In order to give full play to the functions of educational information management, educational resource sharing, distance education teaching, and family community service of the education cloud platform of xxx City, the overall architecture design of the cloud platform is divided into three parts according to the different functional levels of the cloud platform: Cloud platform infrastructure, cloud platform application system and cloud space.

As the underlying construction foundation of the cloud platform, the infrastructure includes the education metropolitan area network, cloud data center and cloud platform support environment. Provide computing resources, storage resources and network resources for the upper layer applications of the cloud platform.

As the carrier of various education, teaching and management activities, the cloud platform application system provides software guarantee for the efficient use of the education cloud platform, the collection of education and teaching data, and the analysis and utilization of smart campus big data.

The cloud space provides a customizable and editable personalized information portal that integrates teaching applications, teaching content, teaching tools and teaching exchanges for the three main roles of individuals, schools and regions.

Cloud platform infrastructure, cloud platform application system and cloud space as an organic whole, provide hardware and network foundation for cloud platform application system and cloud space through cloud platform infrastructure. Collect, clean, and analyze education and teaching data through the cloud platform application system and cloud space, and make full use of the cloud platform infrastructure. The cloud platform can be used efficiently and provide a powerful boost for the rapid improvement of teaching quality. Finally, the design concept of "unified identity authentication, unified data aggregation, unified resource management, and unified desktop presentation" is realized.

1. Overall Construction Concept

The construction concept of the xxx smart campus big data cloud platform is "building a platform, setting standards, applying applications, forming a system, centralized management, and characteristic construction".

1. 1. Build a platform

Build a unified smart campus big data cloud platform, unified planning, and unified implementation; through "two-level deployment and three-level application", an education cloud platform with a city-level big cloud and 14 district and county (city-directed schools) small clouds is formed. The platform realizes the sharing of hardware equipment, high-quality resources, advanced applications and data. Through the unified data center, data user operation center, unified message center, and application access unified access management, the organic interaction of various applications on the platform, the organic flow of data, the elimination of information islands, and the realization of everyone.

1. 1. set standards

Through the planning and construction of the smart campus big data cloud platform in xxx City, a series of standard systems have been formed to meet the needs of unified planning and management at the municipal level. The content of the standard system is as follows:

infrastructure construction standards
District and county-level local area network construction standards
School-level wireless WIFI construction standards
Application Construction Standards:
Data Interoperability Standard
Unified login authentication standard
Unified Messaging Center Standard
Third-party application access system standard
Smart Classroom Construction and Control Standards
Digital Reading Space Construction Standards
Maker Space Construction Standards
Resource Construction Standards
Network marking standards (micro-scan collection system, high-scan collection system)
Construction Standards for Electronic Class Signs
Construction Standards for Recording and Broadcasting Classrooms
Service system standard
platform as a service standard
Operation and maintenance service standards
Operating Service Standards
Training Service Standards
1. 1. Apply on

Deploy and launch a batch of applications that meet the needs of xxx smart campus big data cloud platform construction, including:

A public resource service platform centered on resources: such as education and teaching applications, test analysis and management applications, resource content and management applications;

Public management service platform with OA office and digital campus as the core: such as collaborative office system, digital campus system, teacher principal management system, student management system, etc.

At the same time, the platform is compatible and open, and through the third-party application access standards, it can meet the access of characteristic applications in various districts and counties. Featured applications are deployed on the district and county platforms, following the principle of who builds them and who applies them.

1. 1. Into a system

Through the overall planning and construction of the xxx smart campus big data cloud platform, a series of application systems are built. The content of the application system is as follows:

A smart campus teaching system centered on Internet + technology, including virtual simulation laboratory applications, smart classroom applications, maker spaces, digital reading, lesson preparation systems, online teaching and research systems, etc.;

A resource management system centered on the accumulation, management and use of high-quality resources, including: resource management, school-based resources, personal resources, resource collection, etc. The system is connected to the video cloud platform, smart classroom, and examination analysis system to achieve resource accumulation, Endogenous and exogenous, and sustainable use, let every high-quality resource come alive, be used by teachers and students, and give full play to the huge value of the resource itself.

Examination analysis and management system using micro-scan and high-scan equipment as data collection. It realizes educational administration management, examination management, data collection and data analysis, and connects the data to the cloud data analysis and decision-making platform in real time to provide objective data portraits for objectively evaluating students' learning effects and individualized development. The micro-scanning system is designed to meet the needs of normalized testing and analysis such as classroom testing and unit tests, and the high-scanning system is designed to meet the needs of mass processing of monthly exams, midterms, finals, and mock exams.

Lesson preparation, teaching and research system: The lesson preparation and teaching system is the basic system for applications such as smart classrooms, makerspaces, and online learning. It meets online and offline teaching requirements by invoking resources such as lesson preparation, question bank, and micro-classes in the resource management system. Through the video cloud platform, activities such as online teaching and research and online course evaluation are realized, and a system of lesson preparation and teaching and research is created.

Teacher-student management system: Through the teacher management system, student management system, and principal management system, all-round management of users is realized, including user basic information, growth information, evaluation information, etc.

Digital campus management system: Through digital campus student status, course selection, student moral education information management, daily management, etc., organically interact with the cloud smart campus platform, data docking, build a smart campus, form a virtual simulation laboratory, smart classroom, creative A series of smart spaces such as customer centers, smart libraries, smart laboratories, and smart apartments.

Home-school interaction system: Through the home-school interaction platform, it provides class space, address book, circle and other network activity spaces, allowing parents and schools to communicate closely, displaying various performance and growth information of children in school, and paying attention to students' healthy development.

Collaborative office system: build a smart and convenient mobile office platform through flexible and convenient web and mobile applications such as notification, public information, official document circulation, salary, contract approval, item purchase, and weekly calendar.

Comprehensive evaluation system: The comprehensive evaluation system includes teacher evaluation, student evaluation, dual-track evaluation, examination evaluation, classroom evaluation, education quality evaluation, etc. Evaluation objects include teachers and students. Assessment tools include: dual-track assessment, examinations, and testing.

Educational quality evaluation is an overall evaluation of educational ability and learning ability. Through evaluation methods, the overall level of teachers and students can be improved.

1. 1. centralized tube

Build a centralized management platform and deploy all planned applications on the city-level cloud platform to realize diversified data collection, centralized data collection, standardized data management, and visualized data presentation. The platform will also realize the unified management of users and the distribution of authority applications, so that one account can be used throughout the entire network. Roles and application permissions are personalized to meet the customization needs of various applications.

The "big data and operation and maintenance service visualization" platform that is planned and deployed in a unified manner can dynamically and real-time display the operation status of various devices and applications through a large screen, and can be connected to the inspection system, campus security system, big data visualization analysis system, network Various video data such as operation status, computer room operation status, and operation management center operation status provide evidence for decision makers and intelligent management for operation service personnel.

1. 1. Featured construction

The big data cloud platform of xxx Smart Campus, through the two-level deployment planning of 1 large cloud and 14 small clouds, and through unified access management, meets the needs of districts, counties and schools for personalized application access and deployment.

Featured applications are deployed on district and county-level cloud platforms, following the principle that those who build them will apply them.

1. Overall objective
  1. Talent training goal

Build a multi-level interconnection space from individuals to schools to regions, give full play to the value of intra-school application interconnection, inter-school resource interconnection and district-school data interconnection, and provide various users such as teachers, students, schools and regional education administrative departments An ecological system of sharing, win-win and mutual benefit, realizes the co-construction and sharing of high-quality resources in the region and the collaborative innovation of smart teaching. Through rich teaching methods and high-quality learning resources, students' learning efficiency and learning quality can be improved, and students' learning interest can be improved; through editing management tools and teaching tools, work burden can be reduced and teaching quality can be improved. The ultimate goal is to improve the professional level of teachers as a whole, improve the learning quality of students, and provide all kinds of talents with all-round development for the society.

