Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network

Ye Lei; Yupei Zhang; Yi Lin; Xuequn Shang

doi:10.1007/978-981-99-6222-8_9

Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network

Ye Lei, Yupei Zhang, Yi Lin, Xuequn Shang

School of Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Scopus citations

Abstract

Diagnosing the contents of learning in brain activities is a long-standing research task in cognitive sciences. The current studies on cognitive diagnosis (CD) in education determine the status of knowledge concept (KC) based on the observed responses to test items. However, the learning process of KC in the brain is left with no touch. This paper proposes to solve the problem of knowledge-concept diagnosis (KCD) from fMRIs by identifying the concepts a student focuses on in learning activities. Using the graph convolutional network (GCN), we introduce the STEGCN approach composed of a spatial GCN for brain-graph structure, a temporal GCN for brain-activity sequence, and a fully connected network for KCD. To evaluate STEGCN, we acquired an fMRI dataset that was collected on five concepts when students were learning a computer course. The experiment results demonstrate that our proposed method yields better performance than traditional models, showing the effectiveness of STEGCN in concept classification. This study contributes to a new fMRI-based route for knowledge-concept diagnosis.

Original language	English
Title of host publication	Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings
Editors	Long Yuan, Shiyu Yang, Ruixuan Li, Evangelos Kanoulas, Xiang Zhao
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	100-111
Number of pages	12
ISBN (Print)	9789819962211
DOIs	https://doi.org/10.1007/978-981-99-6222-8_9
State	Published - 2023
Event	Proceedings of the 20th Web Information Systems and Applications Conference, WISA 2023 - Chengdu, China Duration: 15 Sep 2023 → 17 Sep 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14094 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	Proceedings of the 20th Web Information Systems and Applications Conference, WISA 2023
Country/Territory	China
City	Chengdu
Period	15/09/23 → 17/09/23

Keywords

Educational Data Mining
fMRI
Graph Convolution Network
Knowledge-Concept Diagnosis
Space-Time Embedding

Access to Document

10.1007/978-981-99-6222-8_9

Cite this

Lei, Y., Zhang, Y., Lin, Y., & Shang, X. (2023). Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network. In L. Yuan, S. Yang, R. Li, E. Kanoulas, & X. Zhao (Eds.), Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings (pp. 100-111). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14094 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-99-6222-8_9

Lei, Ye ; Zhang, Yupei ; Lin, Yi et al. / Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network. Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings. editor / Long Yuan ; Shiyu Yang ; Ruixuan Li ; Evangelos Kanoulas ; Xiang Zhao. Springer Science and Business Media Deutschland GmbH, 2023. pp. 100-111 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{ed42e120bead4be28e8e5eda3128364d,

title = "Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network",

abstract = "Diagnosing the contents of learning in brain activities is a long-standing research task in cognitive sciences. The current studies on cognitive diagnosis (CD) in education determine the status of knowledge concept (KC) based on the observed responses to test items. However, the learning process of KC in the brain is left with no touch. This paper proposes to solve the problem of knowledge-concept diagnosis (KCD) from fMRIs by identifying the concepts a student focuses on in learning activities. Using the graph convolutional network (GCN), we introduce the STEGCN approach composed of a spatial GCN for brain-graph structure, a temporal GCN for brain-activity sequence, and a fully connected network for KCD. To evaluate STEGCN, we acquired an fMRI dataset that was collected on five concepts when students were learning a computer course. The experiment results demonstrate that our proposed method yields better performance than traditional models, showing the effectiveness of STEGCN in concept classification. This study contributes to a new fMRI-based route for knowledge-concept diagnosis.",

keywords = "Educational Data Mining, fMRI, Graph Convolution Network, Knowledge-Concept Diagnosis, Space-Time Embedding",

author = "Ye Lei and Yupei Zhang and Yi Lin and Xuequn Shang",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.; Proceedings of the 20th Web Information Systems and Applications Conference, WISA 2023 ; Conference date: 15-09-2023 Through 17-09-2023",

year = "2023",

doi = "10.1007/978-981-99-6222-8_9",

language = "英语",

isbn = "9789819962211",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "100--111",

editor = "Long Yuan and Shiyu Yang and Ruixuan Li and Evangelos Kanoulas and Xiang Zhao",

booktitle = "Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings",

}

Lei, Y, Zhang, Y, Lin, Y & Shang, X 2023, Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network. in L Yuan, S Yang, R Li, E Kanoulas & X Zhao (eds), Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14094 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 100-111, Proceedings of the 20th Web Information Systems and Applications Conference, WISA 2023, Chengdu, China, 15/09/23. https://doi.org/10.1007/978-981-99-6222-8_9

Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network. / Lei, Ye; Zhang, Yupei; Lin, Yi et al.
Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings. ed. / Long Yuan; Shiyu Yang; Ruixuan Li; Evangelos Kanoulas; Xiang Zhao. Springer Science and Business Media Deutschland GmbH, 2023. p. 100-111 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14094 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network

AU - Lei, Ye

AU - Zhang, Yupei

AU - Lin, Yi

AU - Shang, Xuequn

PY - 2023

Y1 - 2023

N2 - Diagnosing the contents of learning in brain activities is a long-standing research task in cognitive sciences. The current studies on cognitive diagnosis (CD) in education determine the status of knowledge concept (KC) based on the observed responses to test items. However, the learning process of KC in the brain is left with no touch. This paper proposes to solve the problem of knowledge-concept diagnosis (KCD) from fMRIs by identifying the concepts a student focuses on in learning activities. Using the graph convolutional network (GCN), we introduce the STEGCN approach composed of a spatial GCN for brain-graph structure, a temporal GCN for brain-activity sequence, and a fully connected network for KCD. To evaluate STEGCN, we acquired an fMRI dataset that was collected on five concepts when students were learning a computer course. The experiment results demonstrate that our proposed method yields better performance than traditional models, showing the effectiveness of STEGCN in concept classification. This study contributes to a new fMRI-based route for knowledge-concept diagnosis.

AB - Diagnosing the contents of learning in brain activities is a long-standing research task in cognitive sciences. The current studies on cognitive diagnosis (CD) in education determine the status of knowledge concept (KC) based on the observed responses to test items. However, the learning process of KC in the brain is left with no touch. This paper proposes to solve the problem of knowledge-concept diagnosis (KCD) from fMRIs by identifying the concepts a student focuses on in learning activities. Using the graph convolutional network (GCN), we introduce the STEGCN approach composed of a spatial GCN for brain-graph structure, a temporal GCN for brain-activity sequence, and a fully connected network for KCD. To evaluate STEGCN, we acquired an fMRI dataset that was collected on five concepts when students were learning a computer course. The experiment results demonstrate that our proposed method yields better performance than traditional models, showing the effectiveness of STEGCN in concept classification. This study contributes to a new fMRI-based route for knowledge-concept diagnosis.

KW - Educational Data Mining

KW - fMRI

KW - Graph Convolution Network

KW - Knowledge-Concept Diagnosis

KW - Space-Time Embedding

UR - http://www.scopus.com/inward/record.url?scp=85172267143&partnerID=8YFLogxK

U2 - 10.1007/978-981-99-6222-8_9

DO - 10.1007/978-981-99-6222-8_9

M3 - 会议稿件

AN - SCOPUS:85172267143

SN - 9789819962211

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 100

EP - 111

BT - Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings

A2 - Yuan, Long

A2 - Yang, Shiyu

A2 - Li, Ruixuan

A2 - Kanoulas, Evangelos

A2 - Zhao, Xiang

PB - Springer Science and Business Media Deutschland GmbH

T2 - Proceedings of the 20th Web Information Systems and Applications Conference, WISA 2023

Y2 - 15 September 2023 through 17 September 2023

ER -

Lei Y, Zhang Y, Lin Y, Shang X. Knowledge-Concept Diagnosis from fMRIs by Using a Space-Time Embedding Graph Convolutional Network. In Yuan L, Yang S, Li R, Kanoulas E, Zhao X, editors, Web Information Systems and Applications - 20th International Conference, WISA 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 100-111. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-981-99-6222-8_9