Knowledge Tracing with Sequential Key-Value Memory Networks

Abstract

Can machines trace human knowledge like humans? Knowledge tracing (KT) is a fundamental task in a wide range of applications in education, such as massive open online courses (MOOCs), intelligent tutoring systems, educational games, and learning management systems. It models dynamics in a student's knowledge states in relation to di erent learning concepts through their interactions with learning activities. Recently, several a empts have been made to use deep learning models for tackling the KT problem. Although these deep learning models have shown promising results, they have limitations: either lack the ability to go deeper to trace how speci c concepts in a knowledge state are mastered by a student, or fail to capture long-term dependencies in an exercise sequence. In this paper, we address these limitations by proposing a novel deep learning model for knowledge tracing, namely Sequential Key-Value Memory Networks (SKVMN). is model uni es the strengths of recurrent modelling capacity and memory capacity of the existing deep learning KT models for modelling student learning. We have extensively evaluated our proposed model on ve benchmark datasets. e experimental results show that (1) SKVMN outperforms the state-of-the-art KT models on all datasets, (2) SKVMN can be er discover the correlation between latent concepts and questions, and (3) SKVMN can trace the knowledge state of students dynamics, and a leverage sequential dependencies in an exercise sequence for improved predication accuracy.

Meaningful Answer Generation of E-Commerce Question-Answering

“…Similar to our e-commerce question answering task, several tasks input data in key-value structure instead of a sequence. In order to utilize these data when generating text, key-value memory network (KVMN) [2,75] is purposed to store this type of data. He et al [31] incorporate copying and retrieving knowledge from the knowledge base stored in KVMN to generate natural answers within an encoder-decoder framework.…”

“…To extract the semantic features from each review, we first employ an CNN with a max-pooling operation, then apply a Selective Reading Unit (SRU)-based RNN to obtain final representation for each review. To begin with, a list of kernels with different width are used in the CNN operation, and their outputs are concatenated together, denoted as h r n,t in Equation (2). These different kernels capture different n-grams features.…”

An intelligent knowledge system for designing, modeling, and recognizing the behavior of elderly people in smart space

“…Many techniques based on machine learning and pervasive computing were used to achieve a better prediction of human activities and scenarios. Ta Minh et al (2018) and Abdelrahman and Wang (2019) proposed a key-value modeling. This type of knowledge modeling is based on the simplest data structure to describe a given activity based on flexible units which represent sensor data.…”

R2de

“…Recent literature claims that Deep Knowledge Tracing (DKT) -which was first introduced by Piech et al in [19] and consists in performing KT by means of neural networks -outperforms logistic models in predicting the results of future exams [1,6,32,33], but this advantage is not agreed across the community [7,18,28,31]. Also, DKT does not estimate explicitly the skill level of students nor the latent traits of questions, which makes the interpretation of such models a strenuous task.…”