Construction and Innovative Exploration of Personalized Learning Systems in the Context of Educational Data Mining

Construction and Innovative Exploration of Personalized Learning Systems in the Context of Educational Data Mining

Xingle Ji, Lu Sun, Xueyong Xu, Xiaobing Lei
DOI: 10.4018/IJICTE.346992
Article PDF Download
Open access articles are freely available for download

Abstract

This study examines the current research on educational data mining, educational learning support services, personalized learning services, and personalized learning paths in education. The authors aim to integrate personalized learning concepts into traditional support services by drawing on the latest theoretical and practical research. Using multimodal data fusion techniques, the study conduct exploratory analyses on various data types, including learner academic performance, psychological assessments, learning behavior, and physiological information. This leads to the construction of a personalized education learning support service model. The model focuses on objectives such as monitoring learning behavior, identifying preferences, recognizing abilities, optimizing paths, and recommending resources. The goal is to provide learners with sustained support services throughout the personalized learning process, addressing individual needs, fostering enthusiasm, and maintaining long-term motivation.
Article Preview
Top

Current Research Status of Educational Data Mining

Education Data Mining (EDM) refers to the application of data mining techniques and tools to extract hidden patterns and useful information from large datasets in the field of education. It involves building learner models, customizing learning content and activities, evaluating the learning process, and enhancing online learning systems. EDM is a crucial research direction in the realm of personalized teaching assistance, and in recent years, it has gained increasing attention from both researchers and educators.

The current research on educational data mining mainly focuses on application environments (that is, learning systems), data mining goals and tasks, tools and technologies, and more. Currently, research on the application of educational data mining mainly focuses on the following aspects:

  • 1.

    Construction of learner models: Labib et al. (2017) utilize ontology technology to define semantic relationships among six different standard concepts of learning style models. They linked the dimensions of learning style models to learner characteristics, constructing a learner feature ontology to provide teachers with cross-model teaching assistance. Yue and Chen (2017) constructed remote learner models from the perspectives of personal information, learning style, learning interests, and knowledge models. They also indicated how to obtain data on these four student characteristics.

  • 2.

    Student performance prediction: Pai et al. (2010) designed an improved rough set theory model to predict the academic performance of Taiwanese students. This model can automatically select key information in the data without predetermined influencing factors, providing accurate inference rules for prediction. Qiu (2010) uses the Levenberg Marquardt algorithm in neural networks to predict the academic performance of college students by exploring the mapping relationship between their learning behavior and academic performance.

  • 3.

    Providing feedback for support services: Sohn and Ju (2010) employed a conjoint analysis method to redistribute the weights of four elements (academic ability test, high school academic performance, essay, and interview) in the college entrance exam. This provided useful information for recruiting high-quality science-oriented high school students effectively. Xiao (2012) used the Apriori algorithm in association rules to mine data from distance education teaching evaluations. The findings provided decision-making information for institutions to improve evaluation mechanisms and teaching methods.

  • 4.

    Data analysis and visualization: Anaya and Boticario (2009) have developed an extensible multivariate analysis method. This approach uses statistical indicators of learner interactions in forums as source data. The data is categorized based on learner collaboration, and the collaborative information of learners is ultimately presented to both instructors and students. This presentation aids them in managing the collaborative process. By employing multivariate statistical and social network analysis methods, Chen and Yan (2012) have constructed a visual knowledge map of the academic community in educational technology.

  • 5.

    Student recommendation: Su et al. (2011) proposed a personalized learning content adaptation mechanism. It utilizes clustering and decision tree technologies to manage and analyze a large number of historical learner requests effectively. Using the acquired information, the system intelligently responds to new user requests by delivering personalized learning content. After employing ontology technology, Jiang et al. (2010) recommend personalized learning resources based on students' learning styles.

Furthermore, data mining can be utilized for detecting undesirable learning behaviors in students, analyzing social networks, developing concept maps, and more (Sáiz-Manzanares et al., 2021). Overall, research in educational data mining tends to focus on the conceptual characteristics of big data, new tools and technologies, the impact on the learning process, and studies on coping strategies. It rarely delves into specific practical issues. Additionally, the scope of research is relatively narrow, and there is redundancy in the use of technologies. For instance, in the realm of teaching assessment and feedback, the association rules algorithm are used predominantly, with only a minority employing decision tree classification algorithms. Other algorithms and analytical methods are seldom involved (Baek & Doleck, 2023).

At the same time, we should pay attention when conducting educational data mining for different age groups, as there are varying points to consider (Sukhija et al., 2015):

  • 1.

    Types and sources of data: Different age groups generate different types of data, such as learning behavior data, cognitive ability data, and psychological characteristic data. These data may come from various platforms and systems, such as school education systems, online learning platforms, and mobile applications.

  • 2.

    Feature extraction and analysis: Different age groups have different learning characteristics and needs. Therefore, precise extraction and analysis of features for different age groups are necessary during data mining. For example, for children, factors such as their learning interests and gamification needs may need to be considered, while for adults, their career development needs and learning goals may be of greater concern.

  • 3.

    Model building and prediction: Different prediction models need to be constructed for different age groups to achieve personalized learning support. These models may be based on different algorithms and theories to adapt better to the learning characteristics and needs of different age groups.

  • 4.

    Personalized services and recommendations: The ultimate goal of educational data mining is to provide personalized learning support and recommendations for learners of different age groups. This may include providing customized learning paths, resource recommendations, and learning feedback tailored to different age groups.

Therefore, when conducting educational data mining for different age groups, it is necessary to consider the differences in their characteristics and needs and use appropriate methods and technologies to achieve personalized learning support. The multimodal fusion-based data mining method proposed in this paper requires the collection of academic performance data, psychological assessment data, behavioral performance data, and physiological information data from learners. These data can be collected through online learning platforms, electronic records, scale assessments, surveys, wearable smart devices, and more. Previous studies have indicated that this model algorithm is more suitable for college students.

Complete Article List

Search this Journal:
Reset
Volume 20: 1 Issue (2024)
Volume 19: 1 Issue (2023)
Volume 18: 3 Issues (2022)
Volume 17: 4 Issues (2021)
Volume 16: 4 Issues (2020)
Volume 15: 4 Issues (2019)
Volume 14: 4 Issues (2018)
Volume 13: 4 Issues (2017)
Volume 12: 4 Issues (2016)
Volume 11: 4 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing