[Analytics Reasoning] Impact of Malicious Accounts on Meta

Context & Assumptions On a large social network (similar to Facebook), only 1% of accounts are malicious or "bad" Bad accounts send friend requests at a rate ten times higher than good accounts You have developed a classification model that achieves a 95% true positive rate (TPR) and a 95% true negative rate (TNR) Questions 1. Single Friend Request Probability Estimate the probability that a received friend request comes from a bad account. 2. Multiple Friend Requests Determine the likelihood that, out of five friend requests, at least one originates from a bad account. 3. Model Reliability If the model flags an account as bad, how likely is it to be truly malicious, given the stated TPR and TNR? 4. Data & Features Identify which types of data (e.g., behavior logs, friend-request patterns, reported incidents) would be most relevant to classify accounts accurately. 5. Assessing "Bad Account" Prevalence Propose methods (such as stratified or random sampling) to determine whether the bad-account issue is substantial enough to warrant further intervention. 6. Defining a "Bad User" Outline what characteristics or behaviors would qualify an account as "bad" (e.g., spammer, scammer, bot). 7. Platform Impact Discuss how the presence of malicious users could affect the platform's trust, user experience, and reputation. 8. Friend Request Implications Examine how frequent friend requests from bad accounts might impact legitimate users and overall community health. Follow-Up Considerations Threshold Tuning Discuss how adjusting the classification threshold could alter false positives and false negatives. Cost-Benefit Analysis Evaluate the resource investment needed for automated detection vs. manual review. Long-Term Effects of False Positives Consider how incorrect labeling of legitimate users might erode trust. Advanced Feature Engineering Suggest additional signals (e.g., content quality, login behavior) that could enhance detection accuracy. Scaling & Fairness Explain how to maintain model performance and fairness across billions of accounts, ensuring no disproportionate impact on certain demographics.