Page Recommendations II

Given timestamped user page-view logs, compute for each page the top K pages to recommend based on how frequently users navigate from that page to other pages (excluding itself), with deterministic tie-breaking and ordering. The task focuses on counting/co‑occurrence of consecutive visits and producing per‑page ranked top‑N results — typically solved by grouping/aggregation and ranking (e.g., using window functions or efficient streaming counts).