Most Active Users in a Live Communication Stream

Problem Statement

You are building an analytics service for a live communication platform. You are given a stream of events, where each event is a tuple (timestamp, user_id, message_id).

Your task is to implement a class ActivityTracker that processes these events and identifies the 'Most Active Users'. An active user is defined by the number of unique message_ids they have sent. If a message_id appears multiple times for the same user_id, it should only be counted once.

Implement the following methods:

1. record_event(timestamp, user_id, message_id): Registers a message for a user.
2. get_top_k(k): Returns a list of the top k user_ids based on their count of unique message_ids, ordered by count descending. If two users have the same count, break ties by returning the smaller user_id first.

Constraints

• user_id is an integer.
• message_id is an integer.
• The system must be able to handle millions of events.
• k will be a small positive integer.

Example

tracker = ActivityTracker()
tracker.record_event(1, 101, 1)
tracker.record_event(2, 101, 2)
tracker.record_event(3, 102, 1)
tracker.record_event(4, 101, 1) # Duplicate message_id, ignored

# User 101: 2 unique messages, User 102: 1 unique message
print(tracker.get_top_k(1)) # Output: [101]