Towards a KD4LA Framework to Support Learning Analytics in

Higher Education

Thi My Hang Vu

1,2

Faculty of Information Technology, University of Science, Ho Chi Minh City, Vietnam

Vietnam National University, Ho Chi Minh City, Vietnam

Keywords: Knowledge Discovery, Education Data Mining, Learning Analytics.

Abstract: Learning analytics (LAs) involves the process of collecting, organizing, and generating insights from

educational data, such as learner assessments, learner profiles, or learner interactions with the educational

environment, to support educators and learners in decision-making. This topic has gained attention from the

community for many decades. Nowadays, with advancements in data mining and the availability of large

amounts of data from various educational environments, learning analytics presents both opportunities and

challenges. Especially in higher education, where data is more complex and data analytics is closely integrated

with pedagogical activities and objectives, a consolidated framework is crucial to support both educators and

learners in their tasks. This paper proposes a comprehensive framework, named KD4LA (Knowledge

Discovery for Learning Analytics), which clarifies essential components of common learning analytics tasks

in higher education. These tasks include generating statistical insights on student assessments, segmenting

students based on their acquired knowledge, or evaluating their proficiency in relation to learning objectives.

The proposed framework is validated through several real-world case studies to demonstrate its practical

applicability.

1 INTRODUCTION

Learning analytics (LA) involves the process of

collecting, organizing, and generating insights from

educational data, such as learner assessments, learner

profiles, or learner interactions with the educational

environment, to support educators and learners in

decision-making (Ahmad et al., 2022; Nunn et al.,

2016). This topic has gained attention from the

community for many decades.

Nowadays, with advancements in data mining

technologies and the increasing availability of large

amounts of data from various educational

environments, learning analytics presents both

opportunities and challenges (Nunn et al., 2016). On

the one hand, these advancements enable more

precise tracking of learner progress, personalized

learning experiences , and data-driven decision-

making to enhance educational outcomes

(Hernández-de-Menéndez et al., 2022; Khalil et al.,

2023). Educators can leverage learning analytics to

identify high-risk learners early, adjust learning

activities to accommodate different learner groups

and learning styles, and optimize curriculum design

based on data-driven insights (Bakharia et al., 2016).

For learners, learning analytics allows them to

monitor their own progress in relation to the required

course outcomes, helping them recognize whether

they are at risk or have the potential to achieve a top

ranking in their class (Aldowah et al., 2019; Alyahyan

& Düştegör, 2020). Additionally, students can

compare their performance with the class average,

providing motivation and self-awareness to improve

their learning strategies (Susnjak et al., 2022). On the

other hand, effectively interpreting complex data to

provide actionable insights without overwhelming

educators with excessive or irrelevant information is

essential. The low adoption rate of learning analytics

among educators indicates that current tools do not

fully align with their needs, highlighting the necessity

for more intuitive, user-friendly, and educator-centric

analytics solutions (Bere et al., 2022). Additionally,

to the best of our knowledge, few researchers have

focused on analyzing how to interpret learning

analytics results in relation to learning outcomes to

assess whether the learning design is effectively

supporting the achievement of specific knowledge.

Vu and T. M. H.

Towards a KD4LA Framework to Support Learning Analytics in Higher Education.

DOI: 10.5220/0013571000003967

In Proceedings of the 14th International Conference on Data Science, Technology and Applications (DATA 2025), pages 575-582

ISBN: 978-989-758-758-0; ISSN: 2184-285X

575

This paper introduces KD4LA (Knowledge

Discovery for Learning Analytics), a comprehensive

framework designed to clarify the learning analytics

process and encompass key components to common

learning analytics tasks in higher education. These

tasks include: generating statistical insights on

student assessments, segmenting students based on

their acquired knowledge, and evaluating student

proficiency in relation to learning objectives. To

validate the proposed framework, we conduct real-

world case studies using data from selected courses at

a designated university. These case studies enable in-

depth analysis of student performance, engagement,

and learning progress. By leveraging real academic

data, the case studies demonstrate the practical

applicability of KD4LA in supporting educators with

data-driven decision-making and enhancing student

learning experiences.