1. 1. Promote the goal of modernizing the education governance system and governance capabilities

The separation of management, operation, and evaluation is the basic requirement and urgent task to accelerate the modernization of the education governance system and governance capabilities. The education cloud project will provide an important technical foundation and support for promoting the separation of management, operation, and evaluation, optimize the management process, and promote open and transparent procedures. Through the education cloud, we can master comprehensive and accurate educational basic data, realize the sampling survey of large samples and the accurate analysis of big data, monitor the development of regional education, the quality of education and teaching of schools at all levels and the development level of students' comprehensive literacy; use the education cloud Big data promotes scientific decision-making and efficient management of education.

1. 1. Platform construction goals

Through the construction of the education cloud platform, a high-speed broadband network covering all schools and education management departments in xxx City will be realized. Make full use of the basic advantages of the cloud platform to solve the problem of data islands between established systems, integrate and share education and teaching data, and maximize the operating efficiency of the application system. Combined with the cloud platform support environment and the large-capacity and high-efficiency storage capacity of the cloud data center, it provides basic conditions for the collection, cleaning and analysis of smart campus big data, and then provides data support for educational decision-making. Ultimately achieve six modernizations:

1. 1. 1. Smart campus platform construction standardization

Form a series of construction standards, including infrastructure standardization and platform support standards

standardization, application construction standardization, and platform service standardization.

1. 1. 1. Platform cloudification

Infrastructure cloudification: build computing resource pools, storage resource pools, and network resource pools; conduct unified and centralized management and centralized monitoring through the cloud management platform.

1. 1. 1. Cloudification of business capabilities

The concept of platform as a service deploys various applications on the city-level cloud platform, and users can access remotely through their accounts to build a "fat central server cluster, thin client campus network" service model.

1. 1. 1. Centralization of services

Through unified platform planning and deployment, establish and improve the smart campus big data cloud platform service standard system, and provide unified, standardized, and quantifiable educational cloud services.

1. 1. 1. App Mobility

Form an ecological system of school mobilization, management mobilization, office mobilization, and communication mobilization.

1. 1. 1. application extension

Meet the iterative update of educational informatization applications, and ensure the on-demand access of new technologies and applications.

1. 1. 1. resource sustainability

Adhering to the purpose of "making resources alive", the production, application, and evaluation of resources are formed into a complete system, and the resource management platform is seamlessly integrated with classroom applications, resource accumulation, and test analysis, so that high-quality resources can be used by users , play a huge value.

1. 1. 1. management visualization

Build a "big data and operation and maintenance service visualization" platform to dynamically display the visualization of various devices and applications in real time, so that you can intuitively understand the operation of devices, the use of applications, and the utilization of resources.

1. overall architecture design
  1. Overall architecture

The overall structure of the smart campus big data cloud platform of the xxx City Education Bureau takes the cloud platform infrastructure layer as the underlying hardware foundation, which includes the education metropolitan area network and cloud data center, providing network resources, computing resources and storage resources for the cloud platform. Through the basic support layer, it provides basic services such as unified authentication, unified authorization, log management, system monitoring, multimedia streaming, and operation and maintenance management for the cloud platform application system and Renrentong space.

The cloud platform application system includes two major application system components: the education resource public service platform and the education management public service platform. The cloud platform application system shares business flow and knowledge flow with personal space, school space, and regional space in Renrentong space. Finally, various functions and services are presented and interacted with terminal devices such as teaching terminals, smart phones, electronic whiteboards, and PC terminals.

1. 1. The overall architecture of the cloud platform

1. 1. System technology route design

Adopt the method of "fat central server cluster, thin client campus network" to avoid independent construction of districts, counties and schools, realize centralized management and maintenance, and fully reflect the construction idea of "cloud".

The entire education metropolitan area network is an independent private network, the core layer is in the Municipal Education Bureau, and it realizes high-speed interconnection with all districts and counties, and then connects to each school. However, due to resource constraints and other reasons, the unified Internet export mode is not adopted. All districts and counties set up separate Internet outlets, and the Municipal Education Bureau sets up separate Internet outlets, which must be fully considered when designing network routing.

If the districts and counties need to upgrade and update the existing network equipment and servers, they can apply to the Municipal Education Bureau before the start of the project, and the project will be constructed in accordance with the unified technical standards, and the cost will be the service fee borne by the district and county reflected in.

Smart Campus Big Data Master Plan

2015 is the first year of smart campus big data in China. Governments, enterprises, schools, researchers, managers, teachers, and the public have all begun to pay attention to smart campus big data, and relevant policy documents, research institutions, academic activities, and market products have begun to appeared one after another. However, my country's smart campus big data research and practice are still in the initial stage of exploration, and they are "crossing the river by feeling the stones", involving a series of key issues that need to be solved urgently (such as the natural collection of educational data, the security management and Privacy protection, seamless flow and open sharing of educational data, deep mining of educational data and learning analysis, etc.). Among them, the comprehensive, natural, dynamic, and continuous collection of educational data is the basic and leading work for building smart campus big data. This requires clarifying some basic questions: Where is the source of educational data? What data needs to be collected? What are the commonly used data collection techniques? What should I pay attention to when collecting? Big Data Development Analysis

1. Smart campus big data construction background
  1. strategic opportunity

Cultivating big data talents, using smart campus big data to deepen smart campus management, promoting smart campus reform and development, and using big data platforms to improve smart campus scientific research level and efficiency are not only important tasks for smart campuses, but also strategic opportunities for smart campus development.

Big data is gradually becoming a social infrastructure and a standard configuration for every organization. The "big" of "big data" means more: the data that humans can "analyze and use" is increasing in a large amount. Through the exchange, integration and analysis of these data, humans can discover new knowledge and create new value. And let many normalized cognitions, judgments, thinking patterns, product forms, and service models form a new look and evolution direction.
　　Big data has attracted great attention from the international community, and countries all over the world are accelerating the strategic layout of big data. The big data industry has risen to the height of national strategy and is increasingly penetrating into all aspects of economic development and social life. On September 5, 2015, the State Council issued the "Action Outline for Promoting the Development of Big Data". The document pointed out that "data has become a basic national strategic resource" and clearly stated in the "Public Service Big Data Project", one of the ten major projects launched To build educational and cultural big data.
　　Smart campus big data has risen to the national strategic level, which has attracted widespread attention and great attention from all walks of life. Smart campus big data will first solve the six major problems faced by traditional education (unbalanced development, monotonous methods, information invisibility, extensive decision-making, perceptual school selection, and blind employment), and boost education. All-round reform and innovative development.
　　Establishing the strategic position of smart campus big data in the development and reform of education in my country is an inevitable requirement for the modernization of national education. Smart campus big data is an important national strategic asset, a scientific force for comprehensive reform in the field of education, and the cornerstone of the development of smart campus.
　　Human society has ushered in the "big data era". Cultivating big data talents, using smart campus big data to deepen smart campus management, promoting smart campus reform and development, and using big data platforms to improve smart campus scientific research level and efficiency are not only important tasks for smart campuses, but also strategic opportunities for smart campus development.

1. 1. Big data industry policy support

At present, the policy support for big data is constantly improving, and big data has risen to the national strategy. Since "big data" first appeared in the "Work Report" in March last year, the State Council executive meeting mentioned the use of big data six times within a year. At the executive meeting of the State Council on June 17, Premier Li Keqiang once again emphasized the importance of the use of big data. On July 1, the General Office of the State Council issued "Several Opinions on Using Big Data to Strengthen Services and Supervision of Market Subjects".

On September 5th, with the approval of Premier Li Keqiang, the State Council recently issued the "Action Outline for Promoting Big Data Development" to systematically deploy big data development. Among them, the construction of smart campus big data system is also mentioned.

On December 21, 2015, the first seminar of "White Paper on the Development of China's Basic Smart Campus Big Data" was held in Beijing, which provided exploration and guidance for the construction of national smart campus big data.

1. The source of smart campus big data

Education is an ultra-complex system involving teaching, management, teaching and research, services and many other businesses. Different from the clear, standardized and consistent business process of the financial system, although the education business of different regions and different schools has certain commonalities, the differences are also very prominent, and the differences in business directly lead to more and more educational data sources. Multivariate and data collection is more complex.

