Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050046
Chen Li, Qing Yang, Ming Li, Dou Wen, Yaqun Wang
At present, most deep learning-based analysis of student’s attentional states in class has been studied only for a single model structure, and there is not enough recognition accuracy. To address this issue, an attention classification model FF-BiALSTM is proposed, which integrates an Attention Mechanism and a bi-directional long short-term memory neural network (Bi-LSTM). The Attention Mechanism is used to capture global features better and two Bi-LSTM layers are employed to capture time-domain features more effectively. This study defined two attention states to identify whether students are focused or not. Experiments on the Student EEG and Student Reading datasets show that this algorithm can effectively improve student attention classification performance. This experiment obtained 97.77% accuracy on the Student EEG training set and 91.35% on the Student EEG testing set.
{"title":"Classification of Students’ Attentional States Using Attention Mechanism and BiLSTM Fusion","authors":"Chen Li, Qing Yang, Ming Li, Dou Wen, Yaqun Wang","doi":"10.1109/IEIR56323.2022.10050046","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050046","url":null,"abstract":"At present, most deep learning-based analysis of student’s attentional states in class has been studied only for a single model structure, and there is not enough recognition accuracy. To address this issue, an attention classification model FF-BiALSTM is proposed, which integrates an Attention Mechanism and a bi-directional long short-term memory neural network (Bi-LSTM). The Attention Mechanism is used to capture global features better and two Bi-LSTM layers are employed to capture time-domain features more effectively. This study defined two attention states to identify whether students are focused or not. Experiments on the Student EEG and Student Reading datasets show that this algorithm can effectively improve student attention classification performance. This experiment obtained 97.77% accuracy on the Student EEG training set and 91.35% on the Student EEG testing set.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133678164","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050086
Yantao Wei, Fen Lei, Jie Gao, Xiuhan Li
Automatic recognition of student action is an important means to evaluate students' learning status in the class. It also provides a technique for measuring the effectiveness of teaching. However, the complexity of student action poses a challenge to automatic recognition. In this paper, a student action recognition method based on the fuzzy broad learning system (fuzzy BLS) is proposed. Fuzzy BLS is designed by merging the Takagi-Sugeno (TS) fuzzy system into BLS. As a neuro-fuzzy model, fuzzy BLS overcomes some problems, such as suffering from a time-consuming training stage and a large number of fuzzy rules. To get more abundant local features from student action images, we use the Scale-Invariant Feature Transform (SIFT) descriptor combined with the Local LogEuclidean Multivariate Gaussian $(mathrm{L}^{2}mathrm{E}mathrm{M}mathrm{G})$ descriptor to extract image features. Then, the extracted features are fed into fuzzy BLS after dimension reduction. The experimental results on the self-built dataset have shown that the proposed student action recognition method achieves better performance than other benchmarking methods.
{"title":"Student Action Recognition Based on Fuzzy Broad Learning System","authors":"Yantao Wei, Fen Lei, Jie Gao, Xiuhan Li","doi":"10.1109/IEIR56323.2022.10050086","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050086","url":null,"abstract":"Automatic recognition of student action is an important means to evaluate students' learning status in the class. It also provides a technique for measuring the effectiveness of teaching. However, the complexity of student action poses a challenge to automatic recognition. In this paper, a student action recognition method based on the fuzzy broad learning system (fuzzy BLS) is proposed. Fuzzy BLS is designed by merging the Takagi-Sugeno (TS) fuzzy system into BLS. As a neuro-fuzzy model, fuzzy BLS overcomes some problems, such as suffering from a time-consuming training stage and a large number of fuzzy rules. To get more abundant local features from student action images, we use the Scale-Invariant Feature Transform (SIFT) descriptor combined with the Local LogEuclidean Multivariate Gaussian $(mathrm{L}^{2}mathrm{E}mathrm{M}mathrm{G})$ descriptor to extract image features. Then, the extracted features are fed into fuzzy BLS after dimension reduction. The experimental results on the self-built dataset have shown that the proposed student action recognition method achieves better performance than other benchmarking methods.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128592486","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050066
Min Chen, Chi Zhou
Research suggests the importance of providing learners with contextualized learning content that meets the demands of learning context in the ubiquitous learning (ulearning) environment. Similarly, the positive role of collaborative learning is recognized. However, it is not clear how collaboration may benefit learners if they are provided with contextualized learning content that meets their individual needs in u-learning activities. To bridge the gap, this study explored the cross effect of contextualized learning content and collaborative behaviours on students’ learning effect in a u-learning environment. Thirty-four first-year students at a vocational college in China participated in a sixweek comparison experiment and were interviewed in focus groups. The study found that, regardless of whether contextualized learning content was provided, learners tended to collaborate by cutting the task apart in the ubiquitous environment; contextualized learning content had a positive impact on learners’ learning effect; collaboration by cutting the task apart did not benefit the learning effect; the cross effect of this collaboration and contextualized learning content on the learning effect was not significant. Implications for promoting effective u-learning in terms of learning content and collaboration are proposed.
