CN114328609A - Course arrangement method based on combination of meta-heuristic algorithm and greedy algorithm - Google Patents
Course arrangement method based on combination of meta-heuristic algorithm and greedy algorithm Download PDFInfo
- Publication number
- CN114328609A CN114328609A CN202111673588.3A CN202111673588A CN114328609A CN 114328609 A CN114328609 A CN 114328609A CN 202111673588 A CN202111673588 A CN 202111673588A CN 114328609 A CN114328609 A CN 114328609A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- course
- course arrangement
- meta
- constraint condition
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a course arrangement method based on a meta-heuristic algorithm and a greedy algorithm, and belongs to the field of teaching course arrangement methods. The method comprises the steps of rapidly extracting configured basic information from a database, and generating an initial solution value by using a mechanism with constraint according to the requirement of course arrangement; after data are read, data analysis is carried out by adopting a meta-heuristic algorithm and a greedy algorithm, data solution spaces are separated and then are searched respectively, after local optimal solutions of the solution spaces are obtained, comprehensive matching is carried out on the solutions, and a final result is close to the optimal solution. The method adopts a mode of fusing multiple algorithms, and finds the optimal course arrangement in millions of different course arrangement combinations by combining an optimization meta-heuristic algorithm and a greedy algorithm. The algorithm takes courses as the center, carries out searching and matching and takes the value of the optimal matching. The course arrangement success rate can reach 99 percent at most, and the problem of course arrangement conflict can be solved well, so that limited school education resources are utilized to the maximum extent.
Description
Technical Field
The invention relates to the technical field of teaching course arrangement, in particular to a course arrangement method based on the combination of a meta-heuristic algorithm and a greedy algorithm.
Background
After the middle school realizes the class-moving system, the problem of course arrangement influences the improvement of teaching quality and the full utilization of high-quality teaching resources of the school. How to solve the problem of multi-class resource combination optimization of schools by means of information technology and artificial intelligence, aiming at limited teachers and materials, teaching sites and teaching time resources, the optimal teaching target is achieved to carry out comprehensive and effective planning, and the problem of urgent need to be solved by various middle schools after the current college entrance examination reform is solved instead of artificial course arrangement.
The school timetable contains a plurality of subjects, grades and classes. Each class is around 40 knots each week. Under different permutation and combination, a school can have trillions of different school timetables. Reasonable schedules need to meet multiple requirements, for example, a single subject needs to be arranged in a balanced way in a week, the subject needs to be arranged in the morning as much as possible, the last day of a physical course in the morning and afternoon, and the old teachers do not need to arrange too early and too late. A good course arrangement system needs to find a scheme which can meet most requirements in countless possibilities, and the course arrangement system on the market does not adopt a constraint optimization method to solve the complex course arrangement problem.
Course arrangement, namely course arrangement, means that in order to carry out teaching work normally by a school or an organization, classes, teachers, courses and teaching resources are reasonably arranged, and behaviors of various courses schedules are formulated. The course arrangement is a very complicated matter, various factors such as courses, teachers, students, classrooms and the like must be considered, the course arrangement workload becomes very large, and an effective and flexible course arrangement method can help to train schools to optimize resource allocation and reduce waste. The course arrangement system in the market has computer technology, such as genetic algorithm, particle algorithm and the like, and although the application of the computer technology improves the course arrangement and course selection work efficiency, some problems are gradually revealed along with the development. (1) The course arrangement mode is single. Course arrangement software widely applied to schools is single in course arrangement mode, does not start from the basic characteristics of practice courses, and is difficult to control once the situation occurs. (2) The general course arrangement system in the market is difficult to meet the personalized course arrangement scene.
Disclosure of Invention
Aiming at the prior art, the invention provides a course arrangement method based on the combination of a meta-heuristic algorithm and a greedy algorithm. The course arrangement problem is a combination problem which takes teaching plans and various special requirements as constraint conditions under various elements such as courses, time, teachers, classes, classrooms and the like from the aspect of mathematics, and the essence of the problem is to solve the conflict among all the factors. In the design of the algorithm, in order to reduce the complexity of the course scheduling algorithm and win the course scheduling optimization engine system, the idea of breaking up the whole into parts and the priority algorithm combining the meta-heuristic algorithm and the greedy algorithm are mainly adopted.