Smart campus big data is generated from various educational practices, including teaching activities, management activities, scientific research activities and campus life in the campus environment, as well as learning activities in informal environments such as families, communities, museums, and libraries; Including online education and teaching activities, as well as offline education and teaching activities. The core data sources of smart campus big data are "people" and "things" - "people" include students, teachers, managers and parents, and "things" include information system campus websites, servers, multimedia equipment and other educational equipment.

According to different sources and scopes, smart campus big data can be divided into individual smart campus big data, course smart campus big data, class smart campus big data, school smart campus big data, regional smart campus big data, national smart campus big data and other six types, they are gathered step by step from bottom to top, from small to large:

1. 1. Individual Smart Campus Big Data

Including the basic information of teachers and students collected in the "Education Management Informatization Series Industry Standards (Jiaoji [2012] No. Behavior records, management personnel's various operation behavior records, teachers' teaching behavior records, etc.) and user status description data (such as learning interest, motivation, health status, etc.).

1. 1. Course Smart Campus Big Data

Refers to the relevant educational data generated around course teaching, including basic course information, course members, course resources, coursework, teacher-student interaction behavior, course assessment and other data. The course member data comes from the individual layer and is used to describe and relate to student courses. Learning related personal information.

1. 1. Class Smart Campus Big Data

Refers to various educational data collected in units of classes, including homework data, test data, course learning data, classroom record data, class management data, etc. for each student in the class.

1. 1. school smart campus big data

It mainly includes various school management data stipulated in the standard (such as general situation, student management, office management, scientific research management, financial management, etc.), classroom teaching data, educational affairs data, campus safety data, equipment use and maintenance data, classroom laboratories, etc. data, school energy consumption data, and campus life data.

1. 1. Regional Smart Campus Big Data

It mainly comes from schools, social training and online education institutions, including education administrative management data stipulated by national standards, various behavior and result data generated by regional education cloud platforms, various educational resources required for regional teaching and research, and various regional-level development data. Data on teaching, research and student competitions, as well as data on various social training and online education activities.

1. 1. National Smart Campus Big Data

It mainly gathers various educational data from various regions, focusing on educational management data.

1. Smart campus big data collection technology map

The collection of educational data requires the comprehensive application of multiple technologies, and the scope and focus of data collected by each technology are different. Figure 2 shows the technical system of educational data collection, including 4 categories and 13 common data collection technologies.

1. Big data construction of smart campus faces problems

Big data is a new variable in the transition of the smart campus industry, but it is also facing many challenges:

1. 1. Product homogeneity is serious

It mainly focuses on evaluating and evaluating products, focusing on how to "raise points";

The phenomenon of blindly following the trend is serious, and the products are homogenized with little difference.

1. 1. The analysis end is the overall short board

The current products are mostly concentrated on the relationship maintenance end and data storage end, lacking in-depth data analysis;

The data in the education industry is semi-structured and unstructured, and the technology on the analysis side is generally immature.

1. 1. Lack of uniform industry standards

The industry is going back to the "old road" in the early development stage of education informatization, developing blindly and chaotically;

The standard formulation work specifically for smart campus big data has not yet officially started.

1. 1. The value of big data has not yet been realized

The entire industry lacks influential big data practice cases, and there is generally insufficient recognition of value;

In the absence of influential cases, big data pricing lacks basis and standards.

1. 1. Insufficient scientific nature of the data model

The education business is extremely complex and unique. At present, most products only rely on IT ideas to build education databases, and the selection of data sources and the design of indicators and weights are often divorced from reality.

1. 1. Data rights system is not clear

The ownership, open scope, open methods, and privacy protection of educational data have not been clearly defined;

In the provision of educational data products and services, there are often great policy risks.

1. 1. The scale of data is increasing

The sources of data are diversified, the shared database covers many systems, the source of students in the smart campus continues to expand, and information continues to accumulate, resulting in the continuous increase of information in the database, which in turn brings difficulties to data mining, management and analysis.

1. 1. Lack of a stable and efficient big data environment

The big data environment used by different disciplines and majors in the smart campus mostly relies on the existing IT environment, resulting in chaotic and extremely unstable basic software and hardware environments for big data operation, lack of effective operation and maintenance management, and seriously affecting the normal teaching and scientific research work conduct.

1. 1. underutilized data

Campus information is still in the stage of collection and accumulation. Although the continuous development of mobile terminal systems has brought great convenience to information collection, the collected data only stays in the query stage, and the data has not been integrated, analyzed and sorted out. , so that this information has not been adopted by managers, and there is very little information as a basis for decision-making.

1. 1. New Challenges for Scientific Research Brought by Data Drive

The rapid increase of data leads to qualitative changes caused by quantitative changes, which changes the thinking and behavior patterns of scientific researchers in the traditional subject research field. How to find a new perspective on the research results of this discipline with the help of big data related technologies and resources has become an important topic in the current smart campus research.

1. Construction principles of smart campus big data cloud platform

Data collection is the basic and leading work for the construction of smart campus big data. With the gradual maturity of many new technologies (such as eye tracking technology, voice interaction technology, somatosensory technology, etc.), more and more data collection technologies will be applied to the field of education to promote real-time, continuous and convenient big data in smart campuses. collection. In order to ensure the sustainable collection of high-quality educational data, the following matters need to be paid attention to in the practice of educational data collection:

1. 1. to plan ahead

The construction and application of smart campus big data is a systematic project that requires top-level design in order to collect high-quality educational data purposefully and orderly. The content of planning and design includes: the scope of data collection, the data collection technology used, the deployment of data collection environment, the guarantee measures of data collection quality, the application purpose and scenarios of data collection, the storage scheme of data, the update mechanism of data, the exchange standards, etc. Educational data collection at different levels should have different emphases——National smart campus big data and regional smart campus big data should focus on management data collection, and at the same time pay attention to correlation and cross-analysis with big data in social, medical, transportation and other fields and excavation, focusing on the formulation of education policies and the balanced development of regional education; the big data of schools, classes, and courses should be based on the collection of teaching and learning activity data, and the focus should be on the improvement of teaching quality; individual big data should be based on the individual learners. The collection of behavioral data, state data, and situational data is the main focus, and the focus is on the personalized learning diagnosis and development of learners.

1. 1. have clear boundaries

Although big data has the characteristics of heterogeneity and source diversity, and the storage cost of data is getting lower and lower, it does not have to include all data, and data without value is not worth collecting and analyzing. The same is true for smart campus big data. Its collection should have a clear boundary, rather than blindly collecting any educational activity data. Which data to collect depends on the application purpose of the data. For example, in order to detect and evaluate students' learning progress, it is necessary to collect and analyze data such as course browsing, homework exercises, communication and interaction, and question and answer in real time, instead of collecting data such as students' diet and exercise. Of course, we do not deny the value of data such as diet and exercise in diagnosing the physical condition of students. The "data boundary" mentioned here is relative to the specific application purpose. The construction of any data analysis model needs to rely on a specific data set. Only in this way can the validity of the analysis model and the application value of the analysis results be guaranteed.

1. 1. To maintain continuity and regularity

In many cases, a student's homework performance alone cannot explain the problem, but if the previous homework performance and even the process data of each student in a class are collected, the overall learning effect of the students can be objectively evaluated. Find learning blind spots, diagnose teaching difficulties, and carry out targeted teaching and individualized tutoring. At this time, the homework data has "big" value. The collection of smart campus big data should adhere to the concept of "continuously creating value, standardizing and enhancing value". On the one hand, the collection of educational data should maintain continuity, that is, according to the previous planning and design, collect various educational data regularly, continuously and regularly, and generate big data through long-term accumulation of small data; on the other hand, in order to ensure the integration of later data Interchange and consistent processing, the collection of educational data should follow specific technical standards and norms. At present, the Educational Technology Subcommittee of the National Information Technology Standardization Technical Committee has done a lot of work in the development of education informatization standards, and some technical standards have become national standards. standard etc. In addition, some common international standards are also worthy of reference, such as the IMS-QTI (Question and Test Interaction) standard, xAPI (Learning Experience Record) specification, etc.

