Time-Aware Social Network Follow Graph with Recommendations

Problem: Time-Aware Social Network Follow Graph + Bidirectional Queries + Recommendations Design and implement a data structure to maintain follow relationships between users, where a follow becomes active starting at a given time. Required operations update(A, B, t): user A starts following B at time t. check(A, B, t): at time t, determine whether A follows B. Follow-up 1: Bidirectional queries Extend the structure to support reverse-direction queries (e.g., who follows B at time t, or whom B follows at time t, as defined by the interviewer). Follow-up 2: Recommend users (by number of two-hop mediators) Implement recommend(A) to suggest candidates: If there exists a path A -> B -> X (A follows B and B follows X), then X is a candidate. If a candidate X is reachable via multiple distinct mediators (e.g., both A->B->C and A->M->C), it should rank higher. Example: given A -> B -> C, A -> B -> D, A -> M -> C, recommend C and D to A, with C ranked ahead of D. Notes Clarify whether repeated updates, unfollow operations, and monotonic timestamps are allowed. Clarify whether recommendations are time-dependent (computed on the graph at a query time t).