Design and Implementation of Personalised Learning Support System

Yifei Dou

Information and Computational Science, Xi’an Jiaotong-Liverpool University, Suzhou, China

Keywords: Personalized Learning, System Design, Computer System.

Abstract: Based on the popularity of electronic devices, education is changing from a single learning model to

diversified learning, and the field of education is experiencing a paradigm shift from standardized teaching to

personalized learning. At present, countries are also carrying out the popularization of personalized learning.

This study summarizes the intelligent personalized learning system based on multi-level architecture in view

of the limitations of the traditional education system "one-man", aiming at empowering the education essence

of "teaching students in accordance with their aptitude" through technology. So that students can find a truly

suitable learning mode through the learning system, the system adopts a four-layer architecture design, and

recommends suitable personalized learning systems. Future research will focus on intelligence, explore the

efficiency of personalized learning system, such as solving learning strategy problems through ai, and verify

the usability of personalized learning system, and attract students to accept personalized learning model by

building system innovation. In addition, people's identification and popularization of personalized learning

will also become an important research topic.

1 INTRODUCTION

With the interdisciplinary integration of artificial

intelligence, big data analysis and cognitive science,

the education system has gradually shifted from "one-

to-one" standardized teaching to "people-oriented"

personalized service. This transformation is not only

driven by the development of technology, but also

reflects the return of the essence of education, which

is to respect the individual differences of learners and

realize the goal of teaching students in accordance

with their students' abilities. For learners, finding the

resources that meet their needs from a large number

of learning resources is very important for learning

efficiency and self-study effect (Wang and Gao,

2022).

And with the rapid development of various mobile

devices, such as mobile phones, tablets, laptops,

GPRS technology and Bluetooth technology

popularity, gradually become our main tool for

learning, another new survey conducted by Newzoo's

Global Mobile Market Report 2019 shows that the

popularity of smart phones is high: 82.9% in the UK,

79.9% in Germany, 79.1% in the US, 77.5% in France,

74.3% in Spain, 70.4% in South Korea and 66.3% in

Russia (Bourekkache et al., 2020). In Algeria, the

smartphone penetration rate is 38.1%, which is

considered an important rate (Bourekkache et al.,

2020). Because of the popularity of these devices, it

has become difficult for our study life to leave online

learning, so having an excellent personalized learning

system is essential for learning efficiency. And on a

global scale, The United Nations Educational,

Scientific and Cultural Organization, UNESCO states

that "access to quality education means access to

personalized learning", highly emphasizing the

promise of this approach for addressing uneven

outcomes in school education. As J. Groff puts it,

personalized learning is essential because it is

"entirely consistent with the science of learning"

(Zhao and Wu, 2025).

It is proposed in InfoTech that adaptive resource

recommendations (or building a personalized

learning path) should be made in a dialogue with the

learner, and he should have the possibility to select

some suggested new content or activity. This paper

proposes a knowledge model to ensure adequate

recommendation of dynamically generated learning

paths and to store various approved and well-

annotated learning paths for future reuse (Ivanova,

2023). The architecture design of personalized

learning system is the key to realize personalized

education. An effective personalized learning system

should contain these aspects, including user

Dou, Y.

Design and Implementation of Personalised Learning Support System.

DOI: 10.5220/0014318100004718

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on Engineering Management, Information Technology and Intelligence (EMITI 2025), pages 49-54

ISBN: 978-989-758-792-4

interaction layer, server layer, data access layer,

database layer. The user interaction layer is mainly

responsible for collecting the basic information of

learners, and effectively displaying and

recommending learning resources to students. The

server is mainly responsible for analyzing the user

model and finding a suitable learning path for learners.

The data access layer is mainly responsible for

ensuring the security of data entering the data

database and reviewing whether the data is normal or

not to ensure the security of user data. The database

layer is mainly responsible for storing data and

facilitating learners to call data, and can also give

visual charts for users to view data efficiently and

clearly.