The structure of the paper is as follows. Section 2

presents the methodology adopted to implement the

KD4LA framework. Section 3 summarizes related

works involving learning analytics in higher

education and identifies the research gaps. Section 4

provides details of the KD4LA framework, clarifying

the primary knowledge elements of learning analytics

in higher education and presenting a set of analytics

patterns that serve as blueprints for educators to

perform analytics tasks. Section 5 focuses on

validating the proposed framework through real-

world scenarios in higher education. Finally, Section

6 concludes the paper by discussing the implications

and limitations of the study, as well as suggesting

potential avenues for future research.

2 METHODOLOGY

This research proposes the Knowledge Discovery for

Learning Analytics (KD4LA) framework, a

knowledge-based approach to enhancing learning

design and analytics in higher education. The

framework is developed using the Design Science

Research (DSR) methodology (Dresch et al., 2015),

which focuses on creating innovative artifacts to

solve practical problems through four key phases

(Peffers et al., 2007).

Problem Identification. This phase identifies the

research questions to be addressed for building

KD4LA framework. Two key questions are

identified: RQ#1: What types of essential knowledge

can support analytics? And RQ#2: How can the

knowledge be elaborated and used effectively?.

Solution Definition. This phase defines possible

solutions to solve the identified problems.

Specifically, it defines essential knowledge types in

learning analytics and determines predefined cases to

facilitate analytics tasks. A brief literature review has

been conducted to summarize the current state-of-the-

art in the related field.

Design and Development. This phase involves

creating KD4LA artifacts. These artifacts are

classified in constructs, models, methods, and

instantiation, according to DSR methodology (Peffers

et al., 2007). The constructs define essential

knowledge types, the model formalizes relationships,

and the methods introduces predefined analytics

patterns to guide educators. Several real-world case

studies are also implemented as instantiations to

validate the practical applicability of the framework.

 Constructs clarify fundamental knowledge

types for structuring learning analytical

knowledge elements and ensuring alignment

among these elements. This includes various

types of knowledge: WHO, defining whether

the analysis is for an individual, a group, or

multi-groups; WHAT, specifying the type of

data used for analytics WHY, defining the

analysis purposes; HOW, identifying suitable

analytics methods based on the WHO, WHAT,

and WHY knowledge types; CONTEXT,

adding constraints or conditions for selecting

proper analytics techniques/methods.

 Model organizes the knowledge types defined

in the constructs, a data model is proposed.

This model helps in structuring WHO, WHAT,

WHY, HOW, and CONTEXT knowledge as

interrelated entities, as well as establishing

rules and dependencies to determine how

different knowledge types interact. For

instance, the data model ensures that when an

instructor needs a specific analytics purpose

(WHY) for a given dataset (WHAT), the

system automatically suggests relevant

analytics methods (HOW) while considering

additional contextual constraints (CONTEXT).

 Methods predefine a collection of analytics

patterns to guide educators in performing their

tasks. These patterns encapsulate common

analytical scenarios in higher education and

serve as recommendation templates. For

example, if an instructor wants to compare

(WHY) student final grades (WHAT) across

multiple classes (WHO), the system

recommends descriptive analytics using bar

charts, boxplots, or histograms, or statistical

tests like t-tests (for two groups) or ANOVA

(for more than two groups) to determine if there

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

576

are significant differences between the mean

grades of classes.

Demonstration and Evaluation. This phase

involves validating the proposed framework in real-

world situations. To assess the practical applicability

of our framework, we adopt the data model and apply

the predefined analytics patterns in some courses at a

selected university. This evaluation aims to determine

the potential of the framework in deriving meaningful

insights from learning analytics.

3 RELATED WORKS

This section provides an overview of current research

on data analytics in the education sector and identifies

research gaps that need to be addressed to enhance the

adoption of learning analytics tools among educators.

3.1 Data Analytics in Education

The data analytics, in general, can be classified in

three principal categories: descriptive, predictive, and

prescriptive analytics (Bere et al., 2022).

The first type, descriptive analytics, has gained

considerable attention from the educational

community. This type of analytics is often closely

associated with Learning Analytics Dashboards

(LADs) (Costas-Jauregui et al., 2021). Over the

years, both researchers and practitioners have

endeavored to develop interactive and intelligent

dashboards to enhance understanding and discovery

of student performance. Some studies focus on

tracking learner performance, monitor learner

assessement (Peraić & Grubišić, 2022), explore

learner interactions within a learning environment

(Kaliisa & Dolonen, 2023; Peraić & Grubišić, 2022);

while others incorporate data mining or machine

learning techniques for enhanced prediction and

deeper analysis (Peraić et al., 2025; Ramaswami et

al., 2023). The other recent research on LAD is

comprehensively reviewed in (Barbé et al., 2024;

Jayashanka et al., 2022; Masiello et al., 2024).

The second type, predictive analytics, is often

categorized under educational data mining (EDM)

(Aldowah et al., 2019). This analytics type involves

using machine learning or advanced statistical

techniques to discover hidden patterns, relationships,

or trends within educational data, subsequently

enabling accurate forecasts to support decision-

making. Specifically, predictive analytics can

forecast learner performance or retention (Alyahyan

& Düştegör, 2020, 2020; Batool et al., 2023; Bin

Roslan & Chen, 2022), classify learners into different

groups based on learning styles, behaviors, or

academic results (Dol & Jawandhiya, 2023; Križanić,

2020; Nimy & Mosia, 2023). Further relevant studies

on educational data mining can be found in (Baek &

Doleck, 2023; Romero & Ventura, 2020; Salloum et

al., 2020).

The third type, prescriptive analytics, focuses on

recommending specific actions or strategies to

optimize learning/teaching tasks (Susnjak, 2024). It

leverages advanced techniques, including AI

solutions and optimization algorithms, to suggest the

most effective interventions based on predicted

scenarios. In some instances, prescriptive analytics is

integrated with educational recommendation

systems to provide personalized suggestions tailored

to individual learners' needs and preferences (

Dhananjaya et al., 2024; George & Lal, 2021; Saito

& Watanobe, 2020). Although prescriptive analytics

is less common due to its complexity, it holds

significant potential to enhance educational decision-

making (Rivera et al., 2018).

3.2 Research Gap Identification

A recent empirical study (Bere et al., 2022) highlights

critical determinants influencing the adoption of

learning analytics, revealing that the most significant

barrier is the mismatch between educators'

capabilities and the complexity of available analytics

tools. This mismatch underscores the necessity of

aligning technological solutions with educators'

specific needs and skill levels.

From the brief summary from the related works,

most current research tends to focus heavily on

algorithms, educational models, or the application of

machine learning and traditional data mining methods

to extract meaningful insights supporting teaching

and learning practices, commonly referred to as

Educational Data Mining (EDM). Other studies

concentrate on optimizing visual representations

specifically for educational decision-making. Despite

these advancements, there remains a notable absence

of structured methodologies explicitly connecting

the essential components of data analytics;

including data types, analytical objectives, targeted

user requirements, and suitable visualization

techniques; into a cohesive framework.

To address these gaps, this paper introduces the

preliminary KD4LA framework, which clarifies the

essential components (or knowledge types) for

learning analytics by considering educators' needs

and skills within a set of predefined analytics patterns.

Towards a KD4LA Framework to Support Learning Analytics in Higher Education

577

4 KD4LA FRAMEWORK

This section outlines essential constructs of the

KD4LA framework, as well as predefined analytics

patterns designed to facilitate easier adoption and

application of analytics solutions by educators.

4.1 KD4LA Constructs and Model

The KD4LA constructs are structured in a multi-level

data model to enhance reusability and facilitate future

expansions. The model utilizes the 5W1H model

(who, what, why, when, where, and how), as

introduced by (Jang & Woo, 2012). According to the

5W1H model, the KD4LA encompasses five types of

knowledge for specifying learning analytics tasks:

target users involved in analysis (WHO), types of

learning data for analysis (WHAT), analysis purposes

(WHY), analysis methods/techniques used to process

and interpret learning data (HOW), and additional

conditions for selecting suitable analysis methods

(CONTEXT). Figure 1 illustrates a comprehensive

overview of these knowledge types.

Figure 1: KD4LA Knowledge Elements.

The WHO knowledge type in learning analytics

refers to the target learners involved in the analysis.

This factor determines the scope of analysis, which

can be classified into the following scopes:

 Personal analytics focuses the analysis of

individual learners by monitoring their

performance, behaviors, and learning patterns.

