Preprocessing features of TADAEd Agathe Merceron Kalina Yacef

Preprocessing features of TADA-Ed Agathe Merceron Kalina Yacef

Login Qid Mistake Concept David. M 924 X A David. M 172 X C David. M 172 Y A David. M 172 Z D Mary. T 926 Y D Mary. T 154 X B John. S 924 Z B John. S 924 X A John. S 924 Y A John. S 924 X A John. S 165 Z C John. S 154 W D Login David. M Mary. T John. S ? ? … First we choose an attribute to aggregate the data. Let’s suppose we want to compare students We aggregate the data using the attribute ‘login’

Login Qid Mistake Concept David. M 924 X A David. M 172 X C David. M 172 Y A David. M 172 Z D Mary. T 926 Y D Mary. T 154 X B John. S 924 Z B John. S 924 X A John. S 924 Y A John. S 924 X A John. S 165 Z C John. S 154 W D Login Qid David. M 2 Mary. T 2 John. S 3 … Here we choose to only consider the number of exercises in which each student made one or more mistakes § Qid: Number of types: count the number of distinct value per common attribute value. § Chosen attribute to aggregate: login § Preprocessing performed: • Qid: Number of types

Login Qid Mistake Concept David. M 924 X A David. M 172 X C David. M 172 Y A David. M 172 Z D Mary. T 926 Y D Mary. T 154 X B John. S 924 Z B John. S 924 X A John. S 924 Y A John. S 924 X A John. S 165 Z C John. S 154 W D Login Qid mistake David. M 2 4 Mary. T 2 2 John. S 3 6 … Now we want to only consider the total number of mistakes each student made § Mistake: Total Number : count the number of value per common attribute value § Chosen attribute to aggregate: login § Preprocessing performed: • Qid: Number of types • Mistake: Total Number

Login Qid Mistake Concept David. M 924 X A David. M 172 X C David. M 172 Y A David. M 172 Z D Mary. T 926 Y D Mary. T 154 X B John. S 924 Z B John. S 924 X A John. S 924 Y A John. S 924 X A John. S 165 Z C John. S 154 W D Login Qid We want to know with what concepts the student made mistakes § Rule : Nominal: create one attribute per distinct value. § Chosen attribute to aggregate: login § Preprocessing performed: • Qid: Number of types • Mistake: Total Number • Rule : Nominal mistake A B C D David. M 2 2 1 0 1 1 Mary. T 2 2 0 1 John. S 3 3 1 1