Educational systems typically contain a large pool of items (questions, problems). Using data mining techniques we can group these items into knowledge components, detect duplicated items and outliers, and identify missing items. To these ends, it is useful to analyze item similarities, which can be used as input to clustering or visualization…