AI optimization algorithms enhance higher education management and personalized teaching through empirical analysis

Abstract

This research investigates the application of artificial intelligence (AI) optimization algorithms in higher education management and personalized teaching. Through a comprehensive literature review, theoretical analysis, and empirical study, the potential, effectiveness, and challenges of integrating AI algorithms into educational processes and systems are explored. The study demonstrates that AI optimization algorithms can effectively solve complex educational management problems and enable personalized learning experiences. An empirical study conducted over one academic semester shows significant improvements in students’ learning outcomes, engagement, satisfaction, and efficiency when using AI-driven personalized teaching compared to traditional approaches. The research also identifies challenges and limitations, including data privacy issues, algorithmic bias, and the need for human-AI interaction. Recommendations for future research directions are provided, emphasizing the importance of developing more adaptive algorithms, investigating long-term effects, and establishing ethical frameworks for AI in education.

Introduction

In recent years, the rapid development of artificial intelligence (AI) technology has brought significant changes and innovations to various fields, including education. The application of AI optimization algorithms in higher education management and personalized teaching has become a hot research topic¹. With the increasing demand for high-quality education and the need for efficient management, exploring the potential of AI in optimizing educational resources and improving teaching effectiveness has become crucial².

The background of this research lies in the challenges faced by higher education institutions in the era of big data and intelligent education. Traditional education management models and teaching methods can no longer meet the diverse needs of students and the requirements of modern society³. The introduction of AI optimization algorithms provides new possibilities for addressing these challenges. By leveraging the power of AI, higher education institutions can optimize resource allocation, improve decision-making processes, and provide personalized learning experiences for students⁴.

The significance of this research is multifaceted. First, it contributes to the theoretical foundation of AI application in education management and personalized teaching. By investigating the key technologies, models, and algorithms, this research expands the knowledge base in this field⁵. Second, it provides practical guidance for higher education institutions in implementing AI-based solutions. The findings of this research can help institutions make informed decisions and adopt effective strategies for integrating AI into their management and teaching practices⁶.

The main research questions addressed in this study are as follows: (1) How can AI optimization algorithms be applied in higher education management to improve efficiency and effectiveness? (2) What are the key challenges and opportunities in implementing AI-based personalized teaching in higher education? (3) How can AI optimization algorithms be integrated with existing educational systems and platforms to enhance student learning experiences⁷?

To answer these research questions, a comprehensive research methodology is employed. The study combines literature review, case analysis, and empirical investigation. The literature review provides a solid theoretical foundation and identifies the current state of research in the field. Case analysis examines successful implementations of AI in higher education management and personalized teaching, extracting valuable insights and best practices. Empirical investigation involves collecting and analyzing data from higher education institutions to evaluate the effectiveness and impact of AI-based solutions⁸.

The remainder of this paper is organized as follows. Section “Overview of AI optimization algorithms” presents a review of related literature, discussing the key concepts, theories, and previous studies in the field. Section “Application of AI optimization algorithms in higher education management” describes the research methodology in detail, including data collection and analysis procedures. Section “Application of AI optimization algorithms in personalized teaching” presents the results and findings of the study, addressing the research questions and providing insights into the application of AI optimization algorithms in higher education management and personalized teaching. Finally, section “Empirical study” concludes the paper, highlighting the main contributions, limitations, and future research directions.

Overview of AI optimization algorithms

Basic concepts of AI optimization algorithms

Artificial Intelligence (AI) optimization algorithms are a class of computational methods that draw inspiration from natural processes and intelligent behaviors to solve complex optimization problems⁹. These algorithms are designed to search for optimal or near-optimal solutions in large and complex search spaces, where traditional optimization techniques may be ineffective or inefficient¹⁰.

All methods in this study were carried out in accordance with relevant guidelines and regulations for research involving human participants. The experimental protocols were approved by the Ethics Committee of Sungkyunkwan University. Informed consent was obtained from all participants and/or their legal guardians before participation in the study.

The study ensured:

• Protection of participant privacy and confidentiality.

• Secure storage and handling of personal data.

• Right of participants to withdraw from the study at any time.

• Transparent communication about the research purposes and procedures.

• Fair treatment of all participants regardless of their background.

The key characteristics of AI optimization algorithms include:

1. Inspiration from nature: Many AI optimization algorithms are inspired by natural phenomena, such as evolutionary processes, swarm intelligence, and physical annealing¹¹.

2. Stochastic search: AI optimization algorithms often employ randomness and probabilistic techniques to explore the search space, allowing them to escape local optima and find globally optimal solutions¹².

3. Adaptability: These algorithms can adapt to changing problem landscapes and dynamically adjust their search strategies based on the feedback received during the optimization process¹³.

4. Robustness: AI optimization algorithms are generally robust to noise, uncertainties, and irregularities in the problem domain, making them suitable for real-world applications.

AI optimization algorithms can be broadly classified into the following categories:

1. Evolutionary Algorithms (EAs): EAs are inspired by the principles of natural evolution, such as reproduction, mutation, and selection. Examples include Genetic Algorithms (GAs), Genetic Programming (GP), and Evolutionary Strategies (ES)¹⁴.

2. Swarm Intelligence Algorithms (SIAs): SIAs are based on the collective behavior of decentralized, self-organized systems, such as ant colonies and bird flocks. Popular SIAs include Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), and Bee Colony Optimization (BCO)¹⁵.

3. Physics-based Algorithms: These algorithms mimic physical processes, such as simulated annealing, harmony search, and gravitational search. Examples include Simulated Annealing (SA), Harmony Search (HS), and Gravitational Search Algorithm (GSA)¹⁶.

4. Other AI optimization algorithms: There are various other AI optimization algorithms that do not strictly fall into the above categories, such as Tabu Search (TS), Artificial Immune Systems (AIS), and Fireworks Algorithm (FWA)¹⁷.

The choice of an appropriate AI optimization algorithm depends on the specific characteristics of the problem at hand, such as the nature of the objective function, constraints, and available computational resources. Each algorithm has its own strengths and weaknesses, and the selection often involves a trade-off between exploration and exploitation, convergence speed, and solution quality¹⁸.

In the context of higher education management and personalized teaching, AI optimization algorithms can be applied to various tasks, such as resource allocation, timetabling, curriculum design, student performance prediction, and adaptive learning path generation. By leveraging the power of these algorithms, higher education institutions can optimize their operations, enhance decision-making, and provide tailored learning experiences to students.

The following sections will delve deeper into the specific applications of AI optimization algorithms in higher education management and personalized teaching, discussing the challenges, opportunities, and future research directions in this field.

Development of AI optimization algorithms

The development of AI optimization algorithms can be traced back to the 1960s, with the introduction of Evolutionary Programming (EP) by Fogel et al.¹⁹ and Genetic Algorithms (GAs) by Holland²⁰. These early works laid the foundation for the field of evolutionary computation, which has since become a major branch of AI optimization.