1. 1. The collection granularity should be as small as possible

Data granularity refers to the degree of refinement and comprehensiveness of data. Generally speaking, the higher the degree of refinement, the smaller the particle size; the lower the degree of refinement, the larger the particle size. Data collection should be at an appropriate level of granularity, and the level of granularity should be neither too high nor too low. This is because a low granularity level can provide detailed data, but it takes up more storage space and requires a longer query time; a high granularity level can quickly and easily query, but cannot provide detailed data. As far as the smart campus big data collection is concerned, on the basis of ensuring the validity of the data, the data granularity should be as fine as possible in order to mine more potential value from it. Traditional educational data is centered on scores, and what is collected from an assignment or a test paper is only a numerical symbol representing the grade, that is, the granularity of the collected data is relatively large. If based on the online learning platform or dot-matrix digital pen technology, each student's answering process can be collected, such as the order of doing questions, the dwell time of each question, the number of answer revisions and other more detailed process record data, it will be more accurate. To accurately judge which knowledge points students have doubts and the specific reasons for wrong answers (sloppy or lack of knowledge). Therefore, it can be said that "small particles gather big data, and big data contains great value".

1. 1. Data source analysis of smart campus big data
    1. Data coverage is narrow
Mainly digital campus system generates data
数据的维度少，业务来源不足
1. 1. 1. 有效数据量少
数据量百GB到几TB，数据量少
以结构化数据为主，达不到大数据要求
1. 1. 1. 数据接口不完善

部份数据源有错误，数据不一致
业务接口与数据结构不规范
1. 1. 智慧校园大数据服务用户分析

智慧校园类似一个小社会，用户群体较多，各部门都存在大数据需求，而关注的内容会有较大的区别。比如校领导关注全校基础数据和总体情况，用于战略决策与发展评估。管理部门关注学生的生活、消费和心理状态。教学部门关注学生成绩情况、教师教学质量和学生满意度等。因此，这些特点决定了高教大数据的应用模块和类型会比较丰富。

1. 1. 智慧校园大数据建设责任制问题

智慧校园的特点是数量据并不算大，几万人规模的数据比起我们之前参与的电信几百万人的规模来说不算大，但数据源丰富，而且重视数据关联分析。

现阶段有些智慧校园的大数据是由某些学院自已在搞，没有从全校的层面来进行统一部署，数据处于割裂的状态，大数据价值不明显。智慧校园大数据在很多学校属于一把手工程，需要由学校高层领导牵头，协调和部门的数据，并进行统一的顶层设计和全校规划，由具体的单位比如网络中心来落地建设。

1. 1. 1. 校领导
决策更加科学。
1. 1. 1. 教师
教学更加便捷
教学水平提升
1. 1. 1. 学生
随时随地个性化学习。
1. 1. 1. 家长
及时掌握子女在校情况。
1. 1. 1. 校园环境
更加绿色、环保、安全
信息化覆盖范围和服务水平有效提升
1. 1. 1. 教学管理与服务
管理更加简单高效
服务水平不断提升
1. 1. 1. 社会
更多优质教育资源和科研成果开放共享
1. 建设目标
  1. 实现数据的共享和交换

将学校各应用系统的数据进行集成和整合，使来源各异、种类不一的各类数据可以相互使用，丰富数据的来源，打破系统间的信息孤岛，实现数据的共享和应用。

1. 1. 大数据的采集和存储

研制数据适配接口，对接校内各应用系统获取各类异构数据，并采用大数据主流的框架和系统对数据进行统一存储，为数据的挖掘和分析打好基础。

1. 1. 大数据分析与决策

Use data mining, mathematical statistics and other related technologies to build a big data analysis framework, extract hidden, unknown, and potentially valuable information and laws in the data, and provide educational management, scientific research management, student management, and logistics management for the school. Provide decision-making and guidance for various tasks.

1. Significance of Smart Campus Big Data Platform Construction

As the cradle of high-tech talents and innovative technologies, the smart campus carries the dual mission of scientific research and talent training. In the new wave of science and technology, smart campuses should aim at the forefront of the times, closely integrate teaching and scientific research innovation, professional talent training and big data, and promote schools to make new progress in big data teaching, scientific research and innovation at a higher starting point. steps.

1. 1. Enable Personalized Learning

Integrate educational data mining and learning analysis technology, continuously collect learning behavior data, intelligently analyze, push suitable learning resources and provide personalized learning suggestions.

1. 1. Realize the reconstruction of the educational evaluation system

Collect the data of the whole process of teaching and learning, comprehensively and objectively record the growth trajectory of students, and guide the scientific and healthy development of student training mode and education quality management mode.

1. 1. Achieving Paradigm Transformation of Scientific Research

　　Solve scientific research management problems such as scientific research funds, provide convenient technical support and humanized services, and improve the efficiency of research and the credibility of results.

1. 1. Start a new model of "Big Data Maker"

Complete innovative applications and releases, improve the informatization construction of smart campuses, enhance the strength of smart campuses, accelerate the transformation of maker achievements, promote the industrialization of creativity, and create an influential "maker campus".

1. 1. Realize the reform of teaching mode

　　Teaching data analysis and prediction, changing the teaching mode, realizing personalized education, adjusting the teaching plan, optimizing the teaching method, and improving the teaching quality.

1. 1. Realize scientific education management

　　Pay attention to the identification of relevant relationships, emphasize the determination of causal relationships, discover hidden and useful information, and provide data support for education management and decision-making.

1. Smart campus big data demand analysis
  1. department

Departments formulate various indicators based on statistical reports to guide education development, and the use of big data technology to collect, mine, and analyze such data can better quantify the current status of education and use it to formulate educational resource allocation recommendations and education management policies.

1. 1. 1. Resource allocation

Analyze the relationship between the current situation of the student-teacher ratio and the teaching effect in a city, a province, or even the whole country, and obtain the optimal student-teacher ratio or guiding student-teacher ratio, so as to guide the corresponding policies for teacher transfer in various places;

Analyze the proportion of rural students boarding students in the school, guide the site selection of the new school, the abolition of the old school, etc.;

Do big data analysis on the vacancy rate or utilization rate of classrooms and laboratories in a smart campus, and weigh the redistribution of site resources based on the analysis results, and allocate sites with high vacancy rates (low utilization rates) belonging to certain classes/grades/departments Use tight classes/grades/departments for other venues.

1. 1. 1. education management

In the comprehensive evaluation of students, we can start with the rate of students' physical health reaching the standard, cooperate with special sensors to obtain data related to students' physical fitness, use big data technology to analyze the correlation model between students' physical fitness and students' exercise time, intensity, etc., and propose teaching advice;

Through the analysis of the indicators of the number of full-time teacher training (person-time/100 people), explore the correlation between the continuous training of teachers and the quality of teaching;

Analyze the correlation between the school enrollment ratio of migrant workers' children and local economic development, policy changes, per capita income growth and other factors, and obtain a model related to teaching management programs (new schools, teacher transfers, etc.) to guide the education department Formulate a reasonable teaching management plan.

1. 1. teacher
    1. Precise delivery of teaching resources

Accurate delivery of teaching resources means accurately searching for required teaching plans, teaching courseware, teaching videos, or methods for students to deal with problems, etc. During the search process, you can also refer to the course big data tags to filter and find suitable teaching resources.

1. 1. 1. teacher teaching evaluation

Through big data analysis of teachers' behavior (including course design types, teaching thinking circuit diagrams, etc.), and through horizontal and vertical big data comparisons, we can find out the problems of teachers' own teaching process, carry out targeted training and counseling, and finally get teachers' teaching evaluation.

1. 1. 1. Construction of cloud question bank

Establish a cloud question bank, and on the basis of grade and subject classification, use big data to classify question types, knowledge point types, difficulty levels, and problem-solving thinking types, and label the questions. Teachers can intelligently assemble test papers based on course tags, or assign personalized homework based on student learning data analysis.