 Group analytics concentrates on analyzing

specific groups of learners within a class or

course, providing educators with an aggregated

overview of learning outcomes, participation

levels, and overall student performance.

 Cross-group analytics examines learning

outcomes across multiple groups, classified by

various criteria, to identify potential

imbalances in knowledge and competency

acquisition. This approach helps educators

determine if discrepancies in teaching methods

contribute to varying performance among

different classes or groups.

The WHAT knowledge type refers to the kinds of

data that can be collected, processed, and analyzed to

gain educational insights. In the context of university

research, learning data are primarily collected from

Learning Management Systems (LMS) and can be

classified into the following categories.

 Assessment data stores student results for a

specific course. It can be categorized into

progressive assessment data (P) and final

assessment data (F). The former includes

student grades from labs, quizzes, assignments,

and other learning activities. The latter

represents the overall final course grade.

 Behavioral data captures interactions between

students and the learning environment, such as

the number of clicks on learning activities, time

spent on various activities. This reveals how

frequently different learning activities are

accessed and used by students.

 Learning content data involves specific

concepts, skills, or knowledge areas covered in

a course, typically structured as learning

outcomes. This enables educators to assess

whether learning activities align effectively

with intended learning objectives, identify gaps

in instructional design, and refine content to

enhance knowledge acquisition.

The WHY knowledge type refers to purposes

behind analyzing learning data, influencing the

choice of appropriate analytical methods. Informed

by analytics types (descriptive, predictive, and

prescriptive) and aligned with educators' needs, the

classification of analytical purposes is presented.

 Comparision evaluates differences and

similarities across various learning data to

derive meaningful insights about student

performance, engagement, and learning

behavior.

 Composition analysis examines the distribution

of participation across various learning

activities; such as lectures, quizzes; to identify

which contribute most to student success.

 Distribution analysis visualizes how specific

learning data types are spread across a student

population. This method supports the

identification of student engagement or

knowledge acquisition levels, and highlights

patterns such as outliers or learning gaps.

 Prediction forecasts future outcomes based on

historical data. It uses statistical models and

machine learning techniques to develop

regression models (e.g., linear or logistic

regression) to predict final grades based on

early engagement and assessment data.

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

578

 Classification allows identifying students at

risk of dropping out or likely to succeed, based

on historical patterns, including login

frequency or assignment completion rates.

 Clustering segments students into distinct

groups based on similarities in learning

behaviors or performance, enabling educators

to implement targeted learning strategies or

interventions tailored specifically to each

group.

The HOW knowledge type refers to the specific

analytical methods used to process, interpret, and

derive actionable insights from learning data. The

choice of method depends on the WHAT (type of

data being analyzed), WHY (purpose of analysis),

and WHO (target users of the analysis). By selecting

the appropriate analytical techniques, educators and

decision-makers can effectively translate raw data

into meaningful insights that drive improvements in

teaching and learning. These methods can generally

be categorized into the following key areas.

 Visualization techniques transform data into

visual representations such as graphs, charts,

heatmaps, and dashboards to help educators

and learners quickly determine trends, patterns,

and relationships within the data.

 Statistical methods range from basic

descriptive statistics, like means, medians to

more advanced inferential techniques such as

hypothesis testing, regression models, and

correlation analysis. These methods enable

quantitative assessment of learning outcomes

and help identify significant factors that

influence student performance.

 Machine learning models, Clustering methods,

such as K-Nearest Neighbors (KNN) and

hierarchical clustering, are used to group

students with similar performance patterns or

learning behaviors. Moreover, predictive

models can forecast student outcomes (e.g.,

risk of dropout) based on various learning

indicators, continuously refining their accuracy

as more data becomes available.

The CONTEXT knowledge type acts as a set of

conditions or constraints that further refine HOW is

determined based on WHO, WHAT, and WHY. It

ensures that the selected analytical method is suitable

for the given dataset and scenario.

4.2 KD4LA Methods

In our framework, we define a set of analytics

patterns that encapsulate the four key dimensions

(WHO, WHAT, WHY, HOW) within the educational

context. These patterns serve as predefined templates

that guide the selection of suitable analytical

methods, addressing a critical challenge faced by

many educators who struggle to clearly define their

own analytical needs or choose the appropriate

analytics approach.