The technical scheme adopted by the invention is as follows: a course arrangement method based on a meta-heuristic algorithm and a greedy algorithm is characterized in that configured basic information is quickly extracted from a database, and an initial solution value is generated by using a mechanism with constraint according to the requirement of course arrangement; after data are read, data analysis is carried out by adopting a meta-heuristic algorithm and a greedy algorithm, data solution spaces are separated and then are searched respectively, after local optimal solutions of the solution spaces are obtained, comprehensive matching is carried out on the solutions, and a final result is close to the optimal solution.
Specifically, the basic information includes course, time, teacher, class, and classroom. The course arrangement activity is to arrange and combine teachers and students according to different constraint conditions in time and space, and due to the limitation of course resources, mutual restriction necessarily exists between different courses in resource utilization.
Specifically, the course scheduling method implementation engine of the application adopts java8 to implement the meta-heuristic algorithm and the greedy algorithm.
Specifically, the mechanism with constraint in the method is a limitation condition for class scheduling, and includes, according to priority: the school teaching system comprises a hard constraint condition, a neutral constraint condition and a soft constraint condition, wherein the hard constraint condition is a constraint condition which must be met by a course arrangement, the neutral constraint condition is between the hard constraint condition and the soft constraint condition, and the soft constraint condition is a condition which is met as much as possible for more reasonable class schedule on the basis of meeting the hard constraint condition and the neutral constraint condition according to the practical situation of a school.
The reasonable setting is that the method also comprises manual course arrangement. Therefore, blind search is avoided, the quality of course arrangement is improved, and the backtracking times are reduced.
Compared with the prior art, the invention has the advantages that: the key point of the application is a core course arrangement algorithm: at present, various traditional course arrangement systems at home and abroad relate to a plurality of algorithms, but a whole set of mature algorithm is not provided for the course arrangement of the teaching mode of the 'shift-walking system' of primary and middle schools, so the technical difficulty is very high. The method breaks through tradition, and finds the optimal course arrangement in millions of different course arrangement combinations by combining an optimization meta-heuristic algorithm and a greedy algorithm in a mode of fusing multiple algorithms. The algorithm takes courses as the center, carries out searching and matching and takes the value of the optimal matching.
The success rate of course arrangement can reach 99%, and the problem of course arrangement conflict can be solved well, so that limited school education resources are utilized to the maximum.
Detailed Description
The present invention will be described in further detail with reference to examples.
The school timetable contains a plurality of subjects, grades and classes. Each class is around 40 knots each week. Under different permutation and combination, a school can have trillions of different school timetables. Reasonable schedules need to meet multiple requirements, for example, a single subject needs to be arranged in a balanced way in a week, the subject needs to be arranged in the morning as much as possible, the last day of a physical course in the morning and afternoon, and the old teachers do not need to arrange too early and too late. A good course arrangement method is to find a scheme which can meet most requirements in countless possibilities, and the course arrangement method in the market does not adopt a constraint optimization method to solve the complex course arrangement problem.
Constraint-based optimization, in pedagogical optimization, is the process of optimizing an objective function for certain variables in the presence of constraints on those variables. The objective function is either a cost function or an energy function to be minimized or a reward function or a utility function to be maximized. The constraints may be hard constraints that set conditions for variables that need to be satisfied, or soft constraints that have values of some variable that, if and based on the degree, penalize the condition of the variable in the objective function to be unsatisfactory. Constraint optimization methods are based on feasibility and focus on constraints, narrowing down a large number of different possible solutions to more manageable subset constraints by adding constraints to the problem. The optimization method can search an infinite number of spaces to find a global approximate optimal solution. Each constraint is assigned a weighted penalty point. For example, if a class schedule violates the constraint of up to 3 classes in the morning in a language, the corresponding penalty point is added to the overall score of the schedule. By randomly exchanging courses and evaluating the total points of different schedules, we can find the schdule with the lowest penalty (i.e., violating the least number of important constraints).
