Supplementing extraction rules based on event clustering

This invention helps people make sense of large amounts of raw machine data, like logs or sensor readings. It first groups similar data events into clusters using existing rules, then shows these clusters to a user. Based on what the user sees, they can provide a new rule to refine how the data is grouped, making the system smarter at identifying patterns. The core idea is to iteratively improve how data is extracted and organized based on user feedback.