Research on Optimizing Teaching Resource Allocation Strategies with Machine Learning Models for Intelligent English Teaching Systems

Artificial intelligence is accelerating its integration into education while bringing great impact on people’s life and work. The new era of subversive intelligent technology changing education is coming, and the new round of technology changing education will be led by intelligent education, injecting new ideas into education and teaching, providing new methods and tools, driving the fundamental transformation of education and teaching mode, and promoting the qualitative enhancement of teaching effect [1-4]. Taking English education as an example, the new generation of learners is shifting from passive acceptance of learning to active discovery and inquiry, from large-class teaching to personalized teaching, from consumers of knowledge to creators of knowledge, and the demand for intelligent learning environment, ubiquitous learning resources, and personalized teaching is getting higher and higher [5-7]. English education administrators are able to support the development of personalized learning services based on the ability to extract key concepts in learning resources and access online learning behavioral processes through Artificial Intelligence systems to establish knowledge models, on the basis of which they can intelligently analyze learners’ individual cognitive status and learning needs [8-11]. Artificial intelligence systems can also assess and analyze specific groups, analyze and understand the learning characteristics of the class as a whole, discover the weak links in the learning effect, and provide effective guidance for English education administrators to allocate teaching resources and rationally carry out teaching management [12-15]. At present, artificial intelligence has been able to enhance teaching and learning efficiency in all aspects of learning coaching, teaching assessment and teaching space optimization, enhance the learning experience of English teaching, and make personalized learning a reality [16-18]. Artificial intelligence is leading the innovation of education and teaching and becoming an important factor in the development of education informatization.

Literature [19] proposes an artificial intelligence teaching resources integration model, which utilizes web crawler recognition technology to establish an index database of English teaching resources, and utilizes L-NCD technology as well as ISPO algorithm and K-means algorithm to redundantly remove and integrate the output of English teaching data, respectively. Literature [20] studied the sharing method of English teaching resources in the digital era, in addition to the management of digital teaching resources, it is also necessary to establish a resource evaluation index system and implement objective evaluation, in order to realize the effective sharing of massive teaching resources in a complex environment and improve the quality of English teaching. Literature [21] designs an artificial intelligence information management system based on resource storage system and hardware layer, and simulation experiments found that the system effectively realizes the contribution and management of English teaching resources among schools. Literature [22] develops an intelligent education management platform with cloud computing as the technical theory and service-oriented infrastructure, the platform is deployed in a server cluster environment with hadpoo hosted storage clusters as the data storage center, which provides some insights into the development of intelligent access and intelligent collaborative management in education. Literature [23] shows that open teaching resources in online learning can be used as auxiliary resources in the process of teaching and learning, and constructs a deep learning-based online learning platform (DLOL), which combines fuzzy convolutional neural networks and decision tree algorithms to realize the scheduling of teaching resources, and provides help to improve students’ learning ability. Literature [24] describes the positive impact of intelligent teaching system for students’ personalized learning, which extends the teaching functions of training institutions and traditional education systems, and provides students with better scheduling through the management and allocation of educational resources, which helps students’ personalized learning.

In this paper, for network teaching resource allocation, network resource allocation request is proposed to obtain the sorting and optimization parameters of resources. In the allocation process, the resource allocation delay model is formed to obtain user delays in different modes. Considering from the perspective of a pair of primary users and a pair of secondary users, the Poisson distribution obeys the parameters and calculates the equivalent packet transmission rate of the primary users after the allocation to construct the resource allocation delay model. The resource scheduling problem of network teaching resource allocation is transformed into a nonlinear optimization problem, and the greedy algorithm is integrated to solve the problem in a globally optimal way, and the results of teaching resource allocation are obtained. Based on the mathematical model of an automatic class scheduling system, backtracking and greedy algorithms are jointly used to improve the model. The improved English teaching resource allocation optimization model is applied to Y university, and a test environment is constructed to evaluate the results and application effects of the English teaching resource allocation optimization.

2

Optimizing the allocation of teaching resources

2.1

Principles of Network Teaching Resource Allocation

In the process of carrying out the allocation of network resources, a request for the allocation of network resources is first made, and the search engine searches the Internet and the network resource information base, obtains the sorting of the resources and their optimization parameters, obtains a list of high-quality network resources, and prompts the optimized network resource information to be submitted to the allocation of the user’s choices in the form of a pagination. The specific steps are described in detail below:

Assuming, by A representing the content of n data chunks of the original resource, C_ij representing a linear combination of the local vectors of each resource, and F representing the proposed request for resource allocation, Eq. (1) is utilized to obtain the network resource ordering and its optimization parameters: (1) $S D (i, j) = \frac{F (d x_{k})}{C_{i j} \cdot A \cdot n} R_{M} \cdot P_{S}$

Where, dx_k represents the channel gain, R_M represents the resource requesting node, and P_S represents the unit power, the condition of j = 1, 2, 3… is satisfied.