The success of heuristic algorithms depends to a large extent on the efficiency of the algorithm. To improve efficiency, we introduce a binary index tree (i.e., a Fenwick tree (Fenwick tree) that can efficiently update elements and sum up certain time periods in a schedule.computing run time is reduced from O (n) to O (log n). the Fenwick tree structure is proposed in 1989 by Boris Ryabko and further modified in 1992 These two operations. This is achieved by representing the numbers as a tree, where the value of each node is the sum of the numbers in the sub-tree. The tree structure allows operations to be performed using only O (log n) node accesses. Although Fenwick trees are conceptually trees, in practice they are implemented using a planar array similar to the binary heap implementation as the implicit data structure. Given an index representing a vertex in an array, the index of the parent or child of the vertex may be computed by bitwise operating on the binary representation of its index. Each element of the array contains a sum of a pre-computed range of values and the prefix sum is computed by combining the sum with other ranges encountered during traversal up to the root. When modifying a table value, the sum of all ranges that contain the tree value will be modified during a similar traversal of the tree. The sum of the ranges is defined in this way: both queries and modifications to the table are defined in asymptotically equivalent times O (log n).
In the embodiment, a parallel MapReduce algorithm is adopted, and 60 servers run in parallel on the Alice cloud to generate trillions of different curriculums. 1200 threads are running simultaneously and each thread communicates with the other threads to solve a different part of the problem. The best results were extracted every 10 minutes using MapReduce, which was then used to break and seed new lines. MapReduce is a programming model for parallel operations on large-scale datasets. The concepts "Map" and "Reduce", and their main ideas, are borrowed from functional programming languages, as well as features borrowed from vector programming languages. The method greatly facilitates programmers to operate programs on the distributed system under the condition of no distributed parallel programming. Current software implementations specify a Map function to Map a set of key-value pairs into a new set of key-value pairs, and a concurrent Reduce function to ensure that each of all mapped key-value pairs share the same key-set. MapReduce is a distributed computing model, is proposed by Google, is mainly used for the search field and solves the computing problem of mass data. MR consists of two phases: map and Reduce, users only need to realize Map () and Reduce () two functions, can realize the distributed computation.
The course arrangement problem is a combination problem which takes teaching plans and various special requirements as constraint conditions under various elements such as courses, time, teachers, classes, classrooms and the like from the aspect of mathematics, and the essence of the problem is to solve the conflict among all the factors. In the design of the algorithm, in order to reduce the complexity of the course scheduling algorithm and win the course scheduling optimization engine system, the idea of breaking up the whole into parts and the priority algorithm combining the meta-heuristic algorithm and the greedy algorithm are mainly adopted.
The system quickly extracts the configured basic information from the database, and uses a mechanism with constraint to generate an initial solution value according to the requirements of course arrangement so as to achieve the optimal initial value effect and prepare for the next algorithm processing. And after the data are read, performing data analysis by adopting a hybrid algorithm, namely, combining a meta-heuristic algorithm and a greedy algorithm. The hybrid algorithm separates the data solution spaces and then respectively searches to obtain the local optimal solution of each solution space, and then comprehensively matches each solution to enable the final result to approach the optimal solution. The method gives consideration to time efficiency and result quality, and meanwhile, the system is further optimized through a meta-heuristic algorithm framework. The system designed by combining the meta-heuristic algorithm and the greedy algorithm has strong elastic adaptability and can easily cope with the situations of the curriculum burst request, the class hour change and the like.
By combining the meta-heuristic algorithm and the greedy algorithm, the method is mainly applied to the research and development of the following three conditions:
1) and (3) researching and developing various layered shift conditions:
the class can be divided according to the conditions of randomness, class, course selection combination, score, level layering, score sharing, teachers and the like;
the class can be hooked with the school book course;
the shift can be grouped.
2) Flexible, free-man operation was developed:
the flexible and free arrangement of people helps to quickly shift;
the class selection change of students can be flexibly coped with, and the local adjustment of classes is changed;
the course selection change and the student status change (turning in and out) of the students are processed quickly, and the course arrangement table information is updated in time.
