Using syntactic text analysis to estimate educational tasks’ difficulty and complexity

Abstract

We suggest a routine for automatic estimation of complexity and difficulty of educational tasks. This routine is based on syntactic text analysis, phrases’ predicative structures identification, and semantic network construction. Then we develop a mathematical model which employs a notion on semantic distance between notions–words to calculate the amount of knowledge in a semantic network. We show that the amount of knowledge in a semantic network is a measure in the set of all semantic networks, and the semantic distance makes this set the metric one.