Article
Computer Science, Artificial Intelligence
Jinghui Qin, Zhicheng Yang, Jiaqi Chen, Xiaodan Liang, Liang Lin
Summary: Math word problem solving requires understanding the real-world narrative in the problem text and identifying the relationships between quantifiers and variables. Existing deep learning models have made progress but lack grounding equation logic. Our proposed approach, TCDP, incorporates mathematical logic and real-world knowledge into a PLM-based encoder for improved performance.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Acoustics
Yi Zhang, Guangyou Zhou, Zhiwen Xie, Jimmy Xiangji Huang
Summary: In this paper, a novel hierarchical heterogeneous graph encoding model called HGEN is proposed for solving math word problems. HGEN outperforms existing Graph2Tree-based models by introducing node-level and type-aware attention layers as well as hierarchical node propagation to handle the dependencies of heterogeneous nodes.
IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING
(2022)
Article
Education & Educational Research
Chih-Pu Dai, Fengfeng Ke, Yanjun Pan
Summary: This exploratory mixed-methods case study examines the design heuristics and implications of narrative in a digital game-based learning (DGBL) environment for math problem solving. The study finds positive correlations between narrative usage and successful math problem solving in the game, and also identifies gender effects with girls being more supported by narratives designed with intrinsic integration. The findings emphasize the critical role of game narratives in a sandbox DGBL environment for math problem solving, as well as the importance of gender-inclusive narrative design.
ETR&D-EDUCATIONAL TECHNOLOGY RESEARCH AND DEVELOPMENT
(2022)
Article
Psychology, Multidisciplinary
Jijian Lu, Pan Tuo, Ruisi Feng, Max Stephens, Mohan Zhang, Zhonghua Shen
Summary: The aim of this study is to investigate the responsibility shift in collaborative problem solving during computer-supported one-to-one tutoring using a commognitive responsibility framework. The study found that there is an alternating-led commognitive responsibility shift in both the teacher's and student's behavior during online collaboration.
FRONTIERS IN PSYCHOLOGY
(2022)
Article
Mathematics
Juhaina Awawdeh Shahbari
Summary: The study found that students went through three main stages in the activity: intuitive error, revealing of the error connected with cognitive conflict, and overcoming the error. Similar emotion features were identified among students in each of the three stages. During the revealing of errors stage, indicators of uncomfortable situations were observed, such as high sound or sad tone of voice and certain physical movements.
Article
Chemistry, Multidisciplinary
Bai Zhonghang, Lin Siyue, Zhang Xu
Summary: This study aims to solve the service process problem by analyzing the service touchpoints flow delivery process and propose a method for problem solving based on flow trimming. The flow model of problematic service touchpoints is constructed to visualize the flow delivery process. Service process trimming rules are proposed to trim flow disadvantages, and resource analysis is employed to obtain specific programs. The stochastic dominance rule is used to rank and select the optimal program. The feasibility of the proposed method is verified through problem solving in the medical treatment service process and selecting the optimal program for presentation.
APPLIED SCIENCES-BASEL
(2023)
Article
History & Philosophy Of Science
Davide Rizza
Summary: This paper examines the problem-solving activity in mathematical social science within the context of scientific practice. It identifies three phases of mathematical problem-solving as reference points for investigating aspects of applications that have not been widely discussed in philosophical literature.
Article
Computer Science, Interdisciplinary Applications
Sourav Mandal, Sudip Kumar Naskar
Summary: Automatically solving mathematical word problems is a challenging research problem that has gained momentum in natural language processing, machine learning, and education technology domains in recent years. This article introduces an arithmetic mathematical word problems solver (AMWPS) that surpasses similar systems with a state-of-the-art accuracy of 94.22% on standard datasets. AMWPS combines machine learning classification and rule-based approaches to classify problem types, operations, identify relevant quantities, and generate final answers.
IEEE TRANSACTIONS ON LEARNING TECHNOLOGIES
(2021)
Article
Psychology, Experimental
Li Wang, Chen Cao, Xinlin Zhou, Chunxia Qi
Summary: Open math problem solving is crucial for students to deepen understanding and transfer mathematical knowledge. However, the role of spatial abilities in this process has not been given enough attention. This study found that both spatial visualization and spatial working memory were significantly associated with open math problem solving, and their associations varied depending on the difficulty of the problems.
