Cognitive Complexity Analysis of Learning-Related Texts: A Case Study on School Textbooks

The proposed work analyzes the cognitive complexity of given text-based learning material. Cognitive complexity refers to the thinking skills that are required to process the information which is present in the content of the text. A standard methodology to identify cognitive complexity is the use of Bloom’s Taxonomy. However, as observed from the experiments that some of the action verbs are often present in multiple cognitive levels causing ambiguity about the true sense of cognition. To overcome this drawback, signal words of informational text structure has been used as an added feature. Based on both cognitive action verbs of Bloom’s Taxonomy and signal words together, a computational approach using the SVM model has been used for an experiment on the NCERT dataset. It was observed that using signal words as an additional feature has significantly improved the classification task as compared to using only Bloom’s Taxonomy action verbs.