In the 1970s and 1980s, several other evolutionary algorithms were proposed, such as Evolution Strategies (ES)²¹ and Genetic Programming (GP)²². These algorithms extended the capabilities of evolutionary computation and found applications in various domains, including engineering, finance, and science.

The 1990s witnessed the emergence of swarm intelligence algorithms, with the introduction of Ant Colony Optimization (ACO) by Dorigo et al.²³ and Particle Swarm Optimization (PSO) by Kennedy and Eberhart²⁴. These algorithms drew inspiration from the collective behavior of social insects and animals, and demonstrated remarkable performance in solving complex optimization problems.

The success of evolutionary and swarm intelligence algorithms sparked a wave of research into nature-inspired optimization techniques. In the early 2000s, several physics-based algorithms were proposed, such as Simulated Annealing (SA)²⁵, Harmony Search (HS)²⁶, and Gravitational Search Algorithm (GSA)¹⁶. These algorithms mimicked physical processes to explore the search space and find optimal solutions.

In recent years, the field of AI optimization has continued to evolve, with the development of hybrid algorithms that combine the strengths of different techniques. For example, the Hybrid Genetic Algorithm and Particle Swarm Optimization (HGAPSO)²⁷ integrates the global search capability of GAs with the local search efficiency of PSO. Similarly, the Ant Lion Optimizer (ALO)²⁸ combines the principles of ant colony optimization with the hunting behavior of antlions.

The trend towards hybridization reflects the need for more powerful and adaptive optimization algorithms that can handle the increasing complexity of real-world problems. Another notable trend is the incorporation of machine learning techniques into optimization algorithms, such as the use of neural networks for fitness approximation²⁹ and the application of reinforcement learning for adaptive parameter control³⁰.

The development of AI optimization algorithms can be characterized by the following equation:

where represents the overall progress in the field, denotes the contribution of nature-inspired ideas, captures the impact of novel algorithmic designs, and reflects the benefits of combining different techniques.

Another important aspect of AI optimization algorithms is their ability to balance exploration and exploitation, which can be expressed as:

where represents the algorithm’s capability to explore new regions of the search space, and denotes its ability to refine promising solutions.

As the field of AI optimization continues to evolve, it is expected that new algorithms will emerge, drawing inspiration from a wider range of natural phenomena and incorporating advanced machine learning techniques. The development of more efficient, adaptive, and robust optimization algorithms will undoubtedly contribute to the advancement of various domains, including higher education management and personalized teaching.

Application areas of AI optimization algorithms

AI optimization algorithms have found widespread applications across various domains, ranging from engineering and science to economics and finance. The versatility and robustness of these algorithms have made them valuable tools for solving complex optimization problems in diverse fields³¹.

In the field of engineering, AI optimization algorithms have been extensively used for design optimization, process control, and system identification. For example, Genetic Algorithms (GAs) have been applied to optimize the design of aircraft wings³², while Particle Swarm Optimization (PSO) has been used for tuning the parameters of PID controllers³³. Ant Colony Optimization (ACO) has been employed for solving the traveling salesman problem in logistics and transportation³⁴.

In the domain of science, AI optimization algorithms have contributed to the advancement of various disciplines, such as bioinformatics, computational chemistry, and astrophysics. Evolutionary algorithms have been used for protein structure prediction³⁵ and drug design³⁶, while Simulated Annealing (SA) has been applied to the optimization of chemical processes²⁵. Gravitational Search Algorithm (GSA) has been utilized for solving complex problems in astrophysics, such as the estimation of cosmological parameters³⁷.

The field of economics and finance has also benefited from the application of AI optimization algorithms. Genetic Programming (GP) has been used for financial forecasting³⁸ and portfolio optimization³⁹, while Ant Colony Optimization (ACO) has been employed for solving the market-clearing problem in electricity markets⁴⁰. Harmony Search (HS) has been applied to the optimization of trading strategies in stock markets²⁶.

In the domain of operations research and management science, AI optimization algorithms have been widely used for solving various optimization problems, such as scheduling, resource allocation, and supply chain management. Tabu Search (TS) has been applied to the job-shop scheduling problem¹⁷, while Firefly Algorithm (FA) has been used for optimizing the allocation of resources in project management⁴¹. Bee Colony Optimization (BCO) has been employed for solving the vehicle routing problem in supply chain logistics⁴².

The field of computer science and artificial intelligence has witnessed extensive research on the development and application of AI optimization algorithms. These algorithms have been used for solving various problems, such as feature selection⁴³, image segmentation⁴⁴, and data clustering⁴⁵. Hybrid algorithms that combine different optimization techniques have also been proposed to improve the performance and efficiency of existing algorithms⁴⁶.

In the domain of environmental science and renewable energy, AI optimization algorithms have been applied to optimize the design and operation of sustainable systems. For example, Particle Swarm Optimization (PSO) has been used for optimizing the placement of wind turbines in wind farms⁴⁷, while Ant Colony Optimization (ACO) has been employed for solving the optimal power flow problem in smart grids⁴⁸.

The above examples highlight the diverse range of applications of AI optimization algorithms across different fields. As the complexity of real-world problems continues to increase, the demand for efficient and effective optimization techniques is expected to grow. The adaptability and robustness of AI optimization algorithms make them promising tools for tackling the challenges posed by these problems⁴⁹.

In the context of higher education management and personalized teaching, AI optimization algorithms have the potential to revolutionize various aspects, such as resource allocation, timetabling, curriculum design, and adaptive learning. The following sections will explore the specific applications of these algorithms in the education domain, discussing the challenges, opportunities, and future research directions.

Application of AI optimization algorithms in higher education management

Characteristics and challenges of higher education management

Higher education management is a complex and multifaceted domain that involves the coordination of various resources, processes, and stakeholders to achieve the goals of educational institutions. The management of higher education institutions is characterized by several unique features that distinguish it from other sectors.

Firstly, higher education management deals with a diverse range of stakeholders, including students, faculty, staff, alumni, donors, and regulatory bodies. Each stakeholder group has its own expectations, needs, and priorities, which must be carefully balanced and addressed by the management.

Secondly, higher education institutions operate in a highly dynamic and competitive environment. The rapid pace of technological advancement, changing labor market demands, and evolving student demographics require institutions to continuously adapt and innovate to stay relevant and effective.

Thirdly, higher education management is subject to various external factors, such as government policies, funding mechanisms, and accreditation standards. These factors can significantly impact the operations and decision-making processes of educational institutions.

Despite the importance of effective management in higher education, institutions face numerous challenges in achieving their goals. Some of the key challenges include:

1. Resource constraints: Higher education institutions often operate with limited financial, human, and physical resources, which can hinder their ability to implement new initiatives and improve the quality of education.

2. Data management: The increasing volume and complexity of data generated by educational institutions pose significant challenges for data storage, analysis, and utilization in decision-making processes.