Assuming, by θ represents the list of network resources, Γ represents the data requesting node, Ψ represents the set of processors for the resources, and ι represents the resource service capacity, the average delay for user resource allocation is calculated using equation (2): (2) $\frac{d_{s}}{d h} = \frac{ψ \cdot ι}{θ \cdot Γ} \otimes (δ (r, j) \cdot \frac{l_{2}}{r \cdot ℘})$

Where, l₂ represents the average service time of network users, r represents the maximum total transmission power of terminals, and ℘ represents the current working state of network teaching resources.

Assuming, by x_i represents the allocation cost function of a single resource, and h(τ, κ) represents the total allocation cost of all the resources of the network, the list of network resources is optimized using Eq. (3) to obtain a list of high-quality network resources: (3) $σ (o, p) = \frac{x_{i}}{h (τ, κ)} S D (i, j) \cdot \frac{d_{s}}{d h} ℓ (l . k)$

Where, ℓ(l.k) represents the order in which the network resources are arranged.

Assuming, by kl represents the user node network environment, the allocation of network resources can be accomplished by utilizing Eq. (4): (4) $c k (j) = \frac{σ (o, p) \cdot k l}{℘ \oplus f f (y, k)}$

where ℘ represents the idle downlink power resource, ff represents the channel gain, and (y, k) represents the Gaussian white noise power.

2.2

Multi-Rate Cognition-Based Online Teaching Resource Allocation

2.2.1

Calculation of user delay in different modes

In the process of allocating network teaching resources, the resource allocation delay model is formed from the perspective of the existence of a pair of primary users and a pair of secondary users in the network, and the user delay in different modes is obtained. The specific steps are described as follows:

Considering from the perspective of the existence of a pair of primary users and a pair of secondary users in the network, it is assumed that the Poisson distributions of the resources of the primary and secondary users are obeyed by the parameters represented by λ_p and λ_s, and L_P and L_s represent the corresponding packet lengths, respectively, and pU_s represents the transmitting node of the primary user, pU_D represents the receiving node of the primary user, and R_P and R_PS represent the channel rates of the transmitting resource node SU_S of the secondary user, and R_SP and R_s represent the channel rates of the secondary user, respectively. sending resource node SU_s to the primary user receiving node pU_D and secondary user receiving resource node SU_D, respectively. If the secondary user is involved in the resource allocation process to the primary user, if the primary user resource node grouping is allocated with the secondary user, the post-allocation primary user equivalent packet transmission rate represented by R_CP is calculated using equation (5): (5) $R_{C P} = \frac{p U_{S} \oplus p U_{D}}{S U_{S} [R_{S} + S U_{D}]} * R_{S P}$

Based on the summarized exposition, the resource allocation delay model is formed using equation (6): (6) $q (R_{C P}) = \frac{S U_{S} * R_{P S}}{R_{S P}} S U_{D}$

In the resource use in non-allocated mode, the primary user has the priority right to use the resource, and the secondary user can use the resource only when the primary user is no longer using the resource, and when the primary user needs to use the resource, the secondary user should back off in time and give the right to use the resource to the primary user, then the average delay of resource use of the primary user is formed using equation (7): (7) $E ‖ T_{N C - P} ‖ = E [X_{N C - P}] + \frac{λ_{P} E [X_{N C - P}^{2}]}{2 (1 - ρ_{N C - P})}$

Where, $E [X_{N C - P}]$ represents the average service time of the primary user in the non-allocated model, λ_P represents the resource utilization of the primary user, and ρ_NC−P represents the ratio of the resource utilization of the primary user and the departure rate of the primary user in the non-allocated model [25].

Then the average delay of secondary users is obtained using equation (8): (8) $E ‖ T_{N C - s} ‖ = \frac{E [X_{N C - s}]}{(1 - ρ_{N C - P})} + \frac{λ_{s} E [X_{N C - s}^{2}]}{2 (1 - ρ_{N C - P}) (1 - ρ_{N c - p} - ρ_{N c - S})}$

where $E [X_{N C - s}]$ represents the average service time of the secondary user, ρ_Nc−s represents the ratio of the resource utilization rate of the secondary user and the departure rate of the primary user under the non-distributive model. λ_s represents the resource utilization rate of the user.

In the process of resource use in the allocation model, the secondary user provides resource service to the primary user until the queue of the primary user is empty, and because of the existence of the resource allocation agreement between the primary and secondary users, the primary user will not be in a position to forcibly interrupt the resources that are being used by the secondary user, then the average delay of the primary user’s resource use in the model can be obtained by using Eq. (9): (9) $E ‖ T_{B E - P} ‖ = \frac{E ‖ T_{N C - P} ‖ \times E ‖ T_{N C - s} ‖}{E ‖ X_{B E - P} ‖ \otimes E ‖ W_{B E - P} ‖}$

Where, $E ‖ X_{B E - P} ‖$ represents the average service time of the primary user and $E ‖ W_{B E - P} ‖$ represents the average waiting time of the primary user.

The average delay of the secondary user is obtained using equation (10): (10) $E ‖ T_{B E - S} ‖ = \frac{E [X_{B E - s}] \oplus E [W_{B E - S}]}{E ‖ T_{B E - P} ‖}$

where $E [X_{B E - S}]$ represents the average service time for sub-users and $E [W_{B E - S}]$ represents the average waiting time for sub-users.

