Which data structure is commonly used in Kahn's algorithm to manage the next node with zero in-degree?

Which data structure is commonly used in Kahn's algorithm to manage the next node with zero in-degree?

• A. Stack
• B. Queue
• C. Priority queue
• D. Hash map

Answer: (B)

In Kahn's algorithm, you keep track of vertices whose in-degree is zero and are ready to be processed. The natural way to handle this ready set is with a queue: a First-In, First-Out structure. As soon as a vertex becomes ready, you enqueue it, and you process vertices in the order they were added. This ensures a clean, predictable progression: you remove a vertex, append it to the topological order, and reduce the in-degree of its neighbors, enqueueing any neighbor that hits zero in-degree. The FIFO behavior keeps the processing flow simple and linear, and guarantees you never process a node before all its prerequisites have been handled. Using a stack would change the order of processing (LIFO), which is still valid for a topological sort in many cases but makes the sequence less predictable and harder to reason about. A priority queue would impose an extra, arbitrary ordering among ready nodes and adds overhead without changing the fundamental correctness. A hash map doesn’t manage the ready list at all.

In Kahn's algorithm, you keep track of vertices whose in-degree is zero and are ready to be processed. The natural way to handle this ready set is with a queue: a First-In, First-Out structure. As soon as a vertex becomes ready, you enqueue it, and you process vertices in the order they were added. This ensures a clean, predictable progression: you remove a vertex, append it to the topological order, and reduce the in-degree of its neighbors, enqueueing any neighbor that hits zero in-degree. The FIFO behavior keeps the processing flow simple and linear, and guarantees you never process a node before all its prerequisites have been handled.

Using a stack would change the order of processing (LIFO), which is still valid for a topological sort in many cases but makes the sequence less predictable and harder to reason about. A priority queue would impose an extra, arbitrary ordering among ready nodes and adds overhead without changing the fundamental correctness. A hash map doesn’t manage the ready list at all.