3) Research and develop personalized class schedule for students and teachers
Class schedule for student teaching
Student's full class schedule (including routine administrative class schedule, textbook extended class schedule and teaching class schedule)
Class table for teacher teaching
Teacher's full class schedule (including routine administration class schedule, textbook extended class schedule and teaching class schedule)
Class timetable, class subject timetable, place of class timetable
Through the combination of an optimization meta-heuristic algorithm and a greedy algorithm, the optimal course arrangement is found in millions of different course arrangement combinations.
The automatic course arrangement system realized by the method comprehensively adopts cloud computing and micro-service, is a real-time intelligent course arrangement system applying mainstream web technology, adopts Javascript on an interface, and adopts java8 to realize a meta-heuristic algorithm and a greedy algorithm by a core course arrangement engine. The intelligent school course scheduling system carries out functional modeling through an artificial intelligence algorithm, intelligently analyzes and calculates an optimal solution after the optimization algorithm, meets the individual requirements of course scheduling work of each school, realizes coexistence and individual coexistence, is more flexible and convenient to use, reduces unnecessary work of course scheduling personnel to a certain extent, changes the original management mode, and improves course scheduling efficiency.
The system should be divided into 2 large modules
Course arrangement information management platform
(1.1) initial setup of subsystem Functions
The initial setup module includes two sub-modules, namely: the system comprises an overall condition setting module and a priority setting module;
the overall condition settings include: inputting a school period, selecting the number of weeks in class, selecting a five-day system or a six-day system, and distributing courses each day;
the priority setting includes: inputting a course type and selecting a priority level;
(1.2) basic information subsystem functionality
Adding, deleting and modifying field information, class information, teacher information and course information;
(1.3) user management subsystem functionality
Adding and deleting users, setting and modifying user passwords, setting user permissions and logging in again;
(1.4) Special arrangements of subsystem Functions
The system comprises two submodules, namely a teacher submodule, a class submodule, a course submodule and a course submodule, wherein the resource requirements of the class and the course are different from the resource requirements of the field;
(1.5) safety subsystem function
The safety subsystem is an important component of the whole system platform and comprises system module management, user authority management, role management, user group management and log management. The whole management system is safely controlled, different people see different data, and different data items can be operated. The system also has an access log function, and can track the access condition of the user to the module.
(1.6) Universal tool subsystem functionality
The method comprises the functions of report management, work task definition and the like. The proportion of the report forms in the management system is large, the report forms change frequently, and the report form management can allow a user to define the report forms by himself and meet the requirement that the report forms change continuously.
Class (II) arrangement management system function
The system comprises three submodules, namely an automatic course arrangement engine, a manual course arrangement and a class schedule query. The class schedule query module comprises a printed class schedule.
The scheduling problem is actually a quintuple consisting of Course, Class, Teacher, Classrom, Timetable, a subset of the Class set, Teacher set, Classroom set, and time slice set. The course arrangement activity is to arrange and combine teachers and students according to different constraint conditions in time and space, and due to the limitation of course resources, mutual restriction necessarily exists between different courses in resource utilization.
The optimal belief automatic course arrangement engine considers various course arrangement restriction factors and finds the optimal course arrangement in millions of different course arrangement combinations through the combination of an optimization meta-heuristic algorithm and a greedy algorithm. The algorithm takes courses as the center, carries out searching and matching and takes the value matched firstly; the method has the characteristics of small occupied space and high operation speed.
1) Based on 2018 new requirement of reading from college entrance examination, arrange the class system and arrange the class again in must repairing the class and can choosing the class, accomplish the perfect adaptation of course, mr, student, satisfy the time limit condition of course, satisfy the limit condition of course, satisfy mr and arrange the class needs, satisfy the incidence relation between the course and satisfy multiple arrangement policy:
carrying out routine class-scheduling, wherein class-selecting subjects and qualified subjects (study-taking) are respectively carried out class-scheduling;
expanding class arrangement by class, and carrying out class arrangement by mixing grade (examination) subjects and qualified (examination) subjects;
the whole course of the external voice of the number of the languages is carried out and scheduled;
and (4) carrying out class scheduling a little, and re-dividing administrative classes according to the class selection combination condition of students to realize the micro-class scheduling of part classes and part subjects.