2.2.2

Web-based Teaching Resource Allocation Based on Greedy Algorithm

In the process of optimizing the allocation of network teaching resources, the transmission rate on the allocated resource blocks is dynamically adjusted to model the resource scheduling problem of network teaching resource allocation as a nonlinear optimization problem [26]. The specific steps are detailed below:

Assuming, by M representing the number of resources shared in the uplink band, C = {1, 2, ….M} representing the set of its resources, and D representing the set of resource devices, the received minimum target signal-to-noise ratios at the resource receiving end and at the base station eNB are obtained by utilizing Eqs. (11) and (12) on the basis of the obtained $E ‖ T_{B E - P} ‖$ and $E ‖ T_{B E - S} ‖$ : (11) $γ_{d}^{u L} = \frac{p_{d} G_{d d}}{N_{o} + \sum_{c} y_{c}^{d} p_{c} G_{c d}} E ‖ T_{B E - S} ‖$ (12) $γ_{e N B}^{c} = \frac{p_{c} G_{C B}}{N_{o} + \sum_{c} y_{c}^{d} p_{c} G_{d B}} E ‖ T_{B E - P} ‖$

Where p_c and p_d represent the information gain between all resource devices in the network, G_CB and G_dd represent the transmit power of resource devices, $y_{c}^{d}$ represents the noise power, and N_o represents the transmission power and the spatial distance between terminals.

Eqs. (11) and (12) are defined as the constraints to form the resource scheduling model for network teaching resource allocation, and Eq. (13) is used to model the resource scheduling problem for network teaching resource allocation as a nonlinear optimization problem: (13) $\max \sum_{c = 1}^{M} \log_{2} = \frac{γ_{d}^{u L} + γ_{e N B}^{c}}{y_{c}^{d} (1 + γ_{d}^{U L})}$

Where, $γ_{d}^{U L}$ represents the tolerable signal-to-noise ratio loss that can be tolerated by the system.

Assuming, by p_max represents the transmission constraints of the resource device, the transmission rate on the allocated resource block is dynamically adjusted using equation (14) [27]: (14) $W | ω (n, k) | = \frac{P_{\max}}{\max \sum_{c = 1}^{M} \log_{2}}$

The problem is solved globally optimally using Eq. (15) fused to a greedy algorithm: (15) $p_{d (o p)} = \frac{W | ω (n, k) | \otimes y_{d}^{h i g h}}{p_{\max}}$

where $y_{d}^{h i g h}$ represents the maximum total transmission power of the terminal.

3

English Teaching Scheduling System Based on Greedy Algorithm

3.1

Mathematical model construction of automatic class scheduling system

The elements of the scheduling problem are mainly lesson plans, hardware teaching resources (classrooms), classes, and teachers, and the relationships and constraints between the elements need to be taken into account. The main constraints that need to be considered when designing an automatic scheduling system are basic hard constraints and soft constraints. The basic hard constraints are the value of the automatic scheduling must be achieved, can not violate the logical conditions, generally refers to the teacher, classroom and students in time and space can be achieved in the scheduling of classes must be taken into account, each school’s hardware conditions are not the same, in the design of the design needs to be taken into account in the actual situation of each college and university hard constraints of the design. Soft constraints are the conditions that can be met by automating scheduling as much as possible.

The hard constraints of the design are mainly as follows: 1)

In the scheduling operation of executing courses for a certain class, what needs to be considered is that it is not possible to set the system to let the students complete two kinds of courses in a certain period of time.

2)

In the scheduling operation of a class, it is necessary to consider that the teacher cannot be set to complete two courses in a certain period of time in the system.

3)

In the scheduling operation for a class, it is necessary to consider that it is not possible to set up the system so that the classroom is used for both courses in a certain period of time.

The soft constraints of the design are as follows: 1)

Priority is given to specialized courses

2)

The number of seats in the classroom should be more than the number of students who will attend the class.

3)

In the automatic scheduling of classes in the scheduling to go to the mathematical combination of definitions are:

Teachers assemble for: $T = {T_{1}, T_{2}, \dots T_{t}}$

Classes are assembled as: $C = {C_{1}, C_{2}, \dots C_{c}}$

Curriculum Design Collection for: $L = {L_{1}, L_{2}, \dots L_{l}}$

Classroom assembly for: $R = {L_{1}, L_{2}, \dots L_{l}}$

If’s at the time of scheduling, a teacher T_j inside the college system has a lesson plan L_i, so that a_ij = 1, if no plan then a_ij = 0. In the course scheduling plan L_i includes giving a lesson to class C_i to get b_ij = 1, if the plan does not include class C_i, then b_ij = 0, so that the two correlation matrices can be obtained as: (16) $A = [\begin{matrix} a_{11} & a_{12} & \dots & a_{1 t} \\ a_{21} & a_{22} & \dots & a_{2 t} \\ \dots & \dots & \dots & \dots \\ a_{l 1} & a_{l 2} & \dots & a_{l t} \end{matrix}]$ (17) $B = [\begin{matrix} b_{11} & b_{12} & \dots & b_{1 t} \\ a_{21} & a_{22} & \dots & a_{2 t} \\ \dots & \dots & \dots & \dots \\ a_{c 1} & a_{c 2} & \dots & a_{c t} \end{matrix}]$

