Providing targeted, evidence-based recommendations to improve content by combining static analysis and usage analysis

A computer-implemented method, comprising:

• partitioning digital content stored on the memory device into a sequence of logical content modules;
• utilizing quality scores for the logical content modules along a plurality of quality dimensions, to tag a quality score associated with a quality dimension to a content module in the sequence of logical content modules;
• capturing raw events generated from user interaction with an instrumented content access platform in accessing the digital content;
• detecting and transforming raw event patterns for the logical content modules over a plurality of users, into a set of dynamic states representing potential user experience and performance;
• correlating the quality dimensions of the logical content modules and associated quality scores with the dynamic states;
• and
• given a dynamic state of a user consuming a content module and responsive to finding that the dynamic state does not correlate to at least one of the quality scores associated with the content module, automatically inferring at run-time by a processor, a correlation of the dynamic state to a quality score associated with a pre-requisite content module occurring prior to the content module in the sequence of logical content modules.

A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a device to cause the device to:

• partition digital content stored on the memory device into a sequence of logical content modules;
• utilize quality scores for the logical content modules along a plurality of quality dimensions, to tag a quality score associated with a quality dimension to a content module in the sequence of logical content modules;
• capture raw events generated from user interaction with an instrumented content access platform in accessing the digital content;
• detect and transform raw event patterns for the logical content modules over a plurality of users, into a set of dynamic states representing potential user experience and performance;
• correlate the quality dimensions of the logical content modules and associated quality scores with the dynamic states;
• and
• given a dynamic state of a user consuming a content module and responsive to finding that the dynamic state does not correlate to at least one of the quality scores associated with the content module, automatically infer at run-time by a processor, a correlation of the dynamic state to a quality score associated with a pre-requisite content module occurring prior to the content module in the sequence of logical content modules.