{"title":"The Effects of Contextualized Learning Content and Collaborative Behaviours in a Ubiquitous Learning Environment","authors":"Min Chen, Chi Zhou","doi":"10.1109/IEIR56323.2022.10050066","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050066","url":null,"abstract":"Research suggests the importance of providing learners with contextualized learning content that meets the demands of learning context in the ubiquitous learning (ulearning) environment. Similarly, the positive role of collaborative learning is recognized. However, it is not clear how collaboration may benefit learners if they are provided with contextualized learning content that meets their individual needs in u-learning activities. To bridge the gap, this study explored the cross effect of contextualized learning content and collaborative behaviours on students’ learning effect in a u-learning environment. Thirty-four first-year students at a vocational college in China participated in a sixweek comparison experiment and were interviewed in focus groups. The study found that, regardless of whether contextualized learning content was provided, learners tended to collaborate by cutting the task apart in the ubiquitous environment; contextualized learning content had a positive impact on learners’ learning effect; collaboration by cutting the task apart did not benefit the learning effect; the cross effect of this collaboration and contextualized learning content on the learning effect was not significant. Implications for promoting effective u-learning in terms of learning content and collaboration are proposed.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124787635","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper constructed a multi-modal analysis framework of teacher-student interaction based on intelligent technology. Voiceprint recognition was used to divide the teaching video into slices according to sentences and then used speech recognition, speech emotion analysis, gaze point estimation, and other technologies to recognize and encoded the multimodal behavior of each slice. We analyzed 10 lessons using the event sampling method proposed in the analysis framework in comparison with the classical temporal sampling analysis method and demonstrated the results of multimodal interaction analysis of an instructional video as an example. The results indicated that the event sampling method proposed not only reduces the number of analysis units but also has more complete information about the utterance of each unit, overcoming the incomplete information or information redundancy of analysis units caused by the mechanical segmentation of temporal sampling. The multimodal analysis showed that taking into account both teacher-student verbal and nonverbal interactions can reveal richer and deeper information about classroom teaching and learning. This framework provides an important reference for intelligent multimodal analysis of teacher-student interaction.
{"title":"Intelligent Multimodal Analysis Framework for Teacher-Student Interaction","authors":"Mengke Wang, Liang Luo, Zengzhao Chen, Qiuyu Zheng, Jiawen Li, Wei Gao","doi":"10.1109/IEIR56323.2022.10050044","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050044","url":null,"abstract":"This paper constructed a multi-modal analysis framework of teacher-student interaction based on intelligent technology. Voiceprint recognition was used to divide the teaching video into slices according to sentences and then used speech recognition, speech emotion analysis, gaze point estimation, and other technologies to recognize and encoded the multimodal behavior of each slice. We analyzed 10 lessons using the event sampling method proposed in the analysis framework in comparison with the classical temporal sampling analysis method and demonstrated the results of multimodal interaction analysis of an instructional video as an example. The results indicated that the event sampling method proposed not only reduces the number of analysis units but also has more complete information about the utterance of each unit, overcoming the incomplete information or information redundancy of analysis units caused by the mechanical segmentation of temporal sampling. The multimodal analysis showed that taking into account both teacher-student verbal and nonverbal interactions can reveal richer and deeper information about classroom teaching and learning. This framework provides an important reference for intelligent multimodal analysis of teacher-student interaction.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126585542","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050061
Qinglin Huang, Zhili Zhang, Siyi Jiang, X. Liao, Heng Luo
The “double reduction” policy is a national educational policy issued by the Chinese government in 2021, aiming to reduce the amount of homework and study time of K-12 students. In this study, we collected various data on students’ demographic characteristics, learning patterns, and learning perceptions under the “double reduction” policy using a self-developed questionnaire, and obtained their standardized semester-end test results as measurement of learning outcomes. A total of 8100 5th graders from 45 primary schools in a school district in Wuhan participated in this study. Based on the K-means clustering results, we classified the students into three profile categories: Challenged Learners, Policy Followers, and Competitive Learners and further compared the three learner profiles to identify differences in learning load, learning motivation, and learning outcomes. The study results inform individualized education to accommodate profile differences and inform the sustainable implementation and refinement of the “double reduction” policy.
