GarryLaly/BetterSocial-test

Jawaban C1

Struktur Data

- profileId: number
- profilePhoto: string
- profileName: string
- followedBy: [
      {
         profileId: number
         profileName: string
      }
   ]

Handling Issue

Jika tidak ada pengguna terkait, simple nya tidak menampilkan section tersebut. Namun jika terlalu banyak pengguna yang terkait, bisa dilakukan indexing disisi API dan disisi mobile hanya mengambil data misalnya 10 saja yang terkait. Sehingga bisa meringankan beban load API

Testing

Secara manual bisa dilakukan pendaftaran user A, user B dan user C. Kemudian user A follow user C dan user B follow user C, sedangkan user A dan user B belum melakukan follow. Maka dalam case ini harusnya section "disarankan untuk Anda" ini bisa muncul di user A atau di user B untuk bisa saling follow

Jawaban C2

Tiap user menyimpan favorit komunitasnya

user_profiles = { "user1": ["travel", "photography", "food"], "user2": ["fashion", "fitness", "music"], "user3": ["food", "cooking", "photography"], # ... }

Tiap user menyimpan user yang difollow

user_connections = { "user1": ["user2", "user3"], "user2": ["user1"], "user3": ["user1"], # ... }

Function untuk suggest siapa yang akan difollow

const suggest_users_to_follow = (user)
suggested_users = []

    # Find users with similar interests
    for profile, interests in user_profiles.items():
        if profile != user and len(set(interests) & set(user_profiles[user])) > 0:
            suggested_users.append(profile)

    # Remove already followed users
    suggested_users = [u for u in suggested_users if u not in user_connections[user]]

    # Rank suggested users based on some criteria (e.g., common interests, followers, engagement)
    suggested_users = rank_users(suggested_users)

    # Return a list of suggested users
    return suggested_users

Function untuk rank algoritma

const rank_users = (users)
ranked_users = []

for user in users:
   common_interests = len(set(user_profiles[user]) & set(user_profiles[current_user]))
   follower_count = len(user_connections[user])

   # Calculate a score based on common interests and followers
   # You can adjust the weights for each factor as needed
   score = (common_interests * 0.6) + (follower_count * 0.4)

   ranked_users.append((user, score))

# Sort users by their scores in descending order
ranked_users.sort(key=lambda x: x[1], reverse=True)

# Extract the sorted users from the list of tuples
sorted_users = [user[0] for user in ranked_users]
return users

Penggunaan function

current_user = "user1"
suggested_users = suggest_users_to_follow(current_user)
print("Suggested users to follow for", current_user, ":", suggested_users)