- Designing Personalized Learning Environments | The Role of Learning Analytics. - Learners have the capabilities to interact with the learning analytics system through adequate user interface. - This paper proposes a framework for learning analytics aimed to improve personalized learning environments, encouraging the learner's skills to monitor, adapt, and improve their own learning. - It is an attempt to articulate the characterizing prop- erties that reveals the association between learning analytics and personalized learning envi- ronment. - ¯ndings indicate that educational data for learning analytics are context speci¯c and variables carry di®erent meanings and can have di®erent implications on learning success prediction.. - Keywords: Learning analytics. - Recently, educational data analysis, such as learning analytics (LA), academic analytics (AA) and educational data mining (EDM), has been established as an innovative research area. - It is distributed under the terms of the Creative Commons Attribution 4.0 (CC BY) License which permits use, distribution and reproduction in any medium, provided the original work is properly cited.. - Personalized learning encourages the active involvement of learners in the learning process by improving learning experiences and outcomes, it consists of the following processes 6. - The rest of the paper is organized as follows. - Section 5 of the paper provides the concluding remarks and possible future research.. - 9 Learners interact through the adequate user interface of the LA system, which Vietnam J. - Some frameworks concentrate on Social Learning Analytics (SLA) in which learners' discussion activities are visual- ized using data mining and visualization tools. - The architecture of the tool has been developed to provide an adaptable, modular platform that supports a great number of new modular visualizations. - 22 Predicting learners' learning success and providing practical feedback have been two of the most frequently implemented tasks associated with LA. - for knowing the learners' progress in di®erent skills or the learners' last activity in the di®erent resources). - In addition one example of the new analysis of learning pro- cesses, provided by ALAS-KA, presents the analysis and visualizations of e®ective states. - Based on all the indicators available, individual visualizations can be useful for inspecting the learning styles of learners, though group visualizations can help teachers and learners to make a decision in the learning process. - 26 `Tell Me More' is an application for learning the language that accompanies the results of the exercise as the basis for the visualization of the progress of students. - 2 An overview of these educa- tional LA applications with key features and role in a PLE is presented in Table 1, which embodies an expanded version of the Verbert et al. - However, further personalization and adaptive features of the LA system are desirable for improvement of current learning strategies. - Many universities have established that LA can signi¯cantly improve the promotion of the institution itself in terms of allocation of resources, learner achievements, ¯nance and administration. - This paper, based on comprehensive analysis of the above-mentioned literature and our long-term experience 35–38 with implementation, use and evaluation of a personalized, recommendation-based edu- cation system, proposes a personalized LA framework that can have valuable implications on learning success prediction.. - Personalized Learning Analytics Framework. - Learning analytics features. - 2 Therefore, further personalization and adaptive features of the LA system are required to plan upcoming learning activities or to adapt existing learning strategies.. - Learning analytics role in a personalized learning system. - Another example relates to the use of data in the LMS based on which the teacher can determine if, when and how often the learner has accessed the proper LMS tools. - Likewise, and consistent with the goals of the national attention on assessment, LA can help faculty to improve teaching and learning opportunities for learners 40 and potentially identify areas which are improving by monitoring learner performance and course attendance, as well as examining how this relates to grades. - 43 Assess- ment, as an integral part of the learning process, can foster regulation, under- standing, creation, conclusion and clari¯cation the learning material because learners are more responsible for what they understand when they learn it, and when they show what they have understood. - relation to the learning objective. - Cumulative assessments provide information about the learners, their families, and the sta® of the level of assessment authority regarding the given competencies. - Most of these estimates correspond to the end of the unit or the entire course. - Learning analytics process. - The LA engine, as the central component of the PLE is responsible for ¯nding and then processing all these data based on separate analysis sections (Fig. - (1) Data collecting — the process of gathering quantitative and qualitative in- formation about variables. - Learning analytics action cycle.. - (2) Organization of data — the technique of categorizing and organizing datasets to make them more useful.. - (3) Data cleaning and validation — the process of identifying and adjusting (or removing) incomplete, inappropriate, inaccurate or irrelevant parts of the data and then replacing, modifying or deleting the rough data.. - The purpose of our experimental study was to validate the student pro¯le of the proposed LA framework based on data obtained from Educational software course in Moodle LMS. - In order to identify predictive models for learner success, which is the most important part of the Learning Analytics Process (Fig. - Data analysis for this study was completed to answer the following re- search question: Which predictors are important for learner characteristics that lead to the successful completion of the coursework? This section reports the results of the quantitative data analysis. - Section 4 describes the details of the various statistical tests completed as part of this study. - The research question will be addressed within the discussion and conclusions of the study.. - Most of the learners were from the grammar schools followed from economic and technical schools, and the rest of the respondents were classi¯ed into a vocational school. - With regard to post-secondary education, vocational schools are traditionally distinguished from grammar school by their focus on job-speci¯c training to learners who are typically bound for one of the skilled trades, rather than providing academic training for learners pursuing careers in a