2 THE OVERALL

ARCHITECTURE DESIGN

The user side is responsible for the display and

processing by the web interface, and the interactive

interface directly facing the learner, showing the

graphical interface of the student learning, teacher

monitoring and other modules, as well as the

corresponding security protection. It can intuitively

show the function of the system to the user, provide

users with a convenient and quick experience, but

also consider the user experience, and operation and

maintenance.

The server layer (business logic layer) is the core

area of the system, responsible for receiving and

processing the information transmitted by the user

end, can carry out the centralized management of

complex business, technology selection, code

governance, etc., which may affect the performance.

The database is mainly responsible for providing

an interface for the server to realize the addition,

deletion and modification of data. The use of the

database can make the data more durable and reliable,

and can also make the query and optimization more

efficient, but the problems such as operation and

maintenance complexity and security need to be

solved. Figure 1 shows the overall architecture of

personalized learning.

Figure 1: The overall architecture of personalized learning

(Picture credit: Original).

3 TECHNOLOGY SELECTION

3.1 User Side Development

Figure 2: User-side personalized learning model (Zhang et

al., 2023).

The system is mainly composed of six modules (as

shown in Figure 2): login module, Web log mining

processing module, knowledge base, personalized

recommendation module, regular function module

and learning resource library, as shown in Figure 2

(Zhang et al., 2023). When users log in, they must

first register, and carefully fill in the questionnaire

personalized information materials, and establish the

initial user interest matrix for resources to avoid the

EMITI 2025 - International Conference on Engineering Management, Information Technology and Intelligence

cold start problem of the system (Zhang et al., 2023).

In the system have the function of traditional online

learning system, such as user management, learning

resource management, online testing, online

answering questions and teaching evaluation, etc., at

the same time the system also includes a core of

personalized intelligent module - Web log mining

processing module, in order to realize the learners and

page clustering and learners learning rule, the

analysis of learning style. When a user access to the

system, the operation module mainly complete the

following two aspects: first, using the K - Means

clustering algorithm is analyzed, according to the

learners and page two objects according to the

similarity between them is divided into several groups,

constitute a set of similar objects. If some pages are

found to belong to the same group, other pages in the

same group can be recommended to learners when

learners access the pages (Zhang et al., 2023).

Since the user side is directly oriented to learners,

there are many frameworks that can provide a better

experience for users. For example, FeedbackFlow

Dynamics studies the evolution of real-time feedback

systems and user behavior (Devi et al., 2025). It

investigates the impact of information and timed

feedback on user engagement. By allowing users to

modify timing and input types, interfaces become

more interesting and responsive. The EmoFeedback

interaction model incorporates user emotions into the

design process, recognizing their significant impact

on interface usability and aesthetics, by analyzing and

responding to emotional signals, it enhances the user

experience and contributes to the development of

emotionally intelligent interfaces (Devi et al., 2025).

The DynamicAdapt UX framework ensures a more

satisfying, personalized, and contextualized

experience for customers (Devi et al., 2025). Reactive

flow design model assessment in a variety of screen

sizes and optimize the design of the data flow and

equipment (Devi et al., 2025).

At the same time, interface design plays a vital

role in improving user experience and promoting

product development (Qiu, 2024). By implementing

a series of targeted optimization measures, such as

adopting card design, expanding the search bar, and

increasing the user comment section, we successfully

solved the problem of information overload caused by

user feedback. This makes the overall site clear visual

effect is more intuitive, allows the user to quickly find

interested project (Qiu, 2024). These optimizations

not only improve user efficiency, but also

significantly improve user satisfaction and loyalty

(Qiu, 2024).