APPLIED COGNITIVE PSYCHOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Muhammed Emre Keskin, Mustafa Yilmaz, Chefi Triki
Summary: This article presents a hierarchical windy postman problem, which emphasizes the priority relation of edges and aims to minimize the length of the tour by passing through each edge at least once. An integer non-heuristic linear mathematical model is proposed, and a solution approach is designed. The results show that the proposed heuristic method is faster and generates better quality solutions compared to the commercial solver Gurobi.
Article
Mathematics, Interdisciplinary Applications
S. Lyaqini, M. Nachaoui, A. Hadri
Summary: This paper presents a framework for solving regularized loss minimization problems using machine learning models, and demonstrates the success and efficiency of nonsmooth loss functions. Specifically, the paper formulates supervised learning problems based on L1 fidelity terms and proposes a primal-dual method to solve the nonsmooth optimization problem. Experimental results show that the proposed method outperforms existing methods in terms of convergence speed, quality, and stability of the predicted models.
CHAOS SOLITONS & FRACTALS
(2022)
Article
Education & Educational Research
Qianqian Gao, Si Zhang, Zhihui Cai, Ke Liu, Ning Hui, Mingwen Tong
Summary: This study examined the characteristics of student teachers' collaborative problem solving competency in two online learning tasks. The results suggested that bigram-indicator may be more representative of this competency than unigram-indicator. The competency was found to have a five-factor structure, and the social and cognitive dimensions unfolded orderly in collaborative tasks.
THINKING SKILLS AND CREATIVITY
(2022)
Article
Psychology, Multidisciplinary
Ching- Lin Wu, Hsueh-Chih Chen
Summary: The association between different brain networks and creativity has become a focus of research. This study identified connections between the white matter network and divergent thinking and insight problem solving, and used graph-theoretical analysis to predict creativity. The results showed that effective information transmission across brain regions was correlated with creative problem solving, with different requirements for divergent thinking and insight problem solving.
AMERICAN JOURNAL OF PSYCHOLOGY
(2023)
Article
Psychology, Multidisciplinary
Tao Gong, Lan Shuai, Yang Jiang, Burcu Arslan
Summary: This study examines the process data of fair tests and exhaustive tests in the National Assessment of Educational Progress program. Significant associations are found between item scores and temporal features of preparation time, execution time, and mean execution time. The study shows that these process features can differentiate high- and low-performing students, providing insights on improving performance in large-scale, online scientific inquiry tasks.
FRONTIERS IN PSYCHOLOGY
(2023)
Article
Green & Sustainable Science & Technology
Changbeom Choi, Seungho Yang, Seon Han Choi, Sooyoung Jang
Summary: Modern society comprises various groups based on their interests, emphasizing the importance of citizen participation in decision-making. Living labs (LL) serve as a sustainable approach for stakeholders to collaborate in problem-solving processes. Introducing modeling and simulation (M&S) process aids participants in identifying issues and analyzing solutions effectively in LL.
Article
Psychology, Experimental
M. Asher Lawson, Hemant Kakkar
Summary: Sharing misinformation can have catastrophic consequences, with low conscientiousness conservatives being more likely to spread fake news. Fact-checker interventions are inadequate in preventing the spread of misinformation.
JOURNAL OF EXPERIMENTAL PSYCHOLOGY-GENERAL
(2022)
Article
Multidisciplinary Sciences
M. Asher Lawson, Ashley E. Martin, Imrul Huda, Sandra C. Matz
Summary: Women are underrepresented in leadership positions due to gender stereotypes. However, appointing women to top management roles can change the language used by organizations and associate women with leadership traits. Increasing female representation might help women overcome gender stereotypes without sacrificing communal characteristics.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2022)
Article
Psychology, Experimental
M. Asher Lawson, Shikhar Anand, Hemant Kakkar
Summary: Fake news has negative impacts on political polarization and social cohesion. This research reveals that group members who do not conform to sharing fake news experience reduced social interaction, highlighting the importance of conformity pressure in the spread of misinformation.
JOURNAL OF EXPERIMENTAL PSYCHOLOGY-GENERAL
(2023)
Article
Psychology, Multidisciplinary
M. Asher Lawson, Richard P. Larrick, Jack B. Soll
JUDGMENT AND DECISION MAKING
(2020)