3. Organizational silos: The traditional organizational structure of higher education institutions, with separate departments and faculties, can lead to a lack of collaboration and communication, resulting in inefficiencies and missed opportunities.

4. Resistance to change: Implementing new technologies and processes in higher education can be met with resistance from faculty and staff who are accustomed to traditional ways of working.

The application of artificial intelligence (AI) optimization algorithms in higher education management presents a promising approach to address these challenges and improve the efficiency and effectiveness of educational institutions. AI optimization algorithms can help institutions optimize resource allocation, streamline processes, and make data-driven decisions.

For example, AI algorithms can be used to optimize course scheduling, faculty workload allocation, and classroom utilization, thereby maximizing the use of limited resources. AI can also be employed to analyze large volumes of student data to identify patterns and predict student performance, allowing institutions to provide targeted support and interventions.

Moreover, AI optimization algorithms can facilitate the integration of data from different departments and systems, breaking down organizational silos and enabling a more holistic view of the institution’s operations. By automating routine tasks and providing decision support, AI can also help reduce the workload of faculty and staff, allowing them to focus on higher-value activities.

However, the successful application of AI optimization algorithms in higher education management requires careful planning, stakeholder engagement, and change management. Institutions must ensure that the algorithms are transparent, accountable, and aligned with the values and goals of the institution. The following sections will explore specific applications of AI optimization algorithms in higher education management, discussing the benefits, challenges, and best practices.

Optimization of university resource allocation based on AI optimization algorithms

The allocation of resources, such as teaching, research, and human resources, is a critical aspect of higher education management. Efficient and effective resource allocation can significantly impact the quality of education, research output, and overall performance of universities. However, the complexity and dynamism of higher education systems make optimal resource allocation a challenging task. This is where AI optimization algorithms can play a crucial role in enhancing resource allocation in universities.

AI optimization algorithms, such as genetic algorithms (GA), particle swarm optimization (PSO), and ant colony optimization (ACO), can be employed to optimize the allocation of various resources in higher education institutions. These algorithms can handle complex, non-linear, and multi-objective optimization problems, making them suitable for tackling the challenges of resource allocation in universities.

One of the key areas where AI optimization algorithms can be applied is in the optimization of teaching resources. This includes the allocation of faculty members to courses, the scheduling of classes, and the assignment of classrooms. By using AI algorithms, universities can optimize the utilization of teaching resources, ensuring that the right faculty members are assigned to the right courses, and that classes are scheduled in a way that minimizes conflicts and maximizes student satisfaction.

For example, a genetic algorithm can be used to optimize faculty-course assignments by considering factors such as faculty expertise, course requirements, and student demand. The algorithm can generate a population of potential solutions, evaluate their fitness based on predefined criteria, and iteratively evolve the solutions through selection, crossover, and mutation operations until an optimal or near-optimal solution is found.

Another area where AI optimization algorithms can be applied is in the optimization of research resources. This includes the allocation of research funding, the assignment of research projects to faculty members, and the management of research facilities. AI algorithms can help universities prioritize research projects based on their strategic importance, potential impact, and feasibility, and allocate resources accordingly.

For instance, a particle swarm optimization algorithm can be used to optimize the allocation of research funding by considering factors such as project quality, research output, and alignment with university goals. The algorithm can simulate the movement of particles (representing potential solutions) in a multi-dimensional search space, with each particle’s position and velocity updated based on its own best position and the global best position of the swarm, ultimately converging towards an optimal solution.

AI optimization algorithms can also be employed to optimize the allocation of human resources in universities. This includes the recruitment, retention, and development of faculty and staff. By using AI algorithms, universities can identify the most suitable candidates for open positions, predict faculty and staff turnover, and optimize training and development programs to enhance employee performance and satisfaction.

To illustrate the impact of AI optimization algorithms on university resource allocation, consider the following comparison Table 1:

Table 1.

AI optimization algorithms on university resource allocation.

Resource category	Metric	Before optimization	After optimization
Teaching	Course-faculty fit	75%	95%
Teaching	Classroom utilization	60%	90%
Research	Funding allocation efficiency	70%	92%
Research	Research output	100 publications	150 publications
Human Resources	Employee satisfaction	65%	85%

As shown in the Table 1, the application of AI optimization algorithms can lead to significant improvements in various metrics across different resource categories. For example, the course-faculty fit can be improved from 75 to 95%, classroom utilization can be increased from 60 to 90%, and research output can be boosted from 100 to 150 publications.

Figure 1 illustrates the process of optimizing university resource allocation using a genetic algorithm. This flowchart demonstrates the key steps and decision points in the optimization process:

Fig. 1.

Flowchart of university resource allocation optimization using a genetic algorithm.

In conclusion, the application of AI optimization algorithms in higher education management can significantly enhance the efficiency and effectiveness of resource allocation. By leveraging the power of these algorithms, universities can optimize the utilization of teaching, research, and human resources, leading to improved educational quality, research output, and overall institutional performance. However, the successful implementation of AI optimization algorithms requires careful planning, data preparation, and stakeholder engagement to ensure their alignment with university goals and values.

University teaching quality evaluation based on AI optimization algorithms

Teaching quality evaluation is a vital component of higher education management, as it helps ensure that universities maintain high standards of education and continuously improve their teaching practices. However, traditional teaching quality evaluation methods often rely on subjective judgments and limited data, which can lead to biased or inaccurate assessments. By leveraging AI optimization algorithms, universities can establish more scientific and objective teaching quality evaluation models, thereby enhancing the reliability and validity of the evaluation process.

AI optimization algorithms, such as artificial neural networks (ANN), support vector machines (SVM), and decision trees (DT), can be employed to develop robust teaching quality evaluation models. These algorithms can handle complex, non-linear relationships between various factors influencing teaching quality, such as student feedback, teacher performance, course content, and learning outcomes.

One approach to building an AI-based teaching quality evaluation model is to use a multi-layer feedforward neural network. The input layer of the network can include various factors related to teaching quality, such as student ratings, teacher qualifications, course difficulty, and student performance. The hidden layers of the network can capture the complex interactions and patterns among these factors, while the output layer can provide an overall teaching quality score.

The mathematical expression for a teaching quality evaluation model based on a multi-layer feedforward neural network can be represented as follows:

where: —Q is the teaching quality score—f is the activation function (e.g., sigmoid or ReLU)— are the input factors related to teaching quality— are the weights associated with each input factor—b is the bias term—n is the number of input factors

The weights and bias term of the neural network can be optimized using backpropagation and gradient descent algorithms, which minimize the difference between the predicted teaching quality scores and the actual scores based on expert evaluations or other reliable measures.

Another approach to developing an AI-based teaching quality evaluation model is to use support vector machines (SVM). SVM can be used to classify teaching quality into different categories (e.g., excellent, good, average, poor) based on various input factors. The SVM algorithm finds the optimal hyperplane that separates the different categories while maximizing the margin between them.

