,c,d}

The structure of tags table is as follows:

CREATE TABLE tags (
    id integer NOT NULL,
    name text NOT NULL,
    description text,
    revision_id integer
);