Each analytics pattern takes WHO (target users),

WHAT (learning data), and WHY (analytical

purpose) as input parameters and generates possible

HOW (analytics method) to provide educators with

the most effective analytics method to address their

needs (see Figure 1, 2).

For example, in a common scenarios where a

teacher wants to compare (WHY) the final grades

(WHAT) among different classes (WHO) for a

specific course they teach in a semester. Their goal is

to evaluate the effectiveness of their teaching

methods and identify potential imbalances in student

performance across classes. Some suitable analytics

techniques for this comparison include bar charts and

histograms for visualizing grade distributions.

Additionally, mean hypothesis testing (e.g., t-tests for

two groups or ANOVA for multiple groups) can be

applied to determine whether there is a statistically

significant difference in the mean final grades among

classes. The CONTEXT component ensures that the

selected analytical method is appropriate for the given

dataset and scenario. For example, in mean

hypothesis testing, if comparing two groups with a

sample size of less than 30, a t-test (Student’s t-test)

is the appropriate choice.

Figure 2: Examples of KD4LA Analytics Patterns.

5 VALIDATION

In the validation phase of the KD4LA framework, a

comprehensive suite of analytics is applied to two

specific courses at a selected university, enabling the

confirmation and refinement of insights derived from

educational data. This phase leverages visualization

techniques, such as bar charts, histograms, and box

plots, to transform complex data into intuitive, easily

interpretable formats, allowing educators to quickly

identify trends, patterns, and anomalies.

Towards a KD4LA Framework to Support Learning Analytics in Higher Education

579

To ensure consistency and accuracy, the collected

data undergoes serious preprocessing tasks. First,

data cleaning removes missing or inconsistent entries.

Next, the dataset is organized into a standardized

CSV format to streamline analysis. Then, data

transformation converts categorical variables (e.g.,

engagement levels) into numerical representations

suitable for statistical evaluation. Figure 3 illustrates

sample representations of the processed dataset in

CSV format.

Figure 3: Sample Data for Validating KD4LA Framework.

Three case studies will be conducted using the

sample data to uncover valuable educational insights.

Case Study #1: Assessing Grade Distribution

Across a Course. This case study investigates the

distribution (WHY) of assessment data (WHAT);

specifically, final exam grades and final course

grades; for an entire class enrolled in a course

(WHO). Using histograms (Figure 4) and box plots

(HOW) (Figure 5), the study visualizes how these

grades are distributed, allowing for a comparative

evaluation of exam performance against overall

course outcomes (see Figure 4).

Figure 4: Grade Distribution using Histograms.

Figure 5: Grade Distribution using Boxplots.

Case Study #2: Identifying At-Risk Students.

In this case study, a logistic regression model was

used to identify students who are at high risk of failing

a course based on their quiz average (Quiz_Avg),

midterm exam grade (AC015), and final course grade

(AC020). Figure 6 shows a subset of these students

sorted by their predicted risk probability, illustrating

how certain combinations of low final grades and

inconsistent midterm performance can indicate a

higher likelihood of failure.

Figure 6: Identifying At-Risk Students.

Case Study #3: Clustering Students. This case

study groups students into distinct clusters based on

shared performance patterns across quizzes, midterm,

and final exam grades (see Figures 7 and 8). By

examining each cluster’s average grades and bubble

sizes, educators can design targeted strategies for

improvement (for Cluster 0), maintain steady support

(for Cluster 1), and provide enrichment (for Cluster 2).

Figure 7: Student Clustering.

Figure 8: Cluster Visualization.

6 CONCLUSION

This section summarizes the paper’s contributions

and outlines future research directions.

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

580

In terms of contributions, the paper first proposes

a comprehensive knowledge model that integrates

various types of knowledge within the context of

learning analytics (LA). This model offers educators

a holistic view of how different knowledge types can

be leveraged to facilitate data-driven insights. The

second contribution is a structured method that

defines a set of predefined analytics patterns.

Through the case studies, the paper demonstrates the

feasibility of implementing the proposed framework

in real-world educational settings.

In terms of future research, the framework

remains in an early conceptual stage, presenting

opportunities for further development and

refinement. In future work, a key objective is to create

a web-based tool that streamlines interaction between

educators and learners, enabling them to access and

utilize analytics more intuitively. Additionally,

expanding the repository of analytics patterns is

required to enrich the predefined analytics cases,

providing deeper insights into student performance,

engagement, and other critical learning factors. These

enhancements will not only broaden the framework’s

applicability but also foster more robust, data-driven

decision-making in diverse educational contexts.