User experience evaluation in this framework

employs the PrEmo (Product Emotion Measurement)

instrument, a validated psychometric tool for

quantifying affective responses to design stimuli

through 14 discrete emotional dimensions (Jiao,

2022). Grounded in Desmet's theoretical emotion

framework, the methodology operationalizes four

core components: 1) product evaluation protocols, 2)

attentional engagement metrics, 3) appraisal-induced

affective states, and 4) product attribute-emotion

mapping. Participants interact with purpose-designed

product prototypes that embody specific functional

objectives, aesthetic principles, and attention-guiding

features. Affective outcomes are subsequently

quantified through dynamic visual stimulus

presentations using affect-encoded animation

sequences (Jiao, 2022).

3.2 Server Layer

3.2.1 Micro Server Scheme

The combination of using Spring Boot with Spring

Cloud Spring Boot with its "convention over

configuration" design concept, provides developers

with a quick build Spring application convenient tools.

Spring Cloud further spring-based Boot, focusing on

service governance under the micro service

architecture, provides a set of perfect solutions,

covering service routing, registration and discovery,

load balance, monitoring, and many other aspects.

These features make the Spring Cloud micro service

architecture for building the preferred method of

(Chen et al., 2024). The use of microservice cluster

can disintegrate the complex scientific research

management system into a series of independent and

loosely coupled services, which significantly

improves the automation and information level of the

system (Wang, 2024).

3.2.2 Distributed Server Solution

This distributed server architecture employs multiple

nodes to minimize service delays in real-time systems

(Yanase et al., 2021). Under this design, each user

connects to a dedicated server node for event

transmission. The assigned server aggregates user-

generated events, disseminates them across the

network for synchronization, processes all collective

events through coordinated computation, and

ultimately relays the processed results back to

corresponding users for subsequent operations

(Yanase et al., 2021). The maximum number of users

in the server is given in the form of the server capacity,

Design and Implementation of Personalised Learning Support System

and to process events in the order in which they occur

at the user side, the delay between the user and the

server and the delay between the servers are corrected

to the maximum delay DU between the user and the

server and the maximum delay DS between the

servers, respectively (Yanase et al., 2021). Therefore,

in the Distributed Server Processing (DSP) approach,

it requires 2DU+DS. Compared with the CP, it will

have an extra step to events with other server

synchronization (Yanase et al., 2021). However,

since the server location is closer, the latency between

the user and the server is lower, so the communication

latency can be reduced (Yanase et al., 2021).

3.3 Data Access Layer

The overall system architecture (as shown in Figure

3) includes three parts, which are the front-end and

back-end management interface, the server and the

client, and the program code is deployed on the server.

The system takes the operation and maintenance

control module as the core, including the access

entrance of safety certification, configuration

management and monitoring and audit functions. The

safety certification covers two sub-modules of login

authentication and proxy authentication, and the

monitoring and accounting page sets up three sub-

modules of operation and maintenance monitoring,

session cutting and audit analysis. The configuration

management page covers three sub-modules of

personnel, equipment and policy management (Jia,

2024).

Figure. 3: Overall architecture of Data access layer (Jia,

2024).

There are many kinds of database security threats,

which need to be paid attention to and protected. The

following article will list the security matters that

need to be paid attention to SQL Injection. This is a

form of attack in which malicious code into the front

end (Web application) and transmitted to the backend

database (Jiao, 2022).

Due to SQL injection, computer attackers have

unrestricted access to all the data contained in the

database. There are two possible types of such code

attacks: SQL injection against standard databases and

NoSQL injection against massive databases (Jiao,

2022).

Database vulnerabilities and misconfigurations.

This can also happen when the database is found to

be absolutely inaccessible due to misconfiguration

(Jiao, 2022). Many database systems maintain preset

credentials and setup options by design, a critical

consideration given that malicious actors possess

advanced technical expertise. These cyber attackers

routinely leverage architectural deficiencies and

improper configuration settings as entry points to

compromise organizational networks (Jiao, 2022).

Denial of Service (DoS) incidents specifically

target system availability by overwhelming database

infrastructure, resulting in two primary consequences:

impaired operational efficiency and potential service

interruption that renders critical assets unreachable

(Jiao, 2022). While such attacks preserve data

confidentiality, they force organizations to incur

substantial operational expenses through extended

downtime and remediation efforts (Jiao, 2022).

Ultimately, persistent unavailability of web services

fundamentally undermines the service's core value

proposition, negating its functional purpose despite

maintained data integrity (Jiao, 2022).

Decentralized data management presents critical

security challenges when handling sensitive

information. Organizations frequently maintain

extensive repositories of confidential records without

implementing comprehensive data governance

frameworks, creating attack surface proliferation

through orphaned datasets and legacy storage systems

(Jiao, 2022). Compounding this challenge, the

continuous influx of newly generated critical

information within dynamic data ecosystems makes

systematic monitoring inherently complex. This

operational reality exposes both archival and recently

ingested data assets to potential security breaches

before proper safeguards can be implemented (Jiao,

2022).

Database backup risks. It is best to back up a

proprietary repository within the specified time frame

(Jiao, 2022). Curiously, however, database backup

files are also completely unattacked. As a result,

database backup leaks are a frequent security breach

(Jiao, 2022).

EMITI 2025 - International Conference on Engineering Management, Information Technology and Intelligence

Excessive privilege allocation in database systems

creates substantial security exposure. While role-

based access differentiation is fundamental to

database operations (Jiao, 2022), three primary

privilege escalation vectors emerge: non-essential

privilege retention (maintaining permissions

exceeding operational requirements), legitimate

access overreach (exploiting authorized privileges for

unauthorized activities), and dormant credential

exploitation (leveraging inactive access rights).

Empirical studies indicate approximately 80% of

corporate network breaches originate from insider

threats, with over-provisioned access rights

significantly increasing attack surface vulnerability

through privilege creep mechanisms (Jiao, 2022).

Malicious software propagation vectors

demonstrate sophisticated attack methodologies

through credential hijacking mechanisms. Advanced

persistent threats typically compromise endpoint

devices to establish command-and-control channels,

subsequently exploiting legitimate user credentials

through three operational phases: session co-option

(hijacking authenticated connections), privilege

escalation (leveraging authorized access levels), and

lateral movement (penetrating enterprise networks

via compromised accounts). Forensic analysis

confirms 62% of such breaches originate from

malicious payloads executing privilege escalation

protocols using authenticated user contexts (Jiao,

2022).

3.4 The Database Layer

Cache operates as an advanced object-oriented

database system, utilizing a multi-dimensional

transaction processing architecture to enable

distributed data management. This system establishes

a consistent data model structure, allowing data

manipulation via standard SQL queries while offering

integrated development utilities to accelerate

database implementation, particularly for web-based

applications. Additionally, the platform features

XML object compatibility and enables cross-

language interoperability through support for .NET,

ActiveX, C++, EJB, Java, JDBC, ODBC, SOAP, Perl,

Python, and XML protocols (Goswami and Sharma,

2021).

Teiid, a data integration technology, supports

virtualization of various types of databases; Through

such virtual databases, data sources such as relational

databases, Web databases, and application software

such as ERP and CRM can be accessed in real time

(Wada et al., 2010). Virtualization technology so that

data analysts or other users can use all ubiquitous

databases as if they were a single database, thus

helping to reduce the workload of users (Wada et al.,

2010).

4 CONCLUSION

Based on the interdisciplinary perspective, this study

summarizes an intelligent personalized learning

system. Through the four-layer architecture of "user

interaction layer, business logic layer, data access

layer and database layer", it integrates a variety of

methods and combines a variety of hot core

technologies to solve the pain points of low resource

adaptability and rigid learning path in traditional

education. The study shows that the learning system

is not only a technology, but also the promotion of

education reform. In the future, the technology will

take the essence of education as the carrier to build a

humanized and intelligent learning ecological model.