Matrix A represents the teacher’s schedule, and Matrix B represents the class cohort and class scheduling situation, and the relationship between the scheduling related elements can be determined from Matrix A and Matrix B. If it is assumed that the planned number of days of scheduling is d and the daily schedule of classes is planned to be s, a matrix representation of the class schedule can be made: (18) $Δ = [\begin{matrix} x_{11} & x_{12} & \dots & x_{1 d} \\ x_{21} & x_{22} & \dots & x_{2 d} \\ \dots & \dots & \dots & \dots \\ x_{s 1} & x_{x 2} & \dots & x_{x d} \end{matrix}]$

In equation (18), x_ij = 0 means that in the ird period of the jnd week is no class schedule, otherwise there is a class plan. According to the scheduling plan of Matrix A is carried out, according to the scheduling of teachers, classes and classrooms in the time period of the class schedule matrix, you can get the scheduling matrix of the classroom, the scheduling matrix between the teacher and the class, and the teacher’s teaching scheduling matrix, respectively: 1)

Classroom’s scheduling arrangement matrix (19) $\nabla_{r} = [\begin{matrix} [\begin{matrix} x_{111} & x_{112} & \dots & x_{11 d} \\ x_{121} & x_{122} & \dots & x_{12 d} \\ \dots & \dots & \dots & \dots \\ x_{1 s 1} & x_{1 s 2} & \dots & x_{1 s d} \end{matrix}] \\ Δ_{2} \\ \dots \\ Δ_{r} \end{matrix}]$

2)

Scheduling matrix between teachers and classes (20) $\nabla_{t c} = [\begin{matrix} [\begin{matrix} x_{1111} & x_{1112} & \dots & x_{111 d} \\ x_{1121} & x_{1122} & \dots & x_{112 d} \\ \dots & \dots & \dots & \dots \\ x_{11 s 1} & x_{11 s 2} & \dots & x_{11 s d} \end{matrix}] Δ_{12} \dots Δ_{1 t} \\ Δ_{21} Δ_{22} \dots Δ_{2 t} \\ \dots \dots \dots \\ Δ_{c 1} Δ_{c 2} \dots Δ_{c t} \end{matrix}]$

3)

Teacher’s Teaching Curriculum Matrix (21) $\nabla_{t l} = [\begin{matrix} [\begin{matrix} x_{1111} & x_{1112} & \dots & x_{111 d} \\ x_{1121} & x_{1122} & \dots & x_{112 d} \\ \dots & \dots & \dots & \dots \\ x_{11 s 1} & x_{11 s 2} & \dots & x_{11 s d} \end{matrix}] Δ_{12} \dots Δ_{1 t} \\ Δ_{21} Δ_{22} \dots Δ_{2 t} \\ \dots \dots \dots \\ Δ_{i 1} Δ_{12} \dots Δ_{l t} \end{matrix}]$

Based on the above analysis, in the constraint rule of scheduling, at most one class can be taught in the same classroom in the same time period, which can be obtained from the scheduling matrix of the classroom: (22) $x_{i j i j} \leq 1$

The maximum number of classes that can be scheduled in a class at any one time is obtained from the scheduling matrix between the teacher and the class: (23) $\sum_{j = 1}^{t} x_{i j i j} \leq 1$

A maximum of one class can be scheduled by one teacher at any one time, as obtained from the teacher’s teaching schedule matrix: (24) $\sum_{i = 1}^{t} x_{i j i j} \leq 1$

3.2

Design of automatic class scheduling algorithm based on backtracking and greedy improvement

1)

Greedy algorithm in improved automatic scheduling algorithm design ideas

The greedy algorithm focuses on the work objective of accomplishing the task of specifying the local optimal solution at each stage, rather than reaching the final optimal solution [28]. It enables hierarchical processing in the form of approximate solutions, which can achieve fast solutions while utilizing less time and achieving high efficiency.

For the scheduling problem of teachers, classes and classrooms, the specific implementation is as follows: different courses contain objectives and tasks with their own characteristics, which means that there are n tasks to be performed in n course, and n tasks to be accomplished under the joint collaboration of teachers, classes and classrooms, and the starting time of the tasks to be accomplished is set, and the overlapping of tasks is analyzed with feasibility as the allocation criterion, and the maximum limit of tasks to be handled by a teacher, class or classroom at a certain point of time is only one. The overlapping of tasks is analyzed, and feasibility is used as a criterion for assigning the maximum number of tasks that can be handled by a teacher, class, or room at a given point in time to only one.

2)

Design Ideas of Backtracking Algorithm

One of the characteristics of the backtracking algorithm is that the space occupied is not much, and the use of branch-and-bound method for the deeper levels of the problem in the solution sought can prevent a certain problem from appearing and then recurring again [29].