{"title":"Comparison of Three Learner Profiles under the Influence of the Double Reduction Policy — Evidence from the K-means Clustering Approach","authors":"Qinglin Huang, Zhili Zhang, Siyi Jiang, X. Liao, Heng Luo","doi":"10.1109/IEIR56323.2022.10050061","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050061","url":null,"abstract":"The “double reduction” policy is a national educational policy issued by the Chinese government in 2021, aiming to reduce the amount of homework and study time of K-12 students. In this study, we collected various data on students’ demographic characteristics, learning patterns, and learning perceptions under the “double reduction” policy using a self-developed questionnaire, and obtained their standardized semester-end test results as measurement of learning outcomes. A total of 8100 5th graders from 45 primary schools in a school district in Wuhan participated in this study. Based on the K-means clustering results, we classified the students into three profile categories: Challenged Learners, Policy Followers, and Competitive Learners and further compared the three learner profiles to identify differences in learning load, learning motivation, and learning outcomes. The study results inform individualized education to accommodate profile differences and inform the sustainable implementation and refinement of the “double reduction” policy.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115906314","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050055
Bin Wang, Xinguo Yu, Huihui Sun
Intelligent solutions are an important research field in artificial intelligence education, automatic reasoning and solutions to function problems are a key technology of intelligent tutoring services, which has become a challenging research hotspot in the field of artificial intelligence. This paper aims to provide explanatory solution guidance for intelligent tutor, and proposes a functional problem solving model based on educational cognition and applies it. Specifically, the algorithm model simulates human solution cognition, decomposes the solution process into several progressive combinations of solution states, and designs independent methods to solve subtasks. In each subtask, the cognitive solution is visualized through the mode of cognitive state guiding the closing action to reflect the transformation process between the solution steps. In order to verify the proposed solution model and apply it to the function problem solution, the effectiveness of the new solution model is proved by case analysis and experimental results.
{"title":"Solving Word Function Problems in Line with Educational Cognition Way","authors":"Bin Wang, Xinguo Yu, Huihui Sun","doi":"10.1109/IEIR56323.2022.10050055","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050055","url":null,"abstract":"Intelligent solutions are an important research field in artificial intelligence education, automatic reasoning and solutions to function problems are a key technology of intelligent tutoring services, which has become a challenging research hotspot in the field of artificial intelligence. This paper aims to provide explanatory solution guidance for intelligent tutor, and proposes a functional problem solving model based on educational cognition and applies it. Specifically, the algorithm model simulates human solution cognition, decomposes the solution process into several progressive combinations of solution states, and designs independent methods to solve subtasks. In each subtask, the cognitive solution is visualized through the mode of cognitive state guiding the closing action to reflect the transformation process between the solution steps. In order to verify the proposed solution model and apply it to the function problem solution, the effectiveness of the new solution model is proved by case analysis and experimental results.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114718058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The traditional coding process requires mechanical observation and categorization of the various utterances produced in the classroom. Both the judgment and the professionalism of education of the coders are very challenging. With the development of Automatic Speech Recognition (ASR) and natural language processing (NLP). It is possible for researchers to automate the recognition of speech acts in the classroom. There are also many related studies, but they have not been able to complete the automatic recognition of the classroom interaction speech act(CISA). In order to solve problems, our research proposes a practical CISA coding system. And according to this system, a related CISA dataset is established. A Multi-text classification(MTC) model called Bert-TextConcat is proposed for training on the constructed dataset. The trained model performs automatic classification of CISA while referring to the above. After experiments, We demonstrate the effectiveness of the BertTextConcat model and CISA coding systems.
{"title":"Automatic Recognition of Speech Acts in Classroom Interaction Based on Multi-Text Classification","authors":"Miao Xia, Wei Deng, Sixv Zhang, Meijuan Liu, JiaLi Xu, Peiyun Zhai","doi":"10.1109/IEIR56323.2022.10050047","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050047","url":null,"abstract":"The traditional coding process requires mechanical observation and categorization of the various utterances produced in the classroom. Both the judgment and the professionalism of education of the coders are very challenging. With the development of Automatic Speech Recognition (ASR) and natural language processing (NLP). It is possible for researchers to automate the recognition of speech acts in the classroom. There are also many related studies, but they have not been able to complete the automatic recognition of the classroom interaction speech act(CISA). In order to solve problems, our research proposes a practical CISA coding system. And according to this system, a related CISA dataset is established. A Multi-text classification(MTC) model called Bert-TextConcat is proposed for training on the constructed dataset. The trained model performs automatic classification of CISA while referring to the above. After experiments, We demonstrate the effectiveness of the BertTextConcat model and CISA coding systems.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117106363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050052
Xinyan Zhang, Yuqi Chen, Junjie Hu, Shengze Hu, Tao Huang
Learning situation analysis systems provide personalized learning diagnostic services for students by mining learning data to improve learning efficiency. However, most of the existing systems only focus on partial data from a single learning situation, unable to meet the analysis of changeable and complicated states of students. To alleviate the problem, we propose a novel system I-portrait, which is based on the analysis of multidimensional learning data to provide students with comprehensive portrait services. I-portrait is composed of four modules, cognitive level, subject ability, classroom behavior and emotional attitude. Specifically, we first divide student learning data into static data and dynamic data by concepts and data sources. Then, in each module, I-portrait uses corresponding intelligence artificial technologies to smartly analyze multidimensional student data. Finally, I-portrait integrates analysis results and offers students personalized intelligent learning recommendations, promoting efficient study.