1. 1. 1. Teacher Comprehensive Assessment

On the basis of multiple teaching feedback and student learning feedback, a comprehensive teacher evaluation report is formed to make a comprehensive evaluation of the teacher's teaching type, teaching advantages, teaching defects, and student type matching.

1. 1. student

Through the big data analysis of students' behavior, we can gain insight into the problems of students' learning and thinking, and carry out targeted training to improve.

Based on the analysis of students' weaknesses, combined with the analysis of behavioral data, targeted and specific learning suggestions and even learning programs are put forward.

Record the learning process of students, and accurately push the required teaching resources according to the learning progress.

For learning with clear goals (such as qualification examinations), implement a study plan that breaks down the big goal into small goals, analyze the user's daily schedule to judge the feasibility of the study plan or propose a study schedule. Analyze the user's learning progress and characteristics, and give prediction scores and learning improvement suggestions.

1. 1. Teaching Management

Analyze students' learning styles, ability models, and group learning characteristics, and give guidance on student teaching management systems (such as teaching schedules, teaching facility construction, etc.);

Give guidance on the student teaching management system (such as teaching schedule, teaching facility construction, etc.), and make specific requirements for teacher deployment and teacher teaching ability improvement;

Collect student behavior data (student learning characteristics, time arrangement, etc.), find problem students (including study habit problems, mental health problems) through multi-dimensional comparison of big data, and provide targeted counseling in a timely manner.

1. 1. Educational Technology Service Provider
    1. Platform Technology Service Provider

There are many application requirements for the big data technology of platform technology service providers.

For platform technology service providers, the biggest difficulty lies in the top-level design of the platform, that is, accurately grasping user needs and platform prospect planning.

Although big data technology cannot directly meet user needs, it can quickly and accurately grasp the current market situation and provide trend forecast analysis. From this perspective, the grasp of user needs by big data technology is more informative and practical than traditional questionnaire surveys or field research. Therefore, big data technology is of great guiding significance to the platform construction of platform technology service providers.

1. 1. 1. Speech recognition technology service provider

Speech recognition technology service providers have broad prospects for big data. Speech recognition is a hot field. In addition to educational applications, it also involves many fields.

The key technical indicators of speech recognition are recognition accuracy and recognition speed, and both of these indicators can be optimized through big data technology.

On the basis of establishing a huge speech and vocabulary database, use big data technology to match the user's speech content with the speech database, and understand the meaning expressed by the user according to the vocabulary database.

At present, the English speech recognition system has been gradually improved in the process of use, but due to the complexity of Chinese, there is a technical bottleneck in Chinese speech recognition, and product development needs to be broken through.

At present, iFLYTEK is the leading enterprise in speech recognition technology in the field of education in China, and the engine of iFLYTEK is used in the oral evaluation in all provinces.

1. 1. 1. IoT Technology Service Provider

IoT technology service providers have broad application prospects for big data.

The Internet of Things technology and big data technology are inseparable brothers. The data collected by the Internet of Things needs big data technology for analysis, and the development of big data technology requires the Internet of Things technology for data collection.

1. 1. Education platform service provider

Platform service providers are the focus of future education online industry chain construction, and the future development trend of the education industry will be to combine new information technology to achieve deep online and offline integration. Therefore, the application of big data technology in platform service providers is widely optimistic about the education industry .

1. 1. 1. Educational resource platform

Cloud storage of educational resources, sharing of high-quality educational resources;

Labeling of educational resources makes it easy for teachers and students to find;

Resources are pushed accurately.

1. 1. 1. Education management platform

Student learning progress management, learning situation management, learning situation analysis;

Teachers assign homework, mark test papers, upload course materials, etc.;

Teaching evaluation by both teachers and students.

1. 1. 1. O2O platform _

Accurate matching between teachers and students;

Improve matching efficiency.

1. 1. 1. Learning exchange platform

user portrait;

Exchange partner matching.

1. 1. user

Users are students and parents. Parents are the main payers in the education industry. The needs of students were introduced earlier, so here we only focus on big data applications on the parent side.

The parent-oriented business of the education industry is mainly carried out in the aspect of home-school interaction. Big data technology can be used to synchronize information to parents on the basis of informatization of educational content. Parent feedback can also be analyzed by big data clustering.

The parent-oriented business of the education industry is mainly carried out in the aspect of home-school interaction. Big data technology can combine location data to push the arrival information to parents in real time when students arrive at school or home, solving security problems in the education industry in disguise.

1. 1. The collection process must be ethical

Data privacy and security have always been one of the obstacles to the development of big data. The source of educational data collection comes from students, teachers, parents, and schools. The data is complex and diverse, and many information such as grades, rankings, and family backgrounds involve personal privacy. At present, domestic laws and regulations on education data privacy protection are not perfect, and the awareness of student data protection in schools and educational institutions needs to be strengthened urgently. Due to insufficient supervision, there are many unscrupulous companies in the education industry that privately sell the information of teachers, students and parents for commercial interests. Regardless of whether it is for research, management or commercial purposes, before any educational data is collected, it should follow the ethical code of educational data collection (relevant departments are recommended to compile it as soon as possible), and the data generator should also have a certain right to know and the right to choose. The original intention and ultimate goal of data collection should be based on the principle of "serving the development of education and serving the growth of teachers and students".

1. Scenario Analysis of Smart Campus Big Data Application

1. 1. department
    1. Example of student-teacher ratio

Using big data technology to analyze the teacher-teacher ratio of primary school students in Shanxi Province, it is found that the teacher-teacher ratio of primary school students in the province is 13.12, and the teaching quality of schools with a student-teacher ratio of 12.60-13.40 is generally higher than that of other schools. It is 13.00. Big data also found that the student-teacher ratio in the south of Shanxi Province is generally higher than 13.00, and the student-teacher ratio in the north is generally lower than 13.00. Therefore, the Shanxi Provincial Department can formulate a policy of tilting primary school teacher resources to the south.

1. 1. 1. Example of the proportion of rural students boarding students

The proportion of boarding students in rural areas in Heilongjiang Province is 32%, but the waste of educational resources is too serious, and it is necessary to consider concentrating educational resources by merging or withdrawing schools. Village A and Village B are adjacent villages. Through big data analysis, it is found that the proportion of boarding students in Village A is 20%, and the proportion of boarding students in Village B is 50%. The school in Village B can be withdrawn, and all students from the original school in Village B can be encouraged to transfer to boarding students.

1. 1. 1. Examples of vacancy/utilization scenarios for classrooms and laboratories

Combined with the Internet of Things and using big data technology analysis, it was found that the time utilization rate/space utilization rate of the 160 teaching and research rooms of Fudan University School of Journalism is only 40%, while the utilization rate of 100 teaching and research rooms of the School of Foreign Languages is as high as 80%, and the space is very tight. Therefore, after comprehensive consideration, Fudan University can divide or lend part of the teaching and research sections of the School of Journalism to the School of Foreign Languages, so as to improve the utilization rate of the school's overall site and realize the reallocation of resources.

1. 1. 1. Examples of scenarios for students' physical fitness and health compliance rate

After analyzing the students' physical fitness and health compliance rate collected by the education department of Henan Province, it was found that the average physical fitness and health compliance rate of junior high school students was 60%. Comparing the schools with a physical health compliance rate of more than 80% and the schools with a compliance rate of less than 60%, it is found that the schedule of physical education classes is also arranged twice a week. Most of the schools with a high compliance rate arrange sports such as football, basketball, and badminton. Larger sports events, but schools with low compliance rates can only carry out sports events such as table tennis, relay running, and radio gymnastics due to many factors such as space constraints and funding constraints. Therefore, in the coming year, a special application for education funds can be submitted to the education management department or the arrangement of physical education courses can be changed to improve the rate of students' physical fitness and health.

1. 1. 1. Examples of Scenarios for Reforming Teaching Standards

In the teaching standards for the second grade of elementary schools in a second-tier city, the framers believed that the word "run" should meet the pragmatic standard, but according to the actual data collected, less than 20% of the students can meet this standard. The standard of the word can be achieved only needs to reach cognition.