The application of the improved algorithm to the automated scheduling system is to divide a complete scheduling session into two phases: one is the time allocation phase and the other is the classroom allocation phase. Combined with the characteristics of the greedy algorithm, at the level of time allocation, based on the unallocated time slots, to select the best time unit for the classroom. In case of time allocation lockout, the backtracking algorithm can be used to backtrack upwards, search for the closest node that forms a time conflict, and then start the scheduling operation again to appropriately deal with and solve the conflict situation. Figure 1 depicts the flow chart of an improved automatic scheduling algorithm that uses the backtracking and greedy algorithms.

In the design process, the various elements of the entire scheduling process are studied, and the time that can be scheduled for teachers, classes, and classrooms is represented using the method of constructing a time data set to create a three-dimensional array, where the first dimension of the array represents the number of sections of the course scheduled per day, the second dimension represents the number of days in a week, and the third represents the number of weeks in the first semester of the academic year 2020-2021. Using the three-dimensional array (X, Y, Z) represents the teacher, the classroom and the course of the busy situation, according to the value of the busy field to determine whether or not it can be scheduled, if the corresponding value is “1” means that it can be scheduled, is “0” means that the time period has been used with the task assigned, can not be scheduled.

The priority function is used to calculate the course priority: (25) $L (g) = J (g) * T_{1} + Z (g) * T_{2} + R (g) * T_{3}$

In equation (25), Z(g) represents the number of hours per week of the course, J(g) is the level of the course, according to the requirements of the elective courses and mandatory courses to distinguish between the elective courses and mandatory courses, for 1 is an elective course, for 2 is a mandatory course, R(g) is the number of people electing courses for the course, T₁, T₂ and T₃ are set coefficients. The implementation steps of the improved automatic course scheduling algorithm specifically implemented according to the designed process are as follows:

The first step uses equation (25) to calculate the priority of each course, which is saved in this course table and exists in the corresponding priority field.

The second step queries the available time units. Perform the initialization operation of the maximum acceptable time array of a course to be arranged, understand and get all the classes to be studied for this course by looking for it, and then proceed to analyze and process the teacher’s time array in accordance with a certain order with the time array of the classroom related to its teaching, find and understand the time array of the classes of different classes, and find out the time of the classes of different classes that have already succeeded in completing the arranged course, and at the same time Find out the class time of the different classes that have been successfully scheduled and, at the same time, “match” it with the maximum time array that the course has, so as to obtain the time units that cannot be scheduled for the course. Next, the teacher’s time array is analyzed and processed in a certain order with respect to the classroom time array related to his/her teaching. In the third step, a search is made to find the times that allow the scheduling to be carried out, and the appropriate times are matched with the set number of weekly hours in the time pattern database.

3.3

Application of the Optimization Model of English Teaching Resource Allocation in Y Colleges and Universities

In order to verify the effectiveness of the optimization model of classroom resource allocation in college Y, some raw data of classroom resource allocation in college Y are selected for the application of the optimization model, and the effects of classroom resource allocation before and after the optimization are compared in terms of classroom time slot utilization, classroom capacity utilization, the number of classrooms planned to be used for the same course, and the displacement of teachers and students in completing the tasks in consecutive time slots.

3.3.1

Building the test environment

The CPU in the test environment is 2×2.6 GHz, the server version is Tomcat 6.0 and above, the backend database of the system is MS SQLServer 2000, the running environment of the client is the operating system of Windows 2003, the version of the browser is 6.0 and above, the memory of the system is 1 GB, and the test tool used in the experiment is Load Runner 7.8. Load Runner 7.8 is used as the testing tool.

3.3.2

Time to access information on ELT resources

In the process of simulation testing, different quantities of English teaching resource information are collected, and the regular management system is utilized as the comparison object to conduct the simulation experiment. The simulation curves of different quantities of English teaching resources information are shown in Figure 2.

Before the start of the simulation test, prepare two computers with high configuration and identical models, install the same simulation software, and load the 10 sets of simulation data in Figure 2 into the simulation software. Two no-load experiments were conducted before the simulation test officially started. Then, according to the data interface of the simulation software in the computer, access the conventional English teaching resources information integrated management system and the English scheduling management system based on the greedy algorithm, respectively, and prepare to carry out the system response time simulation experiment. Finally, the obtained data is sampled and analyzed to determine the time required for different experimental groups of users to obtain information on English teaching resources. The resource data information of the English scheduling management system based on greedy algorithm shows a linear growth trend with the increase of simulation data, and when the simulation data is 10 groups, the number of English teaching resource information is 921.

The time for users to obtain the information of English teaching resources obtained from the simulation experiment is plotted as a time curve, as shown in Figure 3. When the number of English teaching resources information is less than 300, the time difference between the two systems to obtain the teaching resources information is not much. However, when the number of English teaching resources information is more than 300, the time for the conventional management system to acquire the teaching resources information is slowly shortening, while the time for the greedy algorithm-based English teaching scheduling system to acquire the teaching resources information is sharply decreasing, and the shortest time to acquire the information is 16.84672s.