ACKNOWLEDGEMENTS

This research is funded by University of Science,

VNU-HCM under grant number CNTT 2023-09.

REFERENCES

Ahmad, A., Schneider, J., Griffiths, D., Biedermann, D.,

Schiffner, D., Greller, W., & Drachsler, H. (2022).

Connecting the dots – A literature review on learning

analytics indicators from a learning design perspective.

Journal of Computer Assisted Learning, jcal.12716.

Aldowah, H., Al-Samarraie, H., & Fauzy, W. M. (2019).

Educational data mining and learning analytics for 21st

century higher education: A review and synthesis.

Telematics and Informatics, 37, 13–49.

Alyahyan, E., & Düştegör, D. (2020). Predicting academic

success in higher education: Literature review and best

practices. International Journal of Educational

Technology in Higher Education, 17(1), 3.

Baek, C., & Doleck, T. (2023). Educational Data Mining

versus Learning Analytics: A Review of Publications

From 2015 to 2019. Interactive Learning

Environments, 31(6).

Bakharia, A., Corrin, L., De Barba, P., Kennedy, G.,

Gašević, D., Mulder, R., Williams, D., Dawson, S., &

Lockyer, L. (2016). A conceptual framework linking

learning design with learning analytics. Proceedings of

the Sixth International Conference on Learning

Analytics & Knowledge - LAK ’16.

Barbé, R., Encelle, B., & Sehaba, K. (2024). Adaptation in

Learning Analytics Dashboards: A Systematic Review:

Proceedings of the 16th International Conference on

Computer Supported Education, 75–86.

Batool, S., Rashid, J., Nisar, M. W., Kim, J., Kwon, H.-Y.,

& Hussain, A. (2023). Educational data mining to

predict students’ academic performance: A survey

study. Education and Information Technologies, 28(1).

Bere, A., Chirilele, P., & Chitiga, R. (2022). Critical

Determinants for Learning Analytics Adoption in

Higher Education. In International Association for

Development of the Information Society. International

Association for the Development of the Information

Society.

Bin Roslan, M. H., & Chen, C. J. (2022). Educational Data

Mining for Student Performance Prediction: A

Systematic Literature Review (2015-2021).

International Journal of Emerging Technologies in

Learning (iJET).

Costas-Jauregui, V., Oyelere, S. S., Caussin-Torrez, B.,

Barros-Gavilanes, G., Agbo, F. J., Toivonen, T., Motz,

R., & Tenesaca, J. B. (2021). Descriptive Analytics

Dashboard for an Inclusive Learning Environment.

2021 IEEE Frontiers in Education Conference.

Dhananjaya, G. M., Goudar, R. H., Kulkarni, A. A., Rathod,

V. N., & Hukkeri, G. S. (2024). A Digital

Recommendation System for Personalized Learning to

Enhance Online Education: A Review. IEEE Access,

12, 34019–34041. IEEE Access.

Dol, S. M., & Jawandhiya, P. M. (2023). Classification

Technique and its Combination with Clustering and

Association Rule Mining in Educational Data

Mining—A survey. Engineering Applications of

Artificial Intelligence, 122.

Dresch, A., Lacerda, D. P., & Antunes Jr, J. A. V. (2015).

Design Science Research: A Method for Science and

Technology Advancement.

George, G., & Lal, A. M. (2021). A Personalized Approach

to Course Recommendation in Higher Education.

International Journal on Semantic Web and

Information Systems.

Hernández-de-Menéndez, M., Morales-Menendez, R.,

Escobar, C. A., & Ramírez Mendoza, R. A. (2022).

Learning analytics: State of the art. International

Journal on Interactive Design and Manufacturing

(IJIDeM), 16(3).

Jang, S., & Woo, W. (2012). 5W1H: Unified User-Centric

Context. 2.

Jayashanka, R., Hettiarachchi, E., & Hewagamage, K. P.

(2022). Technology Enhanced Learning Analytics

Dashboard in Higher Education. Electronic Journal of

E-Learning, 20(2), pp151-170.

