User Likes Tracker - Efficiently Track and Query User Content Likes

Question

## Round 1 - Coding / OOD

## Problem

Design a `UserLikesTracker` that records when users like or unlike content items, and supports efficient queries. Each like/unlike event has a user ID, item ID, and timestamp.

```python
class UserLikesTracker:
    def like(self, user_id: str, item_id: str, ts: int) -> None:
        pass

def unlike(self, user_id: str, item_id: str, ts: int) -> None:
        pass

def is_liked(self, user_id: str, item_id: str) -> bool:
        pass

def liked_items(self, user_id: str) -> list[str]:
        # Returns all item_ids currently liked by user, sorted by like timestamp desc
        pass

def top_liked_items(self, k: int) -> list[tuple[str, int]]:
        # Returns top-k items by total current likes, desc
        pass
```

## Example

```
tracker.like("u1", "post_a", ts=100)
tracker.like("u2", "post_a", ts=101)
tracker.like("u1", "post_b", ts=102)
tracker.unlike("u1", "post_a", ts=103)

tracker.is_liked("u1", "post_a")    # -> False
tracker.liked_items("u1")           # -> ["post_b"]
tracker.top_liked_items(1)          # -> [("post_a", 1)]
```

## Follow-ups

1. How do you make `top_liked_items` O(log k) instead of O(n log n) on each call?
2. How do you handle late-arriving events where unlike arrives before like due to network delays?
3. At 1M users and 100M items, how do you store like state efficiently in memory?
4. How do you expose a feed of "items liked by people you follow" with acceptable latency?

User Likes Tracker - Efficiently Track and Query User Content Likes

Interview Experience

Full Details

About This Question

Topics

More Lead Bank Interview Questions