Calculating the average value of 10 groups of experimental data, it can be concluded that compared with the conventional comprehensive management system of English teaching resources information, the time for users to acquire teaching resources information of the greedy algorithm-based English teaching scheduling system is shortened by 33.48s, and it is suitable to be used for the optimal distribution of English teaching resources information.

3.4

Analysis of the results of the optimization of the allocation of resources to the classroom for course tasks

3.4.1

Utilization of classroom slots

The results after optimization of the classroom resource allocation for the English remedial examination task at the beginning of the school year in the first semester of the first week of the academic year 2021-2022 in the Friday morning time slot at the new campus of University Y are obtained through model solving, and the optimization results are analyzed in this section comparing with the original data of the classroom resource allocation for the English remedial examination task at the beginning of the school year in the same time slot.

Table 1 shows the comparison before and after the optimization of the utilization rate of English course classroom slots.

Table 1.

The English curriculum is compared with the optimization of the classroom

Index name	Classroom floor	Original scheme	Optimization scheme	Optimization effect
Number of classrooms used	1	30	35	5
	2	45	53	8
	3	43	47	4
	4	45	47	2
	5	30	30	0
	6	13	15	2
Number of classroom sessions used	1	1132	1264	132
	2	2048	1942	-106
	3	1624	1764	140
	4	1698	1566	-132
	5	1236	1278	42
	6	501	412	-89
Classroom time utilization rate	1	64.21%	58.32%	-5.89%
	2	69.48%	61.25%	-8.23%
	3	61.45%	59.42%	-2.03%
	4	60.21%	56.31%	-3.90%
	5	65.18%	54.98%	-10.20%
	6	66.18%	56.26%	-9.92%

Compared with the original plan, the optimization plan of classroom resource allocation for English courses in College Y has made more obvious improvements to the problems of insufficient use of classroom resources and insufficient consideration of inter-floor variability in the allocation of classroom resources, and the specific optimization effects can be seen in the following three aspects: 1)

Increase in the number of classrooms: the optimization scheme of classroom resource allocation for the course assignment of college Y has increased the number of classrooms by 21 compared with the original scheme, which makes fuller use of each classroom within the allocation range. Before the optimization, only the classrooms on the 4th and 6th floors of the higher floors were all put into use, while after the optimization, the number of classrooms on the 1st, 2nd, and 3rd floors of the lower floors was significantly increased and all of them were put into use, and at the same time, although the number of classrooms on the 5th floor of the higher floors was increased, one classroom was not yet put into use, and priority was given to the allocation of the classrooms on the lower floors.

2)

The number of time slots used by classrooms on lower floors has increased. Taking each 2 floors as a whole, the analysis of the number of classroom time slots: the number of classroom time slots on floors 1 and 2 changed from 3180 before optimization to 3206, with a certain degree of increase; the number of classroom time slots on floors 3 and 4 changed from 3322 before optimization to 3330, with a certain degree of increase; the number of classroom time slots on floors 5 and 6 changed from 1737 before optimization to 1690, with little change in the value. 1737 to 1690, with little change in value. Overall, priority is given to the allocation of classrooms on the lower floors.

3)

The number of time slots used in the classrooms on the lower floors decreases to a lesser extent. Since the time slot information of the course assignments remains the same before and after optimization, the number of time slots used in the classrooms does not change overall, and after optimization, the increase in the number of classrooms used will inevitably lead to a decrease in the time slot usage rate of the classrooms used. Although the time slot usage rate decreases from 64.45% before optimization to 57.76%, there is a significant difference in the decrease in the usage rate of the classrooms on each floor. Considering each of the three floors as a whole, the sum of the decreases in time slot utilization for classrooms on the lower floors 1, 2, and 3 is 16.15%, and the sum of the decreases in time slot utilization for classrooms on the higher floors 4, 5, and 6 is 24.02%. As a whole, the hourly utilization of classrooms on the lower floors is relatively low, which prioritizes fuller use of classrooms on the lower floors.

3.4.2

Classroom capacity utilization

Table 2 shows the utilization rate of classroom capacity for the English course. Compared with the original scheme, the optimized scheme of classroom resource allocation for the course assignment of college Y has improved the problem of insufficient use of classroom resources, and the specific optimization effect can be seen in the following two aspects: 1)

Expanding the capacity range of classroom use: the optimized plan for classroom resource allocation for course assignments in University Y increases the capacity range of classroom use by adding classrooms with capacities of 130 and 170 compared with the original plan, and classrooms of all capacities are used, which makes fuller use of classroom resources of all capacities within the assigned range.

2)

Increased utilization of classroom capacity. After optimizing, the utilization rate of classroom capacity increased to 55.712%, with a 1.44% increase and a 2.65% percentage increase. Although the capacity utilization rate of some classrooms appears to be lower when calculated using this method due to the large difference in the number of students between course tasks and the uneven number of classroom resources of each capacity, the main purpose of the classroom capacity utilization rate here is to compare the changes in the classroom capacity utilization rate before and after the optimization. Considering all capacity classrooms as a whole, the sum of the classroom capacity utilization rate increase is 70.89%, and the maximum capacity utilization rate increases from 80.69% to 90.23%, with an increase of 9.54% and an increase ratio of 11.82%, and the minimum capacity utilization rate increases from 8.64% to 22.98%, with an increase of 14.34% and an increase ratio of 165.97%, indicating that the optimized classroom resource allocation scheme can improve the capacity utilization rate of the classroom resources for English courses in Y college more substantially, and concentrate the classroom capacity utilization rate in a smaller range and use the classroom capacity more fully.

