Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance Prediction
Knowledge tracing (KT) serves as a primary part of intelligent education systems. Most current KTs either rely on expert judgments or only exploit a single network structure, which affects the full expression of learning features. To adequately mine features of students’ learning process, Deep Knowl...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-07-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/12/14/7188 |
| _version_ | 1797441001850142720 |
|---|---|
| author | Liting Lyu Zhifeng Wang Haihong Yun Zexue Yang Ya Li |
| author_facet | Liting Lyu Zhifeng Wang Haihong Yun Zexue Yang Ya Li |
| author_sort | Liting Lyu |
| collection | DOAJ |
| description | Knowledge tracing (KT) serves as a primary part of intelligent education systems. Most current KTs either rely on expert judgments or only exploit a single network structure, which affects the full expression of learning features. To adequately mine features of students’ learning process, Deep Knowledge Tracing Based on Spatial and Temporal Deep Representation Learning for Learning Performance Prediction (DKT-STDRL) is proposed in this paper. DKT-STDRL extracts spatial features from students’ learning history sequence, and then further extracts temporal features to extract deeper hidden information. Specifically, firstly, the DKT-STDRL model uses CNN to extract the spatial feature information of students’ exercise sequences. Then, the spatial features are connected with the original students’ exercise features as joint learning features. Then, the joint features are input into the BiLSTM part. Finally, the BiLSTM part extracts the temporal features from the joint learning features to obtain the prediction information of whether the students answer correctly at the next time step. Experiments on the public education datasets ASSISTment2009, ASSISTment2015, Synthetic-5, ASSISTchall, and Statics2011 prove that DKT-STDRL can achieve better prediction effects than DKT and CKT. |
| first_indexed | 2024-03-09T12:16:36Z |
| format | Article |
| id | doaj.art-388a7062554b4874a1f03a7e3ac93ba3 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-03-09T12:16:36Z |
| publishDate | 2022-07-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-388a7062554b4874a1f03a7e3ac93ba32023-11-30T22:45:15ZengMDPI AGApplied Sciences2076-34172022-07-011214718810.3390/app12147188Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance PredictionLiting Lyu0Zhifeng Wang1Haihong Yun2Zexue Yang3Ya Li4School of Computer Science and Technology, Heilongjiang Institute of Technology, Harbin 150050, ChinaSchool of Educational Information Technology, Central China Normal University, Wuhan 430079, ChinaSchool of Computer Science and Technology, Heilongjiang Institute of Technology, Harbin 150050, ChinaSchool of Computer Science and Technology, Heilongjiang Institute of Technology, Harbin 150050, ChinaSchool of Computer Science and Technology, Heilongjiang Institute of Technology, Harbin 150050, ChinaKnowledge tracing (KT) serves as a primary part of intelligent education systems. Most current KTs either rely on expert judgments or only exploit a single network structure, which affects the full expression of learning features. To adequately mine features of students’ learning process, Deep Knowledge Tracing Based on Spatial and Temporal Deep Representation Learning for Learning Performance Prediction (DKT-STDRL) is proposed in this paper. DKT-STDRL extracts spatial features from students’ learning history sequence, and then further extracts temporal features to extract deeper hidden information. Specifically, firstly, the DKT-STDRL model uses CNN to extract the spatial feature information of students’ exercise sequences. Then, the spatial features are connected with the original students’ exercise features as joint learning features. Then, the joint features are input into the BiLSTM part. Finally, the BiLSTM part extracts the temporal features from the joint learning features to obtain the prediction information of whether the students answer correctly at the next time step. Experiments on the public education datasets ASSISTment2009, ASSISTment2015, Synthetic-5, ASSISTchall, and Statics2011 prove that DKT-STDRL can achieve better prediction effects than DKT and CKT.https://www.mdpi.com/2076-3417/12/14/7188predictionlearning performancee-learningdeep learningknowledge tracingknowledge representation |
| spellingShingle | Liting Lyu Zhifeng Wang Haihong Yun Zexue Yang Ya Li Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance Prediction Applied Sciences prediction learning performance e-learning deep learning knowledge tracing knowledge representation |
| title | Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance Prediction |
| title_full | Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance Prediction |
| title_fullStr | Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance Prediction |
| title_full_unstemmed | Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance Prediction |
| title_short | Deep Knowledge Tracing Based on Spatial and Temporal Representation Learning for Learning Performance Prediction |
| title_sort | deep knowledge tracing based on spatial and temporal representation learning for learning performance prediction |
| topic | prediction learning performance e-learning deep learning knowledge tracing knowledge representation |
| url | https://www.mdpi.com/2076-3417/12/14/7188 |
| work_keys_str_mv | AT litinglyu deepknowledgetracingbasedonspatialandtemporalrepresentationlearningforlearningperformanceprediction AT zhifengwang deepknowledgetracingbasedonspatialandtemporalrepresentationlearningforlearningperformanceprediction AT haihongyun deepknowledgetracingbasedonspatialandtemporalrepresentationlearningforlearningperformanceprediction AT zexueyang deepknowledgetracingbasedonspatialandtemporalrepresentationlearningforlearningperformanceprediction AT yali deepknowledgetracingbasedonspatialandtemporalrepresentationlearningforlearningperformanceprediction |