Download Completeness Estimation Using Redirect and Media Log Sampling

The patent applies partial action by using only the portion of media server log data that is necessary and accessible (redirect logs with sampled media server logs) rather than requiring complete access to all media server logs. This allows download completeness measurement to proceed with available data, accepting partial information while still achieving meaningful measurement results through statistical sampling and client behavior analysis.

The patent extracts only the essential and accessible components of media server log data (redirect logs and sampled media server logs) rather than requiring access to the complete log system. This extraction approach removes the complexity of full media server log access while retaining sufficient information for download completeness measurement through selective data collection and client software behavior analysis.

The patent implements feedback by analyzing client software behavior patterns and using this information to adjust download counting methodology. By monitoring how client software makes requests (monolithic vs. progressive) and applying this knowledge to interpret redirect log data, the system compensates for the complications introduced by progressive downloads and achieves accurate unique download counting despite the fragmented request pattern.