Table 2.

English curriculum classroom capacity utilization

Index name	Classroom capacity	Original scheme	Optimization scheme	Optimization effect
Utilization rate of classroom capacity	40	73.18%	67.82%	-5.36%
	50	78.65%	63.15%	-15.50%
	60	61.57%	90.23%	28.66%
	70	65.36%	69.15%	3.79%
	79	80.69%	87.26%	6.57%
	80	53.97%	62.36%	8.39%
	83	72.58%	69.48%	-3.10%
	90	65.36%	65.46%	0.10%
	100	62.78%	56.32%	-6.46%
	105	53.98%	76.25%	22.27%
	115	58.41%	72.18%	13.77%
	117	53.95%	59.25%	5.30%
	119	63.18%	62.18%	-1.00%
	120	59.36%	49.36%	-10.00%
	125	60.48%	46.18%	-14.30%
	127	54.32%	61.58%	7.26%
	130	-	46.36%	-
	135	73.65%	53.98%	-19.67%
	140	46.98%	59.36%	12.38%
	145	68.15%	77.25%	9.10%
	150	50.69%	68.21%	17.52%
	155	65.15%	58.36%	-6.79%
	160	60.36%	58.36%	-2.00%
	165	49.15%	45.11%	-4.04%
	170	-	41.69%	-
	175	39.18%	48.59%	9.41%
	178	32.48%	51.36%	18.88%
	180	47.96%	42.69%	-5.27%
	196	38.15%	41.85%	3.70%
	225	40.36%	33.96%	-6.40%
	250	33.85%	50.48%	16.63%
	267	36.98%	34.95%	-2.03%
	283	41.98%	22.98%	-19.00%
	300	39.45%	31.48%	-7.97%
	500	8.64%	24.69%	16.05%

3.4.3

Number of classrooms used for the same lesson plan

Table 3 shows the comparison before and after the optimization of the number of classrooms used in the same course plan. Compared to the original plan, the optimization plan for classroom resource allocation for course tasks in college Y has improved the problem of insufficient use of classroom resources. The specific optimization effects are as follows:

Table 3.

The same course plan USES the number of classroom Numbers to be compared

Pointer name	Number of instructors	Use · Number of classrooms	Original scheme	Optimization scheme	Optimization effect
Number of course plans	1	1	32	55	23
	1	2	21	0	-21
	1	3	2	0	-2
	2	1	3	3	0
	2	1	0	2	2
	3	2	1	0	-1
	3	3	2	1	-1

Distinguish the same course plan by the number of lecturing teachers, and analyze the number of classrooms used for the same course plan: for the same course plan with the number of lecturing teachers of 1, the number of classrooms used before optimization is 1, 2 and 3, and after optimization, the number of classrooms used is 1, and the number of the total course plan is 55, and the number of the classrooms assigned to the same course plan is equal to the number of lecturing teachers of the course plan, and the number of classrooms assigned to the same course plan is equal to the number of lecturing teachers of the course plan. Overall, the number of classrooms assigned to the same course plan is equal to the number of lecturers of the course plan, and the same course tasks of the same lecturer are assigned to the same classroom. The optimized classroom resource allocation scheme can reduce the number of classrooms used in the same course plan of Y college more substantially, so that the same classroom can be more fully used in the same course plan.

3.4.4

Faculty and Students Displaced Between Course Tasks for Completion of Consecutive Sessions

Table 4 shows the comparison between before and after optimizing the displacement between tasks for teachers and students who complete consecutive sessions of the course. 1)

The number of consecutive slots allocated in the same teaching area increased. Analysis of the number of consecutive slots with teachers as the main body: the number of consecutive slots allocated in the same teaching area has increased from 470 before optimization to 1,036, with an increase of 566, and the percentage of increase is 120.43%. Analysis of the number of consecutive time slots by students: the number of consecutive time slots allocated in the same teaching area has increased from 845 before optimization to 1259, with an increase of 414 and an increase of 48.99%. On the whole, since the allocation rule before optimization only considers the displacement of the student body between consecutive time slot course tasks, after optimization, the increase ratio of the number of consecutive time slots allocated in the same teaching area with the teacher as the main body is much higher than that of the number of consecutive time slots allocated in the same teaching area with the student as the main body, and the optimized classroom resource allocation scheme can reduce the number of consecutive time slots allocated in different teaching areas with teachers and students by a relatively large margin. The optimized classroom resource allocation scheme can greatly reduce the situation that teachers and students are assigned to different teaching areas, so that the classrooms in the same teaching area can be more fully utilized for teachers and students to complete the consecutive time slots.

