What do Use Cases Look Like?

Use cases are the simplest kind of UML diagrams; they provide a high-level view of the tasks a user can carry out in a system. Use Cases make no attempt to describe low-level implementation or logic.

They consist of the following shapes:

• Stick-figures, to represent types of users ("Actors");
• Circles, to represent tasks (UML calls them "Use Cases" - a use case diagram basically displays a collection of use cases, which can then be drilled into to provide further diagrams...);
• Two kinds of line: those marked <uses>, and those marked <extends>.

The difference between <uses> and <extends> is important:

• <uses> means: to complete this task, the subtask must be completed, at least once (it might need to be completed many times, but you don't show this in the diagram); the arrow points from the task to the used subtask (use case);
• <extends> means: the subtask is a specialised version of the task it points to; it's the same base task, but is refined or extended in some way; the arrow points towards the use case that is being extended.

There's an excellent discussion on the difference between <uses> and <extends> at: http://www.andrew.cmu.edu/course/90-754/umlucdfaq.html#uediff

And that's it! They are very simple to create, and are there to communicate the functionality a system has, ie. what the user can do with it - so you won't be showing communication between systems, or any low-level applicaiton plumbing, and neither will you be showing error conditions or logic (which should go into an activity diagram).

Next: how to approach the creation of a use case.