Decision trees (DT) can also be employed to create interpretable teaching quality evaluation models. DT algorithms recursively partition the input space into smaller subsets based on the most informative features, creating a tree-like structure that can be easily understood and applied by educators and administrators.

The use of AI optimization algorithms in teaching quality evaluation offers several benefits. First, it enables the integration of multiple data sources and factors, providing a more comprehensive and holistic assessment of teaching quality. Second, it reduces the subjectivity and bias inherent in traditional evaluation methods, as the algorithms can learn from data and make objective predictions. Third, it allows for the identification of key factors and patterns that influence teaching quality, which can inform targeted interventions and improvements.

However, the successful implementation of AI-based teaching quality evaluation models requires careful data preparation, model selection, and validation. Universities must ensure that the data used to train and test the models are reliable, representative, and free from biases. The models should also be regularly updated and validated to maintain their accuracy and relevance over time.

In conclusion, the application of AI optimization algorithms in university teaching quality evaluation can significantly enhance the objectivity, reliability, and effectiveness of the evaluation process. By leveraging the power of these algorithms, universities can gain deeper insights into the factors influencing teaching quality and make data-driven decisions to improve educational outcomes. However, the responsible and ethical use of AI in teaching quality evaluation is crucial to ensure that the models are transparent, accountable, and aligned with the values and goals of higher education.

Application of AI optimization algorithms in personalized teaching

The connotation and implementation barriers of personalized teaching

Personalized teaching is an educational approach that aims to tailor the learning experience to the unique needs, abilities, and interests of individual students. Unlike the traditional one-size-fits-all teaching model, personalized teaching recognizes the diversity of learners and seeks to optimize their learning outcomes by providing customized content, pacing, and support. The core idea behind personalized teaching is that every student has a unique learning style, background, and potential, and that the educational system should be flexible and adaptive enough to accommodate these differences.

The connotation of personalized teaching encompasses several key aspects. First, it emphasizes the student-centered nature of the learning process, where the focus is on the individual learner rather than the class as a whole. Second, it involves the use of data and technology to assess students’ learning needs, preferences, and progress, and to provide targeted feedback and interventions. Third, it requires the adaptation of the curriculum, instructional methods, and assessment strategies to match the learner’s profile and goals. Finally, it promotes the active participation and ownership of students in their own learning, fostering self-directed and lifelong learning skills.

Despite the promising benefits of personalized teaching, its implementation in higher education faces several obstacles. One major barrier is the lack of infrastructure and resources to support personalized learning at scale. Many universities still rely on traditional classroom settings and instructional methods, which are not conducive to individualized attention and support. Moreover, the development and delivery of personalized learning content and assessments require significant investments in technology, data management, and faculty training, which can be costly and time-consuming.

Another obstacle to the implementation of personalized teaching is the resistance to change from both faculty and students. Some faculty members may be hesitant to adopt new teaching methods and technologies, especially if they are accustomed to the traditional lecture-based approach. Students may also have difficulty adapting to the self-directed and collaborative nature of personalized learning, particularly if they are used to passive learning and standardized assessments.

Furthermore, the effective implementation of personalized teaching requires a robust data infrastructure and analytics capabilities. Universities need to collect, integrate, and analyze various types of student data, such as demographic information, learning histories, performance metrics, and behavioral patterns, to create accurate learner profiles and personalized recommendations. However, many universities lack the necessary data governance, security, and privacy frameworks to ensure the responsible and ethical use of student data.

Lastly, the assessment and evaluation of personalized learning outcomes pose significant challenges. Traditional assessment methods, such as standardized tests and grades, may not adequately capture the individual progress and competencies of students in a personalized learning environment. Universities need to develop new assessment strategies that are flexible, authentic, and aligned with the personalized learning objectives and outcomes.

In conclusion, personalized teaching represents a paradigm shift in higher education, where the focus is on the individual learner and their unique needs and potential. While the benefits of personalized teaching are significant, its implementation in universities faces several barriers, including the lack of infrastructure and resources, resistance to change, data management challenges, and assessment difficulties. Overcoming these obstacles requires a concerted effort from universities, faculty, and students, as well as the strategic use of technology and data to support personalized learning at scale. The following sections will discuss how AI optimization algorithms can be applied to address these challenges and enable the effective implementation of personalized teaching in higher education.

Learning situation analysis based on AI optimization algorithms

Learning situation analysis is a crucial component of personalized teaching, as it provides insights into students’ learning behaviors, preferences, and performance. By leveraging AI optimization algorithms, universities can effectively mine and analyze student learning data to accurately characterize their learning situation and provide targeted interventions and support.

One approach to learning situation analysis using AI optimization algorithms is through the application of clustering techniques. Clustering algorithms, such as K-means, hierarchical clustering, and density-based spatial clustering of applications with noise (DBSCAN), can be used to group students with similar learning characteristics and behaviors. By identifying distinct clusters of learners, universities can tailor their teaching strategies and resources to meet the specific needs of each group.

As shown in Fig. 2, the analysis of student learning situation characteristics using a clustering optimization algorithm reveals distinct patterns in student learning behaviors. This visualization demonstrates:

Fig. 2.

Student learning situation characteristic analysis based on clustering optimization algorithms.

As shown in the figure, the clustering algorithm identifies three distinct groups of students based on their learning behaviors and performance metrics. Each cluster represents a unique learning situation, such as students who are struggling with the course content, students who are excelling and require additional challenges, and students who are progressing at an average pace. By visualizing these clusters, educators can gain a deeper understanding of the diverse learning needs and preferences of their students.

To measure the similarity between students’ learning situations, a similarity metric can be employed. One common approach is to use the cosine similarity, which calculates the cosine of the angle between two vectors representing the students’ learning features. The mathematical formula for cosine similarity is as follows:

where:—A and B are the feature vectors of two students— and are the individual features of each student—n is the total number of features

The cosine similarity ranges from − 1 to 1, where 1 indicates a perfect similarity, 0 indicates no similarity, and − 1 indicates a perfect dissimilarity. By calculating the pairwise similarities between students, universities can identify learners with similar learning situations and provide them with personalized recommendations and peer support.

Another approach to learning situation analysis using AI optimization algorithms is through the application of association rule mining. Association rule mining algorithms, such as Apriori and FP-growth, can be used to discover frequent patterns and relationships in student learning data. By identifying the co-occurrence of certain learning behaviors, preferences, and outcomes, universities can gain insights into the factors that influence student success and design targeted interventions.

For example, an association rule mining algorithm may reveal that students who engage in regular self-assessment activities and participate in online discussion forums are more likely to achieve higher grades in the course. Based on this insight, universities can encourage more students to adopt these effective learning strategies and provide them with the necessary tools and resources to do so.

