KDD Cup 2010 Student Performance Evaluation Dataset

Source

Action-Time-Series Notes

  • Dataset type: intelligent tutoring competition dataset.
  • Temporal structure: student-step logs from Cognitive Tutor systems.
  • Actions or interventions: student responses, opportunities, problem steps, hints/attempt-related fields depending on file.
  • Suitability: useful for education dynamics and knowledge tracing; not a clean controllable-action benchmark.

Foundation TSFM Relevance

Agenda slotVerdictEvidenceMissing pieces
Irregular event streamsadjacentThe raw dataset snapshot describes student-step logs with responses, opportunities, problem steps, hints, and attempts.Education logs are not continuous multivariate telemetry and need a modern event-schema wrapper.
Causal and control modelingwarningHints and tutor interactions look intervention-like, but the local suitability note marks the dataset as not a clean controllable-action benchmark.Needs logged tutor policies, candidate actions, and counterfactual learning outcomes.