Unit 9: Computing and Society

Count anonymous activity totals by category while avoiding storage of raw individual identifiers.

Given users, roles, and permissions, determine whether a user can access a resource through assigned roles.

Track failed logins over time and flag accounts that exceed a threshold within a fixed window.

Create a scoring function for security alerts using severity, confidence, and age, then discuss false positives.

Compare recommendation rates across anonymous groups and identify when a model may need additional review.

Detect repeated consent forms by user token and form version while avoiding display of private fields.

Remove records older than a retention window and prove the algorithm does not delete newer records.

Given message features, classify messages with rules and explain where the rule set may be too broad.

Given a dependency graph of services, determine which user-facing features fail when one service is down.

For a simple risk score, return the top contributing factors without revealing sensitive hidden data.

Given a dataset with uneven missing values, identify which summaries may be misleading.

Turn accessibility requirements into test cases for keyboard navigation, labels, contrast, and readable feedback.

Given software dependencies and vulnerability flags, find which applications depend on vulnerable packages.

Limit repeated requests by account or IP-style token and explain the user experience tradeoff.

Given a feature request, decide which fields are necessary, optional, or should not be collected.

Simulate how reports enter review queues and compare FIFO, priority, and hybrid approaches.

Given robotic sensor readings, identify values outside physical ranges and report the safest system response.

Check whether backups exist for each required system and whether the newest backup is recent enough.

Given downtime intervals, compute total availability and identify overlapping outages correctly.

Convert raw numeric activity values into bins and explain why bins reduce detail compared with raw records.

Find users who have both restricted and privileged roles and recommend a review workflow.

Give feedback on password strength without storing or printing the password itself.

Use graph levels to estimate how information could spread and discuss why real behavior is more complex.

Compare current input summaries with training summaries and flag large changes for human review.

Sort bug reports by severity and impact while avoiding rules that ignore accessibility issues.

Update records when a user withdraws consent and ensure future processing skips revoked records.

Given a form specification, identify fields that do not support the stated purpose.

Model an incident as detected, triaged, contained, resolved, and reviewed, with allowed transitions.

Given small group counts, explain why percentages can be unstable and what additional data or review is needed.

Create an assignment that combines algorithm design with privacy, bias, security, and written reflection.