Kaliisa, R., & Dolonen, J. A. (2023). CADA: A teacher-

facing learning analytics dashboard to foster teachers’

awareness of students’ participation and discourse

patterns in online discussions. Technology, Knowledge

and Learning, 28(3), 937–958.

Towards a KD4LA Framework to Support Learning Analytics in Higher Education

581

Khalil, M., Prinsloo, P., & Slade, S. (2023). The use and

application of learning theory in learning analytics: A

scoping review. Journal of Computing in Higher

Education, 35(3), 573–594.

Križanić, S. (2020). Educational data mining using cluster

analysis and decision tree technique: A case study.

International Journal of Engineering Business

Management, 12.

Masiello, I., Mohseni, Z. (Artemis), Palma, F., Nordmark,

S., Augustsson, H., & Rundquist, R. (2024). A Current

Overview of the Use of Learning Analytics

Dashboards. Education Sciences, 14(1), Article 1.

Nimy, E., & Mosia, M. (2023). Web-based Clustering

Application for Determining and Understanding

Student Engagement Levels in Virtual Learning

Environments. E-Journal of Humanities, Arts and

Social Sciences, 4–19.

Nunn, S., Avella, J. T., Kanai, T., & Kebritchi, M. (2016).

Learning Analytics Methods, Benefits, and Challenges

in Higher Education: A Systematic Literature Review.

Online Learning, 20(2).

Peffers, K., Tuunanen, T., Rothenberger, M. A., &

Chatterjee, S. (2007). A Design Science Research

Methodology for Information Systems Research.

Journal of Management Information Systems, 24(3),

Article 3.

Peraić, I., & Grubišić, A. (2022). Development and

Evaluation of a Learning Analytics Dashboard for

Moodle Learning Management System. In G.

Meiselwitz, A. Moallem, P. Zaphiris, A. Ioannou, R. A.

Sottilare, J. Schwarz, & X. Fang (Eds.), HCI

International 2022—Late Breaking Papers. Interaction

in New Media, Learning and Games (Vol. 13517, pp.

390–408).

Peraić, I., Grubišić, A., & Pintarić, N. (2025). Machine

Learning in Learning Analytics Dashboards: A

Systematic Literature Review. 2025 IEEE 23rd World

Symposium on Applied Machine Intelligence and

Informatics (SAMI), 000047–000052.

Ramaswami, G., Susnjak, T., Mathrani, A., & Umer, R.

(2023). Use of Predictive Analytics within Learning

Analytics Dashboards: A Review of Case Studies.

Technology, Knowledge and Learning, 28(3), 959–980.

Rivera, A. C., Tapia-Leon, M., & Lujan-Mora, S. (2018).

Recommendation Systems in Education: A Systematic

Mapping Study. In Á. Rocha & T. Guarda (Eds.),

Proceedings of the International Conference on

Information Technology & Systems.

Romero, C., & Ventura, S. (2020). Educational data mining

and learning analytics: An updated survey. WIREs Data

Mining and Knowledge Discovery, 10(3), e1355.

Rowley, J. (2007). The wisdom hierarchy: Representations

of the DIKW hierarchy. Journal of Information

Science, 33(2).

Saito, T., & Watanobe, Y. (2020). Learning Path

Recommendation System for Programming Education

Based on Neural Networks: International Journal of

Distance Education Technologies, 18(1), 36–64.

Salloum, S. A., Alshurideh, M., Elnagar, A., & Shaalan, K.

(2020). Mining in Educational Data: Review and Future

Directions. In A.-E. Hassanien, A. T. Azar, T. Gaber,

D. Oliva, & F. M. Tolba (Eds.), Proceedings of the

International Conference on Artificial Intelligence and

Computer Vision.

Susnjak, T. (2024). Beyond Predictive Learning Analytics

Modelling and onto Explainable Artificial Intelligence

with Prescriptive Analytics and ChatGPT.

International Journal of Artificial Intelligence in

Education, 34(2), 452–482.

Susnjak, T., Ramaswami, G. S., & Mathrani, A. (2022).

Learning analytics dashboard: A tool for providing

actionable insights to learners. International Journal of

Educational Technology in Higher Education, 19(1),

12.

DATA 2025 - 14th International Conference on Data Science, Technology and Applications

582