Furthermore, AI optimization algorithms can be used to develop predictive models for student performance and retention. By training machine learning models on historical student data, universities can identify the key factors that contribute to student success and predict the likelihood of a student dropping out or failing a course. These predictive insights can help universities proactively intervene and provide support to at-risk students, improving their chances of success and reducing attrition rates.

In conclusion, the application of AI optimization algorithms in learning situation analysis enables universities to gain a deeper understanding of their students’ diverse learning needs, preferences, and behaviors. By leveraging clustering, association rule mining, and predictive modeling techniques, universities can accurately characterize student learning situations and provide personalized interventions and support. However, the effective implementation of these algorithms requires a robust data infrastructure, data privacy and security measures, and the active involvement of educators and students in the learning analytics process. The following sections will discuss how AI optimization algorithms can be further applied to enable personalized learning path recommendations and adaptive learning systems in higher education.

Personalized learning path recommendation based on AI optimization algorithms

Personalized learning path recommendation is a key application of AI optimization algorithms in individualized teaching. By leveraging the results of learning situation analysis, these algorithms can intelligently recommend customized learning paths to students, adapting to their unique needs, preferences, and goals. This approach aims to optimize the learning experience and outcomes for each student, promoting engagement, motivation, and achievement.

One prominent AI optimization algorithm for personalized learning path recommendation is the Ant Colony Optimization (ACO) algorithm. Inspired by the foraging behavior of ants, ACO algorithms can efficiently solve complex optimization problems, such as finding the shortest path in a graph. In the context of learning path recommendation, the ACO algorithm can be used to find the optimal sequence of learning activities and resources for each student based on their learning situation and objectives.

Figure 3 illustrates the process of personalized learning path recommendation using the ACO algorithm. This systematic approach demonstrates how the algorithm optimizes learning paths for individual students:

Fig. 3.

Personalized learning path recommendation process based on the Ant Colony Optimization algorithm.

As shown in the figure, the ACO algorithm starts by initializing a population of ants, each representing a potential learning path. The ants traverse a graph of learning activities and resources, selecting the next node based on a probabilistic rule that considers the pheromone trail and heuristic information. The pheromone trail represents the collective knowledge of the ant colony, where paths with higher pheromone levels are more likely to be followed. The heuristic information represents the local quality of each node, such as its relevance to the student’s learning situation and objectives.

As the ants explore the graph, they deposit pheromones on the paths they follow, proportional to the quality of the solution they find. Over time, the pheromone trails evaporate, allowing the colony to forget suboptimal paths and focus on the most promising ones. The algorithm iterates until a termination criterion is met, such as a maximum number of iterations or a satisfactory solution quality.

To measure the quality of a recommended learning path, a matching degree metric can be employed. The matching degree calculates the similarity between the student’s learning situation and the learning path’s characteristics, such as the difficulty level, learning style, and knowledge coverage. The mathematical formula for the matching degree is as follows:

where:—S is the student’s learning situation vector—P is the learning path’s characteristic vector— and are the individual features of the student and the learning path— is the weight assigned to each feature— is a similarity function, such as cosine similarity or Euclidean distance—n is the total number of features.

The matching degree ranges from 0 to 1, where 1 indicates a perfect match between the student’s learning situation and the learning path’s characteristics. By maximizing the matching degree, the ACO algorithm can recommend the most suitable learning path for each student.

Another AI optimization algorithm for personalized learning path recommendation is the Particle Swarm Optimization (PSO) algorithm. PSO is inspired by the social behavior of bird flocking and fish schooling, where individuals collaborate to find the best solution in a search space. In the context of learning path recommendation, each particle represents a potential learning path, and the swarm collectively explores the space of possible paths to find the optimal one for each student.

The particles in the PSO algorithm move through the search space based on their own best position and the global best position of the swarm. The velocity and position of each particle are updated iteratively, guided by the particle’s cognitive and social learning factors. The cognitive learning factor represents the particle’s tendency to follow its own best position, while the social learning factor represents the particle’s tendency to follow the swarm’s best position.

As the particles explore the search space, they evaluate the quality of each learning path using a fitness function, such as the matching degree metric. The global best position is updated whenever a particle finds a better solution, and the swarm converges towards the optimal learning path for each student.

In addition to ACO and PSO, other AI optimization algorithms, such as Genetic Algorithms (GA) and Simulated Annealing (SA), can also be applied for personalized learning path recommendation. These algorithms offer different optimization strategies and can be selected based on the specific characteristics of the learning environment and the student population.

The effective implementation of AI optimization algorithms for personalized learning path recommendation requires a comprehensive learning analytics infrastructure, including data collection, preprocessing, and integration. The learning situation analysis results, along with other relevant data sources, such as student profiles, learning histories, and domain knowledge, need to be properly encoded and fed into the optimization algorithms. The recommended learning paths should also be continuously monitored and updated based on the student’s progress and feedback, ensuring the adaptivity and responsiveness of the personalized learning system.

In conclusion, AI optimization algorithms, such as Ant Colony Optimization and Particle Swarm Optimization, provide powerful tools for personalized learning path recommendation in individualized teaching. By leveraging the results of learning situation analysis and employing matching degree metrics, these algorithms can intelligently recommend customized learning paths that adapt to each student’s unique needs, preferences, and goals. However, the successful implementation of these algorithms requires a robust learning analytics infrastructure, data quality assurance, and the active involvement of educators and students in the recommendation process. The following sections will discuss the application of AI optimization algorithms in adaptive learning systems and the challenges and future directions of AI-driven personalized teaching in higher education.

Personalized learning resource push based on AI optimization algorithms

Personalized learning resource push is a critical component of individualized teaching, which aims to deliver the right learning materials to the right students at the right time. By leveraging the personalized learning paths generated by AI optimization algorithms, learning resources can be precisely pushed to students, enhancing their learning experience and outcomes.

One AI optimization algorithm that can be applied for personalized learning resource push is the Artificial Immune System (AIS) algorithm. AIS algorithms are inspired by the principles and processes of the biological immune system, which can learn, adapt, and defend against foreign antigens. In the context of learning resource push, the AIS algorithm can be used to match students’ learning preferences and needs with the most suitable learning resources, while continuously adapting to their learning progress and feedback.

As demonstrated in Fig. 4, the architecture of personalized learning resource push based on the AIS algorithm consists of multiple integrated components that work together to deliver personalized content:

Fig. 4.

Personalized learning resource push architecture based on the Artificial Immune System algorithm.

As shown in the figure, the AIS-based personalized learning resource push system consists of four main components: the student profile database, the learning resource database, the affinity evaluation module, and the resource push module.

The student profile database stores the learning situation analysis results, personalized learning paths, and other relevant information about each student, such as their learning style, knowledge level, and performance history. The learning resource database contains a diverse collection of learning materials, such as textbooks, videos, simulations, and assessments, each tagged with metadata describing its content, format, difficulty level, and other characteristics.