1. 1. educational institution
    1. Examples of teacher scenarios

Precise allocation of teaching resources: A teacher of a primary school in a mountain village in Guangxi Province is about to teach "Youth Runtu". Due to the low level of his own teaching courseware, he hopes to find a teaching courseware suitable for use in Guangxi Province, so that local students can understand it. After logging in to an educational resource platform, search for the required courseware through keywords such as "Guangxi", "Juvenile Runtu", "Courseware", "Picture Rich", and then quickly find the required resources by viewing the course label of each courseware. After a few times, the platform draws a teaching portrait of this teacher through big data technology, and predicts the teaching content she will need in the future based on her search interval and the sixth grade teaching schedule, referring to the course labels of her previous selection of courseware. Push the teaching resources that the teacher may need. After logging in to the platform, the teacher can immediately receive the push information of the platform and make a choice. If things go on like this, with the increase of the teacher's behavior data, it is possible for the teacher to directly receive the educational resources sent to her by email (or other means) without having to log in to the platform after being screened by big data.

Intelligent paper composition in the cloud question bank: A teacher in Guangxi Province is about to issue a midterm exam paper, hoping to find some reference questions online. After logging in to the cloud question bank system of an educational resource platform, first select the math question bank of the sixth grade last semester, and then select the first three chapters of the last semester, and then you can see some midterm exam papers to choose from. After the teacher selects a test paper labeled "Moderate", click on the multiple-choice question section to enter the multiple-choice question bank and select the question to be replaced according to the label of the inspection knowledge point, and then select fill-in-the-blank questions, judgment questions, calculation questions, and application questions in sequence. Questions and other types of questions to carry out the same operation. During the selection process, the teacher can clearly see the statistics of the number of knowledge point labels and the scores of the difficult and easy questions of the selected topics, and make appropriate adjustments according to the statistics. Finally, after selecting some topics and changing the numbers, a suitable midterm can be generated Exam papers.

1. 1. 1. Example of a student scene

Tracking Xiao Ming's behavioral data, big data analysis found that Xiao Ming's average listening time in English class is only 30min/45min, and the error rate of English multiple-choice questions is the highest (30%), and grammar questions are the most (50%). After data collection and analysis on Xiao Ming’s problem-solving process, it is believed that Xiao Ming is a student with slow thinking and clear thinking. Therefore, in general, Xiao Ming should strengthen the speed of problem-solving and grammar learning. In addition, he can also give skills for multiple-choice questions.

A student has an average penalty of 10 points in mathematics due to a calculation error, while the average penalty for a student in the same grade is 18 points due to a calculation error. The analysis shows that the student's calculation weakness is not obvious, and he can focus on other weak points first. , usually pay a little attention to the calculation.

小红想报名三个月后的雅思考试，目标为6.5分，已通过英语六级考试500分。结合物联网采集数据和通话数据，运用大数据技术分析后认为小红每日可学习时间仅有4小时，即使再缩减其他事务，最多只能学习6小时，且分布于中午和晚上。根据大数据对雅思考生的观察，六级500分相当于雅思5.5分，同水平考生平均花费3个月每天8小时进行学习，最后雅思考试6.5分通过率为80%。因此不建议小红参加3个月后的考试，建议小红将学习时间延长到4个月，考试通过率较高。

1. 1. 1. 教学管理场景举例

大数据发现李老师和赵老师同为物理老师，班级平均分均为77分。李老师教导风格属于“全民提升型”，擅长讲解基础知识和难度适中的题型，全班物理成绩在80-85分的人数占70%；而赵老师教导风格属于“优升劣降”型，擅长讲解提升型知识和难题偏题怪题，全班物理成绩在90分以上人数和60分以下人数基本相等，且优生和差生之和占全班人数的55%。基于这种情况，可考虑让李老师带差生班，赵老师带优生班，发挥二人的教学特长。

结合物联网数据和家长反馈情况，发现广东省某初中班级学生完成课后作业时间平均为2小时，低于1小时的学生占20%，高于3小时的学生占10%。在低于1小时的学生中，差生占40%，可将此情况告知教师和家长，督促学生认真完成作业。在高于3小时的学生中，优生占70%，可基于3小时完成的作业量的不同重点关注做题效率较低的优生，由老师进行针对性辅导，找到低效原因，提高做题速度。另外还可结合物联网数据，对女生结伴上厕所的现象做统计，找到半年以来始终单独上厕所的女生，重点关注这类群体的心理健康，提前排查学生心理问题。

1. 1. 教育服务商
    1. 技术服务商场景举例
      1. 平台技术服务商

By integrating various external data (Baidu search data, educational institution student data, course data, etc.), and using big data technology analysis, it is found that the field that users are most interested in is language training (35%), and the market pain point in language training is spoken language Practice (60%), so the platform should focus on oral evaluation, combined with the functions or effects that users hope to achieve in oral practice (importance ranking). In addition, according to the user's field of interest ranking and market analysis, formulate the implementation steps of platform functions, and do a good job in the top-level design.

1. 1. 1. 1. Speech recognition technology service provider

In the process of establishing the voice database, big data technology is used to classify words and sentences with similar pronunciation and meaning. After hearing the voice of the user, the sound wave is quickly converted into an analog signal and then converted into a structured Data, quickly find the corresponding word/sentence in the voice database, and realize the voice recognition function.

1. 1. 1. Example of Platform Service Provider Scenario
      1. Educational resource platform

On the basis of the top-level design of the provincial and even national educational resource platforms, an educational resource sharing platform in Changsha, Hunan Province was built to integrate all electronic teaching resources in Changsha and upload them to the cloud using cloud storage technology. Teachers and students can easily view cloud teaching resources through the Internet, breaking the information islands of schools and counties in the past, and realizing resource sharing. When a school uploads electronic teaching resources, it needs to fill in the basic information of each resource, including grades, subjects, etc. After the upload is successful, the background collects user feedback and labels each teaching resource, such as "suitable for eugenics" and "difficult to understand". "Chinese and English subtitles", etc., are convenient for users to select suitable teaching resources. For the precise push of educational resources, please refer to the first scenario about teachers in the educational institution in the application scenario above.

1. 1. 1. 1. O2O platform

Ms. Wu hopes to find a teacher who can tutor her daughter in physics on weekends for her daughter in the second grade of junior high school. The students are weak, so she hopes to tutor the light and sound parts first. After Ms. Wu submitted her daughter's learning needs to the O2O platform, the background combined with map travel data and used big data technology to quickly lock in a physics teacher who was free on Saturday morning. 30 minutes on foot. After informing the teacher and parents of the matching result, both parties are very satisfied with the result, and the improvement effect after the trial lecture is also good.

1. 1. 1. 1. Learning exchange platform

Integrate the bbs platform of the national smart campus, connect K12 students and university students, and help students form a basic understanding of their future majors before going to university through platform communication. With the increase of platform coverage and question questions, big data technology can be used to effectively sort out and analyze, and plan a clear guidance route. For example, for a student who does not know any major, a systematic question and judgment link can be designed to guide students to choose Continue to learn more about a professional direction.

1. 1. user

Users are students and parents. Since parents are the main payers in the education industry, we divide the analysis into two types of application scenarios: students and parents.

Although parents are not the direct audience of the education industry, as payers and guardians, they can construct many big data application scenarios, such as synchronous management and analysis of educational content in schools and students' homes, and analysis of student safety positioning systems. Real-time understanding of students' dynamic system for the purpose of ensuring student safety, combined with LBS technology, records and analyzes students' action paths during daily learning, and realizes analysis of abnormal situations that are sent to parents and schools in a timely manner to solve security problems in the education industry. In addition to the most direct student safety services, the integrated student dynamic data can be combined with other impact indicators to generate new big data application scenarios (such as analyzing data such as the number of times students go to the library and the number of times they stay in the stadium).