2) Resolving multiple constraints
The lesson-arrangement restriction conditions can be roughly classified into the following three categories according to the priority:
hard constraints
I.e. the constraints that the course arrangement must satisfy. If a class is ensured not to have two classes in the same time period; ensuring that the same classroom does not have two classes in the same time period; ensuring that the teacher can not give class to two classes in the same time period; all courses must be scheduled, and so on.
Neutral constraint
The special constraint condition can be called as a special constraint condition, and is often between a hard constraint condition and a soft constraint condition according to the practical situation of a school. If the school has multiple school zones and is far away, the requirement is met; if the two courses of the teacher are in different school zones, the time interval between the two courses should meet the requirement; the teacher has special requirements for the time of class, etc.
Soft constraint condition
Namely, on the basis of meeting the two basic conditions, in order to make the school timetable more reasonable, certain conditions are met as much as possible. If the main class course is arranged in the morning as much as possible; the class course is a sub class course, and if the class is used for multiple times in a week, the interval of 1-2 days is needed; if the teacher attends the class in the same school zone, the teacher tries to arrange in the morning or afternoon.
3) Basic flow:
in order to avoid blind search, improve the quality of course arrangement and reduce the backtracking times, the idea of manual course arrangement is added into the algorithm, namely a manual and automatic combined mode, which is a more scientific type in the course arrangement software at present. Although most course arrangement software is called manual and automatic combination, little or no course arrangement software can be really done. The real hybrid manual part should have enough course arrangement guide, and the automatic part should have precise condition setting, so that the combined schedule of the manual course arrangement and the automatic course arrangement can meet the desire of the course arranger.
The course arrangement success rate of the method or the system can reach 99 percent at most, and the problem of course arrangement conflict can be solved well, so that limited school education resources are utilized to the maximum extent.
Claims (5)
1. A course arrangement method based on the combination of a meta-heuristic algorithm and a greedy algorithm is characterized in that: quickly extracting configured basic information from a database, and generating an initial solution value by using a mechanism with constraint according to the requirement of course arrangement; after data are read, data analysis is carried out by adopting a meta-heuristic algorithm and a greedy algorithm, data solution spaces are separated and then are searched respectively, after local optimal solutions of the solution spaces are obtained, comprehensive matching is carried out on the solutions, and a final result is close to the optimal solution.
2. The course scheduling method based on the combination of meta-heuristic algorithm and greedy algorithm as claimed in claim 1, wherein: the basic information comprises course, time, teacher, class and classroom.
3. The course scheduling method based on the combination of meta-heuristic algorithm and greedy algorithm as claimed in claim 1, wherein: the method implementation engine adopts java8 to implement the meta-heuristic algorithm and the greedy algorithm.
4. The course scheduling method based on the combination of meta-heuristic algorithm and greedy algorithm as claimed in claim 1, wherein: the mechanism with constraint is a limitation condition of course scheduling, and comprises the following components according to the priority: the school teaching system comprises a hard constraint condition, a neutral constraint condition and a soft constraint condition, wherein the hard constraint condition is a constraint condition which must be met by a course arrangement, the neutral constraint condition is between the hard constraint condition and the soft constraint condition, and the soft constraint condition is a condition which is met as much as possible for more reasonable class schedule on the basis of meeting the hard constraint condition and the neutral constraint condition according to the practical situation of a school.