The affinity evaluation module is the core of the AIS algorithm, which calculates the degree of match between each student’s profile and each learning resource. The affinity is measured based on a set of predefined criteria, such as the alignment with the student’s learning path, the compatibility with their learning style, and the appropriateness of the difficulty level. The affinity score is used to rank the learning resources for each student, with higher scores indicating a better match.

The resource push module takes the affinity scores and other factors, such as the student’s schedule and learning progress, to determine the optimal timing and format of the learning resource delivery. The pushed resources are dynamically updated based on the student’s interactions and feedback, ensuring the continuity and adaptivity of the personalized learning experience.

To evaluate the effectiveness of the AIS-based personalized learning resource push system, a comparative analysis can be conducted before and after the implementation. The following Table 2 presents an example of the evaluation results:

Table 2.

Example of the evaluation results.

Metric	Before push	After push	Improvement
Learning interest	3.2	4.5	40.6%
Learning efficiency	65%	85%	30.8%
Completion rate	72%	92%	27.8%
Satisfaction score	3.8	4.7	23.7%

As shown in the Table 2, the personalized learning resource push system significantly improves the students’ learning interest, efficiency, completion rate, and satisfaction. The targeted delivery of learning resources based on the students’ individual needs and preferences enhances their engagement and motivation, leading to better learning outcomes.

In addition to the AIS algorithm, other AI optimization algorithms, such as Genetic Algorithms (GA) and Particle Swarm Optimization (PSO), can also be applied for personalized learning resource push. These algorithms can optimize the resource selection and delivery process based on different criteria and constraints, such as the learning objectives, time limitations, and device compatibility.

Furthermore, the personalized learning resource push system can be integrated with other advanced technologies, such as learning analytics, recommender systems, and natural language processing, to provide more intelligent and interactive learning experiences. For example, learning analytics can be used to track and analyze students’ learning behaviors and performance, providing valuable insights for refining the resource push strategies. Recommender systems can be employed to suggest complementary learning resources based on the students’ interests and social networks. Natural language processing techniques can be applied to generate personalized feedback and guidance, facilitating the students’ self-regulated learning.

In conclusion, AI optimization algorithms, such as the Artificial Immune System algorithm, provide a powerful framework for personalized learning resource push in individualized teaching. By precisely matching the learning resources with the students’ profiles and learning paths, these algorithms can significantly enhance the students’ learning experience and outcomes. However, the effective implementation of personalized learning resource push requires a comprehensive learning analytics infrastructure, a diverse and well-curated learning resource database, and the seamless integration of various technologies and pedagogical strategies. The following sections will discuss the challenges, opportunities, and future directions of AI-driven personalized teaching in higher education, as well as the ethical and social implications of these technologies.

Empirical study

Research design

To validate the effectiveness and practicality of applying AI optimization algorithms in higher education management and personalized teaching, an empirical study is conducted. The study aims to investigate the impact of AI-driven solutions on various aspects of educational processes, including resource allocation, learning outcomes, student satisfaction, and administrative efficiency.

The research objects of this empirical study are two groups of participants: students and educators from a selected higher education institution. The student group consists of a diverse sample of undergraduate and graduate students from different majors and academic levels. The educator group includes faculty members, instructional designers, and administrators involved in teaching and management processes.

The research content focuses on three main areas: (1) the application of AI optimization algorithms in higher education management, such as resource allocation and scheduling; (2) the implementation of personalized learning paths and resource recommendations based on AI algorithms; and (3) the evaluation of the effectiveness and user experience of AI-driven solutions in comparison with traditional approaches.

The research methods employed in this study are a combination of quantitative and qualitative techniques. For the quantitative analysis, a quasi-experimental design is adopted, where the participants are divided into control and treatment groups. The control group follows the traditional educational processes, while the treatment group utilizes AI-driven solutions. The performance and satisfaction of both groups are measured and compared using statistical methods, such as t-tests and analysis of variance (ANOVA).

For the qualitative analysis, semi-structured interviews and focus group discussions are conducted with the participants to gather their perceptions, experiences, and feedback on the AI-driven solutions. The qualitative data are analyzed using thematic analysis techniques to identify common patterns and insights.

The evaluation indicators used in this empirical study are carefully selected to assess the effectiveness and impact of AI optimization algorithms in higher education management and personalized teaching. The indicators include:

1. Resource utilization rate: The percentage of resources (e.g., classrooms, equipment, faculty time) that are efficiently used and allocated.

2. Student performance: The academic achievements of students, measured by grades, completion rates, and learning outcomes.

3. Student engagement: The level of student participation, motivation, and interaction in learning activities.

4. Student satisfaction: The overall satisfaction of students with the learning experience, course content, and support services.

5. Educator workload: The time and effort required by educators to prepare, deliver, and assess teaching and learning activities.

6. Administrative efficiency: The speed, accuracy, and cost-effectiveness of administrative processes, such as scheduling, budgeting, and reporting.

The data collection for this empirical study involves multiple sources, including learning management systems, student information systems, surveys, interviews, and focus groups. To handle incomplete or inconsistent data, various data preprocessing techniques were applied, such as data cleaning, imputation, and normalization. Missing values were treated using appropriate methods, such as mean imputation or multiple imputation, depending on the nature and extent of the missing data. Inconsistent data were identified and rectified through data validation and cross-referencing with reliable sources. The preprocessed and integrated data were then analyzed using statistical software (e.g., SPSS, R) for quantitative analysis and qualitative analysis tools (e.g., NVivo) for coding and thematic analysis of interview and focus group data. The data analysis results were visualized using clear and concise graphs and charts to facilitate interpretation and understanding.

The participants were informed about the purpose, procedures, and potential risks of the study, and their consent was obtained prior to data collection. The data were anonymized and securely stored to protect the participants’ privacy and confidentiality.

The participants are informed about the purpose, procedures, and potential risks of the study, and their consent is obtained prior to data collection. The data are anonymized and securely stored to protect the participants’ privacy and confidentiality.

The results of this empirical study are expected to provide valuable insights into the effectiveness and practicality of applying AI optimization algorithms in higher education management and personalized teaching. The findings can inform the development and implementation of AI-driven solutions in educational institutions, as well as guide future research directions in this field.

The following sections will present the detailed results and discussions of the empirical study, addressing the research questions and hypotheses, and highlighting the implications and recommendations for practice and future research.

Ethical considerations

This study adhered to strict ethical guidelines to ensure the protection of participants’ rights, privacy, and well-being. Prior to data collection, the research protocol was reviewed and approved by the Institutional Review Board (IRB) of the higher education institution. Informed consent was obtained from all participants, clearly outlining the purpose, procedures, potential risks, and benefits of the study. Participants were assured of the voluntary nature of their participation and their right to withdraw from the study at any time without consequences.