2)

Less displacement between students and faculty completing continuous hour course assignments. Since a large increase in the number of consecutive periods assigned to the same teaching region is likely to cause an increase in the displacement between consecutive period course tasks assigned to the same teaching region, it is necessary to take the ratio of the displacement between consecutive period course tasks assigned to the same teaching region to the number of consecutive periods assigned to the same teaching region as the average displacement between course tasks for each consecutive period to conduct the Analysis. After the optimization, the average displacement between course tasks per consecutive time slot changes from 1.15 to 1.13 for teacher-oriented courses, and the average displacement between course tasks per consecutive time slot changes from 1.02 to 1.22 for student-oriented courses, and the values do not change much and remain above and below 1. From the previous analysis, it can be seen that prioritizing the fuller use of low-floor classrooms, where students and faculty complete fewer displacements between course tasks in consecutive periods, will further enhance the use of low-floor classrooms between course tasks in consecutive periods.

Table 4.

The students and students completed the continuous shift of the course

Index name	Body	Whether it is assigned to the same teaching area	Original scheme	Optimization scheme	Optimization effect
Continuous period quantity	Teacher	YES	470	1036	566
	Teacher	NO	561	24	-537
	Students	YES	845	1259	414
	Students	NO	452	5	-447
Continuous time shift between tasks	Teacher	YES	540	1166	626
	Teacher	NO	-	-	-
	Students	YES	863	1536	673
	Students	NO	-	-	-

4

Analysis of the Improvement of English Achievement by Optimization of Teaching Resources

4.1

English Learning Achievement

4.1.1

Distribution of accomplishments

Data analysis of the results of the English teaching assessment of the whole year, the greedy algorithm based English course scheduling system has its unique advantages, Figure 4 shows the distribution of the students’ English grades, the greedy algorithm based English course scheduling system through the optimization of the English teaching resources, thus affecting the students’ English grades, most of the students’ grades are concentrated in the interval [80,90], the number of people is 35, which accounted for 32.9%.

4.1.2

Comparison of achievements

Figure 5 shows the English grade passing rate comparison, comparing the English passing rate of this paper’s system and the traditional English learning system. Students who use the English teaching resources optimization system based on greedy algorithm have an English passing rate of 90%, while the traditional English learning system has a passing rate of less than 60%, with a difference of 36% between the two systems.

4.2

Scoring

4.2.1

Reading comprehension

Figure 6 shows the scores of reading comprehension questions. The average score for using this paper’s system for the optimal allocation of English teaching resources is 27.075, and the average score rate is 66.65%. Class A1 has the highest average score, with some students achieving the highest score of full marks and an average score of 31.5.

4.2.2

Written expression

Figure 7 shows the scores for the Written Expression question type, the average score for Written Expression for the whole year was 10.575, similarly the A1 class scored the highest score for the whole year with a maximum score of 15 and an average score of 13, which is a score of more than 80%.

5

Conclusion

In this paper, based on the principle of network teaching resources allocation, the optimal allocation of English teaching resources is implemented by combining multi-rate cognition. The mathematical model of an automatic class scheduling system is constructed, backtracking is introduced using a greedy algorithm, and an improvement scheme for the system is proposed. Taking Y university as an example, the scheduling system constructed in this paper is applied to it, and the effect of optimizing English teaching resources is studied through simulation experiments and real case analysis. Load 10 groups of simulation data in the computer, respectively access to the conventional English teaching resources information integrated management system, and the greedy algorithm based English scheduling management system, the response time of the system is analyzed by simulation experiments, and the greedy algorithm based English scheduling management system shows linear growth, and when the simulation data is 10 groups, the number of English teaching resources is 921. The analysis of the optimization results of classroom resource allocation shows that the optimized classroom capacity utilization rate is 55.712%, with an improvement of 1.44%, and the percentage of improvement is 2.65%. The sum of the overall classroom capacity rate improvement is 72.33%, and the maximum capacity utilization rate has increased from 80.69% to 90.23%, with an improvement of 9.54%. This indicates that optimizing the allocation of classroom resources can substantially increase the capacity utilization rate of English course teaching resources in Y college, and the allocation of resources is more reasonable. After using the English course scheduling system in this paper, the subjects’ grades are mainly concentrated in the interval of [80,90], the number of which is 35, accounting for 32.9%.

Lingua:: Inglese

Frequenza di pubblicazione:: 1 volte all'anno
Argomenti della rivista:: Scienze biologiche, Scienze della vita, altro, Matematica, Matematica applicata, Matematica generale, Fisica, Fisica, altro

Feed RSS della rivista

Research on Optimizing Teaching Resource Allocation Strategies with Machine Learning Models for Intelligent English Teaching Systems

Wenjing Lv

Pubblicato online: 24 mar 2025

Ricevuto: 28 ott 2024

Accettato: 15 feb 2025

DOI: https://doi.org/10.2478/amns-2025-0706

Parole chiaveNetwork Teaching Resource Allocation, Multi-Rate Cognition, Nonlinear Optimization, Greedy Algorithm, Global Optimal Solution, English Teaching

© 2025 Wenjing Lv, published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Parole chiave
Network Teaching Resource Allocation, Multi-Rate Cognition, Nonlinear Optimization, Greedy Algorithm, Global Optimal Solution, English Teaching