There are still many challenges in personalized

learning system, such as data collection and privacy,

algorithm bias, user diversity and effect verification.

In the future, both technology developers and users

will collaborate to solve these challenges.

REFERENCES

Bourekkache, S., Tigane, S., Kazar, O., and Kahloul, L.:

‘Mobile and personalized learning system for computer

science students,’ 2020 Sixth International Conference

on e-Learning (econf), Sakheer, Bahrain, 2020, pp.

189-193

Chen, Z., Wu, J., Ma, F. Z.: ‘Application Development

Research based on microservice architecture’.

Technology News. 2024

Devi, V. S. A. et al.: ‘Designing Intuitive User Interfaces in

Human-Computer Interaction for Enhanced Digital

Experience,’ 2025 International Conference on

Intelligent Control, Computing and Communications

(IC3), Mathura, India, 2025, Pp. 637-643

Goswami, P. K., and Sharma, A.: ‘An Intelligent Shift

Towards High-Performance Database Engines

Including Autonomous Databases Using Derived

Transient Objects from the Persistent at the Runtime,’

2021 IEEE International Conference on Technology,

Research, and Innovation for Betterment of Society

(TRIBES), Raipur, India, 2021, pp. 1-5

Ivanova, T. I.: ‘Knowledge-Based Semi-Automatic

Selection of Personalized Learning Paths,’ 2023

International Conference on Information Technologies

(InfoTech), Varna, Bulgaria, 2023, pp. 1-4

Jia, J. Y.: ‘Application of Big Data Technology and

MySQL Database’. Integrated Circuit Applications.

2024

Design and Implementation of Personalised Learning Support System

Jiao, Z.: ‘Research and Application of Human-Computer

Interface Based on User Experience,’ 2022 7th

International Conference on Intelligent Computing and

Signal Processing (ICSP), Xi'an, China, 2022, Pp.

1417-1420

Jiao, Z.: ‘Research and Application of Human-Computer

Interface Based on User Experience,’ 2022 7th

International Conference on Intelligent Computing and

Signal Processing (ICSP), Xi'an, China, 2022, Pp.

1417-1420

Qiu, L.: ‘Application of Deep Learning-Based User Intent

Recognition in Human-Computer Interaction,’ 2024

5th International Conference on Intelligent Computing

and Human-Computer Interaction (ICHCI), Nanchang,

China, 2024, pp. 549-552

Wada, Y., Watanabe, Y., Syoubu, K., Sawamoto, J., and

Katoh, T.: ‘Virtual Database Technology for

Distributed Database,’ 2010 IEEE 24th International

Conference on Advanced Information Networking and

Applications Workshops, Perth, WA, Australia, 2010,

pp.214-219

Wang, H., and Gao, Z.: ‘Research on Personalized Service

Path of Learning Resources by Data Driven,’ 2022

International Symposium on Educational Technology

(ISET), Hong Kong, Hong Kong, 2022, pp. 143-147

Wang, Y.: ‘Research on Scientific Research Management

System Based on Microservice Architecture’.

Electronic Engineering and Products World. 2024

Yanase, S., He, F., and Oki, E.: ‘Approximation Algorithms

to Distributed Server Allocation With Preventive Start-

Time Optimization Against Server A Failure’, IEEE

Networking in Letters, vol. 3, no. 4, pp. 191-195, Dec.

2021

Zhang, J., Jing, S. Y., Linag, S.: ‘Research on personalized

learning resource recommendation model based on

Web log mining’. 2023 3rd International Conference on

Electronic Information Engineering and Computer

Communication (EIECC). 2023

Zhao, J. Y., Wu, J.: ‘Artificial Intelligence Empowering

Personalized Learning: The Path Analysis of K-12

Education Digitization in the United States’. Journal of

Comparative Education. 2025

EMITI 2025 - International Conference on Engineering Management, Information Technology and Intelligence