Chen Wen-xiang

Notice of Retraction A Stepwise Augment Algorithm for the Construction of Reduced Q-Matrix

Cognitive Diagnosis is more and more important for education and has been a research focus. Tatsuokas Q-matrix theory has been widely used in Cognitive Diagnosis, in which Reduced Q-matrix (Qr matrix) plays an important role. Two methods generate Qr matrix are considered. One is Tatsuokas method, another is augment algorithm. Based on augment algorithm, a stepwise augment algorithm is proposed. Considering the number of valid items, the stepwise augment algorithm is compared with Tatsuokas method and augment algorithm upon running time. Mathematical models with 8 attributes are built for the three algorithms by linear regression analysis.

An augment algorithm based on sufficient formal context

Formal Concept Analysis(FCA) is one of the powerful tools for cognitive science. A concept lattice induced by a context is the core data structure of FCA. A formal context expresses the incidence relation between objects and attributes. It is obvious that the constraint relation between attributes exists. An attribute hierarchy constituted by the attribute relation inevitably restricts the formation of context. Based on an attribute hierarchy, studying the property of the attribute hierarchy, sufficient formal context is defined, the augment algorithm of generating sufficient context is given, and its performance is shown.