In addition, parents and students live in different social environments, and the topics of concern are naturally different. How to let parents pay attention to the environment in which students grow up every day, and communicate with students better, instead of just staying on topics such as exam rankings. Student growth is also important. Use big data to mine topics that students care about, push them to parents, and give corresponding suggestions. For example: a new schoolbag with cartoon characters has recently been released and is very popular among elementary school students. Parents can buy their children’s birthday gifts after receiving similar information; a school organized a football league for middle school students. Parents can care about their children’s sports achievements . If we can collect the topics of every day or even every class, it will also have important reference significance for building a student learning analysis platform.

With the construction of the basic hardware layer as the underlying support, using cloud computing, big data data integration, data security, server clusters, data computing and mining analysis and other technologies, follow the "1 8" (1 central platform 8 major technical standards) big data The standard system builds a big data data center through HIVE and HBASE. On this basis, the company has independently developed the "Guozi Data Rubik's Cube" business development platform. Through the business development platform, information management of application development, application release, and service registration can be performed, and the application can be displayed to users with a visual interface. Decisions provide a data basis.

1. Smart campus big data architecture

1. 1. basic hardware layer

　　The basic hardware layer is built from a combination of inexpensive PCs or servers. The basic hardware layer mainly carries the tasks of data storage, calculation, fault tolerance, scheduling and communication, and executes and gives feedback to the instructions issued by the basic application layer.

1. 1. data integration

　　The characteristics of big data are manifested in real-time, interactive, massive, etc., and mainly semi-structured and unstructured data, with low value density. In order to better "let the data speak" and give full play to the value effect of big data, we should adhere to According to the principle of "collecting all that can be collected", the coverage of data sources should be as large as possible.

1. 1. Data Computing and Analysis Mining

　　The big data platform covers commonly used computing scenarios in big data scenarios, including offline computing, real-time computing, stream computing, data mining, and machine learning. It is easier and more convenient to build a data lake with a full life cycle, which enriches the process of data processing, processing, and innovation, thereby realizing greater value of data.

1. 1. Data Security

　　The big data platform realizes the security, maintainability, availability, and credibility of data resources through a series of authentication authorization and resource isolation mechanisms. Providing unified authentication service is responsible for verifying the user's identity; providing unified authorization service is responsible for controlling the user's resource access rights; providing unified resource scheduling is responsible for isolating the underlying resources used by users.

1. 1. server cluster

　　A server cluster is a collection of many servers for the same service. The cluster can obtain higher computing speed and can also be used as a backup. If any server is damaged, the entire system can operate normally. Clustering operations can reduce the number of single points of failure and achieve high availability of clustered resources.

1. 1. Big Data Technical Standards

　　A central platform: big data business development platform.

Eight major technical standards: basic standards, data representation standards, data processing standards, data storage standards, data service standards, data security and privacy standards, industry big data standards, and big data product testing standards.

1. 1. Big data data center

　　Through related operations such as collection, preprocessing, analysis, processing, and storage of various types of structured, semi-structured, and unstructured data information, build a unified, standardized, and comprehensive big data data center to provide data support for related work .

1. 1. Big data business development platform

　　Taking Hadoop as the core, integrating excellent technologies, providing an open data and business development platform, and carrying out application development, application publishing, application registration, and application service information process management, thereby improving the experience of big data applications, which is conducive to exerting the spirit of innovation, Create unlimited value.

1. 1. Big data business visualization analysis

Big data business visual analysis can mine information and knowledge hidden in massive data, provide users with a visual operation and analysis interface, and provide data basis for users' related activities, thereby improving work efficiency.

1. Smart campus big data platform standard system

Big data information standards provide guidance and reference standards for data collection, data processing, data storage, data analysis and mining, accelerate the establishment of technical standards for smart campus information collection, storage, disclosure, sharing, use, quality assurance and safety management, and guide the establishment of Standards and specifications for information sharing and exchange, promote the development and utilization of information resources, realize the aggregation and integration of big data, and provide a strong basic support for big data and cloud computing.

Do a good job in the construction of the big data standard system, promote the implementation of the national big data strategy, and respond to three aspects of demand: facing the needs of smart campus education, develop open and shared big data standards; help the innovation and development of smart campus education, formulate relevant standards in typical fields; guarantee resources Security, protection of personal privacy, development of security standards, etc., data security management with data as the basic element, standardize the entire process of data sharing, use and management, and solve cross-platform data interaction, data open sharing and other problems.

The company sorts out the existing standards in my country, the standards under research and the standard plans to be proposed, based on the big data technology system, analyzes from different angles such as foundation, technology, products, applications, etc., and forms the framework of the big data standard system, according to "1 8" concept, that is, supported by the "Guozi Data Rubik's Cube" business development platform, to build basic standards, data representation standards, data processing standards, data storage standards, big data service standards, big data security and privacy standards, industry Eight technical standards for big data application standards and big data product testing.

Through the construction of platforms and technical standards, integrate and guide resources, activate technological elements, promote independent innovation and open innovation, and promote the healthy development of big data; accelerate technology accumulation, technological progress, and promotion of innovative achievements, and accelerate the wide application of big data in smart campuses , to promote the comprehensive, coordinated and sustainable development of smart campuses; to solve problems such as difficult data sharing, inconsistent data formats, non-standard data standards, and repeated data construction.

1. 1. basic standard

Big data terminology, big data reference architecture, and big data platform architecture standards can describe related operations more realistically, form a unified data standard, and provide basic support and services for the database.

1. 1. data representation standard

　　Data coding specifications, metadata specifications, unstructured data, unified description specifications of datasets, etc., ensure the interactive sharing of data information, thereby eliminating information islands.

1. 1. Data Processing Standards

　　Data quality evaluation standards, data collection standards, data organization standards and other standards and specifications related to the big data processing stage can eliminate the influence of variables' own variation and numerical value, and lay a good foundation for big data applications.

1. 1. data storage standard

　　New storage system-related specifications under the background of big data, such as non-relational database specifications and unstructured data management system specifications, are helpful for interactive transmission and management of data, improving storage capacity and storage speed, and further providing rapid mining of big data, Extraction and analysis provide the basis.

1. 1. Big Data Service Standards

　　Provide standardized description and access of a series of big data services such as big data real-time analysis service and visualization service, improve the correlation between data, reduce the complexity of data analysis, and greatly improve the accuracy of analysis.

1. 1. Big Data Security and Privacy Standards

　　When big data is used for external services, standards for data storage security, data transmission security, and data analysis and mining security are formulated to provide reliable data storage, safe mining and analysis, and strict operational supervision for internal management and external attacks faced by security.

1. 1. Industry big data application standards

　　Standards for big data applications in related fields, classification and coding of big data in related fields can more accurately regulate the data standards of various industries, and launch products that match the industry.

1. 1. Big Data Product Testing Standards

Test scenarios, test indicators, test tools, etc. of big data products. The establishment of big data product testing standards can fairly and objectively evaluate the functions and performance of big data products, and has important reference value for people to choose suitable big data products.

1. Smart campus big data business development platform

In order to provide many convenient and easy-to-use development frameworks and service engines, so that users can quickly grasp, recognize and use the open data on the platform, and choose the appropriate service engine for secondary development according to different application scenarios, the company independently developed "national "Sub-Data Rubik's Cube" business development platform, while providing open data, also allows users to develop big data applications through this platform, providing users with integrated application development, testing, deployment, operation, management, monitoring and other hosting environments, enabling Application developers do not need to care about the underlying hardware and infrastructure construction of the application, thereby improving work efficiency.

1. 1. Smart campus big data business development platform architecture diagram

1. 1. n big data data center

　　The big data data center is not a simple integration of hardware devices, nor is it just a center for data storage, but a center for data circulation and application services. It has very rich information resources, safe and reliable computer room facilities, high-level network management and very complete value-added services. The data center is one of the basic projects of smart campus information construction.

　　The data center realizes data exchange and sharing between application systems through a unified data format. The smart campus data center has the following construction significance:

1. While collecting and storing various types of data, it can effectively manage the data, break the existence of "information islands", provide unified data services for various application systems of the smart campus, and ensure data consistency.

2. Provide real-time data for school departments and leaders. Various departments can easily view the public data of other departments; leaders can view the business data of all departments in the whole school as a whole, and can intuitively understand the situation of the school.

