Dynamic scheduling system for web monitoring/crawling system

TL;DRAbstract

This research studies the updating speed of web pages to provide a dynamic scheduling system for web monitoring. It combines two approaches; Multiple Classification Ripple-Down Rules (MCRDR) and Detector Constructor (DC-1). MCRDR is used to retrieve articles from web pages and classify them into folders. The DC-1 then checks if there is any unusual posting activity in these folders to inform the MCRDR to schedule a new revisiting time sooner than the originally scheduled time. This system aims to keep the user updated with fewer visiting times and less delay time between publishing (a change in web page) and collecting times (revisiting time).

Chat with Paper

AI Agents for this Paper