professional discipline. - Economic school introduce economics as a key strand of History, Government, and Social Studies, and to develop a critical understanding of the assumptions underpinning economics.. - machining operator, drive technician, bookbinders, mechanical technician) were employed to do one part of the production that required a variety of skilled workers.. - The age of the learners in the study population were categorized into four groups. - We have collected data about previously ¯nished high schools of the learners in the study population ( N ¼ 174 ) and compared them with data obtained for the ¯rst-year learners ( N ¼ 444. - The majority of the ¯rst-year learners. - Table 2 displays the distribution of the set of known characteristics of the ¯rst- year learners by gender, previous education, and age. - Additionally, the distribution of age in the study population shifts from the ¯rst-year learner, because the course we have chosen is the elective course from the ¯rst to the fourth year.. - First-year population Population in the study N (447) Population % N (174) Population % Gender. - The learners enrolled in the course Educational software ranged from sophomore to senior learners. - The distribution of the academic level of learners enrolled in the classes can be seen in Fig. - The results of the analysis, we used to create several prediction models, will be presented in the rest of this section. - These results were used to create ¯gures and tables for the predict phases of the LA process. - The overall average ¯nal grade for the course was 8.017 (out of 10) with 65.517% of the learners receiving a grade 8 or better. - Nearly 17% of learners enrolled in the course ¯nished it earning a grade 10, while 8.89% earned grade 5 (not passed exam) or withdrew from the course. - A series of four di®erent logistic regression models are created in the process of identifying the best model for predicting success. - For this model, the Nagelkerke R 2 estimate re°ects the variability of success that can be attributed to the variables included in the logistic regression model. - The combination of demographic variables used in the model accounts for two in°uences on the likelihood of success ( R 2 ¼ 0:0482. - Because the model explains such a low percentage of the likelihood of success, it was only an accurate predictor 55.16% of the time, based on the area under the curve (ROC Curve Model).. - The Nagelkerke R 2 estimate showed that academic variables accounted for 42.16% of the likelihood of success across the study population, and the area under the curve indicated the model was accurate in predicting success 79.44% of the time.. - The Nagelkerke's R 2 estimation showed that variables related to the previous education in°uenced 5.13% of the likelihood of success. - When variables related to the previous education were used as a prediction model for the study population the area under the curve showed the model was accurate in predicting success 63.18% of the time.. - While each of the described models addresses some fea- tures of the predictors of success for learners, the full model includes complete view based on all analyzed variables. - For the study population the test of the full model was statistically signi¯- cant, X 2 (46, N p <. - This model properly predicts success for 88.15% of the learners in the study population, according to the ROC Curve Model, with a spec- i¯city of 44.8% and sensitivity of 95.1%.. - Data analysis for this study was performed to identify relationships between the variables of the learning pro¯les and the model of success prediction. - The results of the analysis will be discussed in this section. - As well, a comparison of the literature considerations and ¯ndings from the data collected at our institution for this study will be given. - Studies reviewed in the literature had mixed results based on the use of age as a predictor, so these results coincide with the previous studies. - While, there are studies stated that the high school average grade point is one of the greatest predictors of graduation at college. - In the literature, we can ¯nd classi¯cation of learners based on previous education which would allow the administrative and academic sta® to identify learners who would be \at risk". - The authors acknowledge ¯nancial support of the Ministry of Education, Science and Technological Development of the Republic of Serbia (Grant No. - Siemens, Learning analytics: Envisioning a research discipline and a domain of prac- tice, in Proc. - Learning Analytics and Knowledge (ACM, 2012), pp. - Ifenthaler, Are higher education institutions prepared for learning analytics? Tech- Trends . - Conde, Using learning analytics to improve teamwork assessment, Comput. - Kanai, Learning analytics methods, ben- e¯ts, and challenges in higher education: A systematic literature review, Online Learn. - Ventura, Educational data mining: A review of the state of the art, IEEE Trans. - Basiron, Demystifying learning analytics in personalised learning, Int. - Mah, Learning analytics and digital badges: Potential impact on student retention in higher education, Technol. - Lee, Personalised and self regulated learning in the web 2.0 era:. - Greller, The pulse of learning analytics understandings and expectations from the stakeholders, in Proc. - Ferguson, Learning analytics: Drivers, developments and challenges, Int. - Mariño, Visualization improvement in learning analytics using semantic enrichment in State-of-the-Art and Future Directions of Smart Learning (Springer, 2016), pp. - Inventado, Educational data mining and learning analytics, in Learning Analytics (Springer, 2014), pp. - Kloos, Glass: A learning analytics visualization tool, in Proc. - Jovanovi ć , A qualitative evaluation of evolution of a learning analytics tool, Comput. - Joksimovic, Current state and future trends: A citation network analysis of the learning analytics ¯eld, in Proc. - Learning Analytics and Knowledge (ACM, 2014), pp. - Kloos, Alas-ka: A learning analytics extension for better understanding the learning process in the khan academy platform, Comput. - Nanopoulos, Recommender systems in e-learning environments: A survey of the state-of-the-art and possible extensions, Artif.. - Ivanovic, Learning analytics – new °avor and bene¯ts for educational environments, Inform. - Corrin, Learning analytics: A case study of the process of design of visualizations, J. - Yasmin, Application of the classi¯cation tree model in predicting learner dropout behaviour in open and distance learning, Distance Educ. - Berry, Using learning analytics to predict academic success in online and face-to-face learning environments, Doctoral dissertation (2017).
Xem thử không khả dụng, vui lòng xem tại trang nguồn hoặc xem
Tóm tắt