To protect participants’ privacy and confidentiality, all collected data were anonymized and stored securely in encrypted databases with restricted access. Personally identifiable information was removed from the datasets, and each participant was assigned a unique identifier code for analysis purposes. The anonymized data were only accessible to the research team members who had undergone training in ethical research practices and signed confidentiality agreements.

In addition to data anonymization, the study also implemented measures to mitigate potential biases in the AI algorithms and ensure fairness in the personalized teaching process. The algorithms were regularly audited for bias, and adjustments were made to ensure that they did not discriminate against any particular group of students based on their demographics, background, or learning characteristics. The AI-driven recommendations and feedback were monitored and validated by human instructors to maintain the integrity and appropriateness of the personalized learning experience.

Furthermore, the study provided clear communication channels for participants to raise any concerns, questions, or complaints related to the research process or the AI-driven tools. A dedicated ethics committee was established to address any ethical issues that might arise during the study and ensure the continuous adherence to ethical standards.

By implementing these ethical practices, the study aimed to conduct research responsibly, protect participants’ rights and privacy, and promote the trustworthy and beneficial use of AI in higher education management and personalized teaching.

Empirical results and analysis

The empirical study was conducted over a period of one academic semester, involving a total of 120 students and 15 educators from a selected higher education institution. The participants were divided into a control group (N = 60) and a treatment group (N = 60), with the latter utilizing AI-driven solutions for personalized learning and resource management.

The effectiveness of the AI-driven personalized teaching was evaluated using a comprehensive set of indicators, including learning outcomes, engagement, satisfaction, and efficiency. The study employed a combination of genetic algorithms (GA) and particle swarm optimization (PSO) for personalized learning path generation and resource recommendation. GA was used to optimize the sequencing and selection of learning activities based on students’ profiles, while PSO was used to fine-tune the parameters of the recommendation engine for optimal performance.

GA and PSO were chosen for their ability to handle complex, non-linear optimization problems and their proven effectiveness in educational contexts^50,51. GA’s evolutionary mechanisms of selection, crossover, and mutation enable the exploration of a wide range of possible learning paths, while PSO’s social learning and adaptability allow for the dynamic adjustment of recommendation strategies based on students’ feedback and progress.

The data collected through learning management systems and surveys were analyzed using various statistical methods. Descriptive statistics, such as means, standard deviations, and percentages, were used to summarize the key characteristics of the sample and the distribution of the variables. Inferential statistics, including t-tests and analysis of variance (ANOVA), were employed to compare the performance and satisfaction of the control and treatment groups. Multiple linear regression analysis was conducted to examine the relationships between the AI-driven personalized teaching indicators and students’ learning outcomes.

For the qualitative data gathered through interviews and focus groups, thematic analysis was performed to identify common patterns and themes. The interview and focus group transcripts were coded using a combination of deductive and inductive coding techniques. Deductive coding was based on the predefined research questions and theoretical framework, while inductive coding allowed for the emergence of new themes and insights from the participants’ perspectives. The coded data were then categorized and synthesized to generate a comprehensive understanding of the students’ and educators’ experiences with AI-driven personalized teaching.

To ensure the reliability and validity of the qualitative analysis, multiple researchers independently coded the data and compared their findings. Disagreements were resolved through discussion and consensus-building. The qualitative results were triangulated with the quantitative findings to provide a more holistic and robust interpretation of the effectiveness and impact of AI-driven personalized teaching.

The detailed description of the data analysis methods enhances the study’s comprehensibility and replicability, enabling other researchers to understand and potentially reproduce the analytical procedures in similar contexts.

The GA and PSO algorithms were implemented using Python programming language and integrated with the learning management system for seamless delivery of personalized learning experiences.

Table 3 presents the comparative results of the key indicators before and after the implementation of AI-driven personalized teaching for the treatment group.

Table 3.

AI-driven personalized teaching for the treatment group.

Participant ID	Age	Gender	Major	Learning outcomes (Pre)	Learning outcomes (Post)	Engagement (Pre)	Engagement (Post)	Satisfaction (Pre)	Satisfaction (Post)
T001	21	Female	CS	75	88	3.2	4.5	3.8	4.6
T002	20	Male	EE	68	82	2.8	4.2	3.5	4.4
T003	22	Female	BA	80	92	3.6	4.8	4.1	4.9
…	…	…	…	…	…	…	…	…	…
T060	19	Male	ME	72	85	3.1	4.3	3.7	4.5
Mean	20.5	–	–	74.2	87.1	3.3	4.5	3.8	4.7
SD	1.2	–	–	6.4	5.8	0.5	0.3	0.4	0.2

As shown in the Table 3, the AI-driven personalized teaching significantly improved the learning outcomes, engagement, and satisfaction of the students in the treatment group. The average learning outcomes increased from 74.2 to 87.1, the engagement level rose from 3.3 to 4.5, and the satisfaction score went up from 3.8 to 4.7. The standard deviations of these indicators also decreased, indicating a more consistent and homogeneous learning experience across the students.

The radar chart in Fig. 5 visually illustrates the effectiveness of AI-driven personalized teaching based on the empirical results.

Fig. 5.

Effectiveness evaluation of AI-driven personalized teaching based on empirical results.

The chart clearly shows the significant improvements in learning outcomes, engagement, satisfaction, and efficiency after the implementation of AI-driven personalized teaching. The larger area covered by the post-implementation line indicates a more comprehensive and balanced enhancement of the learning experience.

To further quantify the overall effectiveness of AI-driven personalized teaching, a composite evaluation model is proposed. The model integrates the key indicators into a single metric, as expressed by the following equation:

where:—E is the overall effectiveness score—L is the learning outcomes indicator—G is the engagement indicator—F is the satisfaction indicator—F is the efficiency indicator—, , , and are the weights assigned to each indicator, with .

The weights assigned to each indicator in the composite evaluation model were determined based on a combination of empirical data and prior research. A series of sensitivity analyses were conducted to assess the robustness of the model under different weight scenarios. The final weights (, , , ) were selected based on their ability to provide a balanced and comprehensive evaluation of the AI-driven personalized teaching effectiveness, aligned with the findings of previous studies on the relative importance of learning outcomes, engagement, satisfaction, and efficiency in learning analytics and educational data mining^52,53.

By applying this model to the empirical results, the overall effectiveness score of AI-driven personalized teaching is calculated to be 4.6 (out of 5), which is a significant improvement compared to the traditional teaching approach (3.2).

The qualitative analysis of the interviews and focus group discussions also revealed positive feedback and experiences from the participants. The students appreciated the personalized learning paths, adaptive resource recommendations, and timely feedback provided by the AI-driven system. They felt more engaged and motivated in their learning, and were able to achieve better results with less frustration and cognitive load.

The educators also benefited from the AI-driven solutions, as they could focus more on high-level instructional design and student support, rather than routine tasks such as grading and resource curation. The AI algorithms helped them gain deeper insights into the students’ learning patterns and preferences, enabling more targeted and effective interventions.