{"title":"I-portrait: A Multidimensional Student Portrait System for Learning Situation Analysis","authors":"Xinyan Zhang, Yuqi Chen, Junjie Hu, Shengze Hu, Tao Huang","doi":"10.1109/IEIR56323.2022.10050052","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050052","url":null,"abstract":"Learning situation analysis systems provide personalized learning diagnostic services for students by mining learning data to improve learning efficiency. However, most of the existing systems only focus on partial data from a single learning situation, unable to meet the analysis of changeable and complicated states of students. To alleviate the problem, we propose a novel system I-portrait, which is based on the analysis of multidimensional learning data to provide students with comprehensive portrait services. I-portrait is composed of four modules, cognitive level, subject ability, classroom behavior and emotional attitude. Specifically, we first divide student learning data into static data and dynamic data by concepts and data sources. Then, in each module, I-portrait uses corresponding intelligence artificial technologies to smartly analyze multidimensional student data. Finally, I-portrait integrates analysis results and offers students personalized intelligent learning recommendations, promoting efficient study.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"5 6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121737928","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050065
Chensi Li, Xinguo Yu, Rao Peng
Problem solving technology is a hot research issue in intelligent education. Linear function scenario problem is one of the important types of problems. This paper presents a linear function relation identification algorithm for solving linear function problems. Firstly, the problem text was transformed into semantic vectors through the BERT model. Secondly, a linear function relation candidate set is created and a Bi-LSTM based identification model is used to select the correct set of linear relations among candidates. Finally, a two-stage solving method is used to obtain the implicit and explicit relations from the correct set of linear relations to get the result. The experiment was tested on 486 linear function scenario problems. The result shows our algorithm achieved 86.1% accuracy in finding the correct set of linear relations and 59.4% accuracy in solving linear function scenario problems.
{"title":"Linear Function Relation Identification Based on BERT and Bi-LSTM","authors":"Chensi Li, Xinguo Yu, Rao Peng","doi":"10.1109/IEIR56323.2022.10050065","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050065","url":null,"abstract":"Problem solving technology is a hot research issue in intelligent education. Linear function scenario problem is one of the important types of problems. This paper presents a linear function relation identification algorithm for solving linear function problems. Firstly, the problem text was transformed into semantic vectors through the BERT model. Secondly, a linear function relation candidate set is created and a Bi-LSTM based identification model is used to select the correct set of linear relations among candidates. Finally, a two-stage solving method is used to obtain the implicit and explicit relations from the correct set of linear relations to get the result. The experiment was tested on 486 linear function scenario problems. The result shows our algorithm achieved 86.1% accuracy in finding the correct set of linear relations and 59.4% accuracy in solving linear function scenario problems.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"507 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131448730","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-12-18DOI: 10.1109/IEIR56323.2022.10050084
Fucheng Guo, Pengpeng Jian
Geometry problem solving is a long-standing problem in artificial intelligence. The task requires generating explainable solving sequences based on text and diagram descriptions. Existing approaches have made great progress in geometry formal language extraction and interpretable solving. However, they neglect the graph structure information in formal language. This leads to poor prediction effect of the theorem, and too long reasoning time for problem solving and affects the accuracy of problem solving. In this paper, we construct the formal language graph and use a graph convolutional network to encode structure information of formal language. We propose an improved diagram parser for better diagram relation set extraction. The experimental results show that our method achieves better performance in interpretable geometry problem solving.
{"title":"A Graph Convolutional Network Feature Learning Framework for Interpretable Geometry Problem Solving","authors":"Fucheng Guo, Pengpeng Jian","doi":"10.1109/IEIR56323.2022.10050084","DOIUrl":"https://doi.org/10.1109/IEIR56323.2022.10050084","url":null,"abstract":"Geometry problem solving is a long-standing problem in artificial intelligence. The task requires generating explainable solving sequences based on text and diagram descriptions. Existing approaches have made great progress in geometry formal language extraction and interpretable solving. However, they neglect the graph structure information in formal language. This leads to poor prediction effect of the theorem, and too long reasoning time for problem solving and affects the accuracy of problem solving. In this paper, we construct the formal language graph and use a graph convolutional network to encode structure information of formal language. We propose an improved diagram parser for better diagram relation set extraction. The experimental results show that our method achieves better performance in interpretable geometry problem solving.","PeriodicalId":183709,"journal":{"name":"2022 International Conference on Intelligent Education and Intelligent Research (IEIR)","volume":"98 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134140799","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：