Six Degrees of Collusion - Risk Scoring

Problem: Risk Scoring and Blocking Not every connected cluster is malicious. Each transaction log now contains a user, a device, a credit-card hash, and a risk score: user_id,device_id,credit_card,risk_score Users are still connected through shared device_id or shared credit_card, forming a Fraud Ring. Each user has a risk_score in [0, 100]. The same user always has the same risk score across all logs. If the average risk score of users in the Fraud Ring containing target_user is strictly greater than 75, return the string true; otherwise return false. Critical Exception: Trusted Users Users with risk score 0 are trusted users: They still maintain graph connectivity and may connect otherwise separate groups; They must be excluded from both the sum and the count when computing the average; Each unique user is counted only once. If the target user does not exist, or the target ring has no non-zero-risk users, return false. Input Format n transaction_1 transaction_2 ... transaction_n target_user Output Format Print true or false. Constraints 0 <= n <= 2 * 10^5 All fields are non-empty strings without commas 0 <= risk_score <= 100 A user may appear multiple times, but their risk score is consistent Example Input 3 Alice,D1,CC1,90 Bob,D1,CC2,0 Charlie,D2,CC2,80 Alice Output true