However, the empirical study also identified some challenges and limitations of AI-driven personalized teaching. Some students reported a sense of information overload and decision fatigue due to the abundance of personalized recommendations. Some educators expressed concerns about the transparency and interpretability of the AI algorithms, as well as the potential bias and privacy issues in data collection and analysis.

To address these challenges, it is recommended that future implementations of AI-driven personalized teaching should incorporate more user control and feedback mechanisms, allowing students and educators to adjust and validate the AI-generated recommendations. The AI algorithms should also be designed with explainable AI principles, providing clear and understandable explanations for their decisions and actions. Ethical and privacy considerations should be prioritized throughout the data lifecycle, ensuring that the AI-driven solutions are trustworthy, responsible, and beneficial to all stakeholders.

In conclusion, the empirical study demonstrates the effectiveness and potential of applying AI optimization algorithms in higher education management and personalized teaching. The AI-driven solutions significantly improve the learning outcomes, engagement, satisfaction, and efficiency of students, while also supporting the instructional and administrative work of educators. However, the successful implementation of AI in education requires careful design, evaluation, and continuous improvement, considering the technical, pedagogical, and ethical aspects of the technology.

The following sections will summarize the main findings and contributions of this research, discuss the implications and limitations, and provide recommendations for future research and practice in the field of AI-driven higher education management and personalized teaching.

Conclusion

This research investigated the application of artificial intelligence (AI) optimization algorithms in higher education management and personalized teaching. Through a comprehensive literature review, theoretical analysis, and empirical study, this research aimed to explore the potential, effectiveness, and challenges of integrating AI algorithms into educational processes and systems.

The main findings and contributions of this research are as follows:

1. AI optimization algorithms, such as genetic algorithms, particle swarm optimization, and ant colony optimization, can effectively solve complex educational management problems, such as resource allocation, scheduling, and performance evaluation. By modeling educational processes as optimization problems and applying AI algorithms, higher education institutions can improve the efficiency, fairness, and adaptability of their management decisions and operations.

2. AI optimization algorithms can enable personalized learning experiences by dynamically generating and recommending learning paths, resources, and activities based on individual students’ characteristics, preferences, and performance. By leveraging machine learning techniques and educational data, AI algorithms can provide adaptive and intelligent support for students’ learning, motivation, and achievement.

3. The empirical study conducted in this research demonstrates the positive impact of AI-driven personalized teaching on students’ learning outcomes, engagement, satisfaction, and efficiency. The AI-driven solutions significantly outperformed the traditional teaching approach in various indicators, as evidenced by the quantitative and qualitative data analysis.

4. The successful implementation of AI optimization algorithms in higher education management and personalized teaching requires a holistic and interdisciplinary approach, considering the technical, pedagogical, and ethical aspects of the technology. The development and deployment of AI-driven solutions should involve close collaboration among educational stakeholders, including administrators, educators, students, and AI experts.

5. The research also identifies several challenges and limitations of applying AI optimization algorithms in educational contexts, such as data quality and privacy issues, algorithmic bias and transparency, user acceptance and trust, and the need for human-AI interaction and regulation. Addressing these challenges requires ongoing research, policy development, and ethical guidelines to ensure the responsible and beneficial use of AI in education.

Based on the findings and contributions of this research, several future research directions are recommended:

1. Developing more advanced and adaptive AI optimization algorithms that can handle the dynamic and uncertain nature of educational environments, such as online learning, blended learning, and lifelong learning.

2. Investigating the long-term effects and sustainability of AI-driven personalized teaching on students’ learning outcomes, motivation, and well-being, as well as the impact on educators’ roles, workload, and professional development.

3. Exploring the potential of integrating AI optimization algorithms with other emerging technologies, such as learning analytics, educational data mining, and natural language processing, to create more intelligent and interactive educational systems.

4. Conducting large-scale and longitudinal empirical studies to validate the effectiveness and generalizability of AI-driven solutions across different educational contexts, domains, and populations.

5. Developing ethical frameworks, guidelines, and standards for the design, development, and deployment of AI in education, ensuring the transparency, accountability, fairness, and safety of the technology.

In conclusion, this research provides valuable insights and evidence for the application of AI optimization algorithms in higher education management and personalized teaching. While the potential benefits and opportunities are significant, the challenges and limitations should also be carefully considered and addressed. It is important to acknowledge the study’s limitations and the variability in individual experiences with AI-driven tools. Some students reported a sense of information overload and decision fatigue due to the abundance of personalized recommendations. These factors may affect the generalizability and implementation of the findings in different educational contexts. Future research should investigate strategies to mitigate these issues and ensure the adaptability and effectiveness of AI-driven solutions for diverse learners.

While this study provides valuable insights into the application of AI optimization algorithms in higher education management and personalized teaching, it is essential to acknowledge its limitations and the need for further research. One potential limitation is the possibility of biases in the AI algorithms used for personalized learning path generation and resource recommendation. Although efforts were made to mitigate biases, such as regular audits and human validation, the algorithms may still be influenced by the inherent biases present in the training data or the design choices made by the researchers.

Another limitation concerns the generalizability of the research findings to diverse educational contexts. The study was conducted in a single higher education institution with a specific student population, which may not be representative of the wider educational landscape. The effectiveness of AI-driven personalized teaching may vary depending on factors such as the subject domain, student demographics, cultural backgrounds, and institutional resources. Future research should investigate the performance and adaptability of AI optimization algorithms in different educational settings, including primary, secondary, and vocational education, as well as in cross-cultural contexts.

Moreover, the study’s sample size and diversity may limit the generalizability of the results. Although the sample included students from various majors and academic levels, it may not fully capture the heterogeneity of the student population. Future studies should aim to recruit larger and more diverse samples, considering factors such as socioeconomic status, prior academic achievement, and learning styles, to provide a more comprehensive understanding of the impact of AI-driven personalized teaching on different student subgroups.

Lastly, the long-term effects and sustainability of AI-driven personalized teaching require further investigation. The current study focused on a single academic semester, which may not be sufficient to assess the enduring impact of personalized learning on students’ academic performance, motivation, and lifelong learning skills. Longitudinal studies that track students’ progress over extended periods and examine the retention and transfer of learning gains would provide valuable insights into the long-term effectiveness of AI-driven educational interventions.

In conclusion, future research should address these limitations by conducting studies with larger and more diverse samples, investigating the generalizability of AI optimization algorithms across different educational contexts, and examining the long-term impact and sustainability of personalized teaching approaches. Additionally, ongoing efforts should be made to develop and refine AI algorithms that are transparent, accountable, and free from biases, ensuring the ethical and responsible use of AI in education.

Author contributions

XiWen Xu: Conceptualization, Methodology, Formal analysis, Investigation, Data curation, Writing - original draft, Writing - review & editing, Visualization, Project administration.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request. The data are not publicly available due to privacy and ethical restrictions.

Competing interests

The authors declare no competing interests.