5. The course scheduling method based on the combination of meta-heuristic algorithm and greedy algorithm as claimed in claim 4, wherein: the method also comprises manual course arrangement.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111673588.3A CN114328609A (en) | 2021-12-31 | 2021-12-31 | Course arrangement method based on combination of meta-heuristic algorithm and greedy algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111673588.3A CN114328609A (en) | 2021-12-31 | 2021-12-31 | Course arrangement method based on combination of meta-heuristic algorithm and greedy algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114328609A true CN114328609A (en) | 2022-04-12 |
Family
ID=81021664
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111673588.3A Pending CN114328609A (en) | 2021-12-31 | 2021-12-31 | Course arrangement method based on combination of meta-heuristic algorithm and greedy algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114328609A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116843525A (en) * | 2023-08-28 | 2023-10-03 | 湖南强智科技发展有限公司 | Intelligent automatic course arrangement method, system, equipment and storage medium |
-
2021
- 2021-12-31 CN CN202111673588.3A patent/CN114328609A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116843525A (en) * | 2023-08-28 | 2023-10-03 | 湖南强智科技发展有限公司 | Intelligent automatic course arrangement method, system, equipment and storage medium |
| CN116843525B (en) * | 2023-08-28 | 2023-12-15 | 湖南强智科技发展有限公司 | Intelligent automatic course arrangement method, system, equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106127355A (en) | The cource arrangement method of a kind of high efficiency smart and system | |
| WO2015089865A1 (en) | Intelligent simulation system for urban comprehensive emergency and operation method therefor | |
| Xu | [Retracted] Improved Genetic Algorithm to Solve the Scheduling Problem of College English Courses | |
| Wen-jing | Improved Adaptive Genetic Algorithm for Course Scheduling in Colleges and Universities. | |
| Xie et al. | Learning behavior analysis using clustering and evolutionary error correcting output code algorithms in small private online courses | |
| Chen et al. | Intelligent course plan recommendation for higher education: a framework of decision tree | |
| Dai et al. | Design and research of intelligent educational administration management system based on mobile edge computing internet | |
| CN114328609A (en) | Course arrangement method based on combination of meta-heuristic algorithm and greedy algorithm | |
| Adrianto | Comparison using particle swarm optimization and genetic algorithm for timetable scheduling | |
| Jia et al. | [Retracted] An Association Rule‐Based Multiresource Mining Method for MOOC Teaching | |
| Wang | College student employment management recommendation system based on decision tree algorithm | |
| Che et al. | Application of decision tree in PE teaching analysis and management under the background of big data | |
| Gao et al. | Evaluation method of creative dance teaching quality based on fuzzy comprehensive evaluation | |
| Stapleton et al. | Control Challenges for Social Systems Milestone Survey: Current Status and Future Vision | |
| Hajiyeva | Digital education in the context of New Economy | |
| Yang | Design on generating test paper based on simulated annealing algorithm | |
| Li | Research on the innovation of Intelligent Student management Assessment Information System based on python software design | |
| Du | RETRACTED ARTICLE: Application of English education big data system based on multi-source information fusion and machine learning | |
| Ma | The university students personal knowledge management strategy study | |
| Nie et al. | [Retracted] Application of Edge Computing in Physical Education Teaching Management | |
| Liu et al. | Exploration and Practice of P2G Teaching Mode of Big Data Architecture and Mode Experiment Based on Clustering Algorithm | |
| Xiong et al. | Research and Development of Education Informatization Platform Based on Multi-dimensional Education Big Data Collection and Analysis Algorithm | |
| OZOR | AN ANALYTICAL NOTE ON “TECHNOLOGICAL INNOVATION, ENTREPRENEURSHIP DEVELOPMENT, SCIENTIFIC RESEARCH AND DEVELOPMENT FOR ECONOMIC TRANSFORMATION” | |
| Qi et al. | Instruction Ability Intelligent Evaluation System (ES) on Account of Neural Network Algorithm (NNA) | |
| Zheng et al. | Distributed Education Talent Cultivation Model and Practice in Higher Vocational Colleges |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 214400 A601, 159 Chengjiang Middle Road, Jiangyin hi tech Zone, Wuxi City, Jiangsu Province Applicant after: Jiangsu Youlixin Technology Co.,Ltd. Address before: 214400 A601, 159 Chengjiang Middle Road, Jiangyin hi tech Zone, Wuxi City, Jiangsu Province Applicant before: Jiangsu youlixin Education Technology Co.,Ltd. |