3. It is convenient for later application system development, separates application and data, reduces the difficulty of application system expansion and development, and lays a solid foundation for the comprehensive integration of smart campus application systems.

1. 1. Big data business platform layer

　　The big data business platform is constructed by components, big data processing engine, APP, and BI engine. The big data business platform is built based on the idea of PAAS and follows the SAAS standard.

1. 1. 1. components

Components are the most basic elements of the big data business platform. The component interface standard is built in the big data business platform, and all components follow this standard. Define the component's inputs, privates, and outputs in the standard. In the process of business development, one or more components are created, edited and associated through a graphical interface, so as to be combined into a data processing service and released to the outside world.

1. 1. 1. Big Data Processing Engine

　　The big data processing engine is the execution center of the entire business platform, and realizes the smooth operation of the entire service by performing operations such as parsing, scheduling, executing, iterating, and merging related components in the publishing service.

1. 1. 1. APP

APP is a light application composed of one or more published data processing services. The user edits the data and chart forms displayed by the APP through the graphical interface.

1. 1. 1. BI drawing

The BI engine parses and presents the data binding service and chart form of the built APP.

1. Key technologies for the construction of smart campus big data platform

The construction of the big data platform draws on the advanced concept of the open source system, adopts the Hadoop open source system, and makes full use of the reliability of HDFS. Good performance in terms of ease of use and performance.

1. 1. Hadoop technology

The Hadoop framework is an open source large-scale data processing platform and tool, mainly derived from the MapReduce programming framework proposed by Google, GFS file system and BigTable storage system and other technologies. Hadoop has a huge family system, and the construction of this platform mainly involves the distributed file system HDFS and MapReduce model of the Hadoop framework. As the bottom layer of the Hadoop framework, the distributed file system is mainly responsible for the distributed storage and management of analytical data, and the MapReduce model is mainly responsible for computing and processing large-scale data sets. Hadoop uses the HDFS file system sub-framework to achieve its storage capabilities, and uses the MapReduce programming model framework to achieve its computing capabilities. The ingenious combination of the two makes Hadoop have efficient storage and computing capabilities.

1. 1. HDFS technology

HDFS distributed file system is an effective tool for distributed storage and management of large-scale data. It is also the storage basis of distributed computing. It has high fault tolerance and scalability, and provides high throughput for data reading and writing. HDFS realizes the distributed storage of data, enables applications to access large-scale data sets more flexibly, and also provides a data platform for subsequent analysis of large-scale data. The HDFS distributed file system adopts a typical master/slave structure, which greatly simplifies the system architecture, makes the system more concise and facilitates system management. Hadoop's distributed file system HDFS is mainly composed of a main controller and data nodes. The main controller manages the name space and data nodes, and manages the mapping of data blocks to data nodes.

As a data node, the DataNode in the file system mainly stores actual data, is mainly responsible for storage management on the physical node where it is located, and executes commands issued by the main controller. Data nodes can receive read and write requests sent by customers in a timely manner, and complete corresponding operations for these requests. From the perspective of the structure of the distributed file system, data files are stored and divided into multiple data blocks and stored on each data node. Each data node stores data blocks from multiple files. At the same time, each data node also stores Multiple copies of these data blocks will be stored to ensure the accuracy of subsequent data operations.

1. 1. MapReduce technology

MapReduce technology is based on a distributed file system. By writing corresponding processing procedures, parallel computing and processing of large-scale data sets can be realized. By writing related MapReduce processing functions for different analysis modules, accurate analysis of large-scale data can be realized. At the same time, it can Control each node to complete efficient task scheduling. MapReduce distributes the operation to each node on the network, and each node will periodically return the work it has completed and the latest status, so as to realize the operation of large-scale data sets. This processing method ensures the reliability of the operation .

The way MapReduce technology handles is to first decompose a specific task into several small tasks, then distribute the decomposed tasks to each sub-node, manage and schedule the sub-node tasks through the master node, and then get The results processed by the sub-nodes are then integrated to obtain the final results. Through the mutual cooperation and scheduling between multiple nodes, the calculation and processing of large-scale data sets can be realized. Generally speaking, MapReduce is based on the idea of "divide and conquer" to realize "decomposition of tasks and summary of results".

1. The construction effect of smart campus big data platform

The construction of a big data analysis platform for colleges and universities is an innovative exploration based on the strategic development plan of the smart campus and based on the outline of the information construction of the smart campus. Based on the top-level design of big data construction, collect and integrate data generated from all aspects of smart campus education, extract valuable information and models from the data, and promote the comprehensive innovation of smart campus education.

1. 1. Carry out the top-level design of big data, and comprehensively promote the development of the school with the application of big data

Big data will become a new driving force to promote the development of schools. Through the top-level design of big data, a comprehensive plan is made for the acquisition, collection, arrangement, and utilization of big data. Starting from application requirements, the purpose and path of construction are clarified, and what should be done and what should not be What to do, what should be done first, what should be done later, what model to use, what degree to achieve, and what effect to achieve, so as to guide the school's big data construction in the next 3-5 years.

1. 1. Rapidly promote the informatization of teaching and management work, and establish rich data sources

Use big data methods to comprehensively analyze existing teaching and management work, build or upgrade information systems, record the entire process of teaching and management, and establish rich data collection channels.

For example, by comprehensively upgrading the existing distance education system, jumping out of the concept of distance education, realizing comprehensive support for the teaching process, and recording the learning behavior data of each student in detail, including course learning data, video viewing data, data review data, homework Complete data, interactive communication data, performance data, etc., subdivide the data into the details of each behavior, so as to provide data fineness beyond the traditional system and objectively reflect the actual status of learning.

Through the establishment of Internet of Things applications, it can achieve strong support for the management of items, personnel, security, etc., and improve the quality of management while accumulating a large amount of management data and behavior data.

1. 1. Based on personalized service requirements, establish a big data analysis model

Providing excellent personalized service is one of the important goals of education and management, big data application is a necessary condition for providing large-scale personalized service, and the quality of big data analysis model determines the value of data. A data that is usually ignored will have an unimaginable effect in a good model.

Through the analysis of learning behavior data, we can understand students' learning interests and learning effects, and study which learning methods are most acceptable, which course designs are the most popular, or specific to the correct answer to a certain homework question Rates and horizontal and vertical comparisons, and show the reasons in depth. These data are provided to teachers, which will provide the most direct support for teaching innovation.

Through the analysis of the behavioral data of the card, we can understand the students' daily behavior and consumption rules, understand the relationship between student behavior, academic performance, and school effects, and provide a basis for student management innovation.

1. 1. Comprehensive application of big data results to promote comprehensive innovation in schools

Through the comprehensive application of big data, the analysis and judgment of various education and management work can be established, applied to actual work, and comprehensive innovation of schools can be promoted from all aspects.

Through the comprehensive analysis of the teaching process, learning behavior, academic performance, teaching satisfaction, teacher demand, professional teacher quality, professional maturity, and action track, establish teacher portraits, student portraits, and professional portraits to intuitively understand the advantages and disadvantages, and predict process status.

Through comprehensive analysis of the number of students, dormitory allocation, classroom use, energy consumption, network consumption, canteen consumption, library utilization, etc., the establishment of resource utilization index, data visualization, and guide the refinement of management work and the flattening of management .

Through cluster analysis of card, book borrowing, major distribution, course distribution, grades, learning behavior, etc., special groups with certain characteristics and their unique behavior patterns are found, and correlation data are used to mine to find rules .

Through comprehensive analysis of teaching data, teaching effects, attendance records, personal data, and management data, a scientific, authentic and objective teacher performance evaluation system can be established to change the subjectivity of traditional human evaluation and allow big data to select truly excellent teachers.

1. Smart campus big data common business system

The following are common business systems in smart campuses.

This is the data flow of a certain smart campus business system. We can find that the business system and the data center have not completely exchanged and shared data. Therefore, when we plan to build big data in the smart campus

It is necessary to coordinate the resources of various functional departments to confirm the data interface of the business system. At the same time, due to the inconsistency of data among various business systems, we still need to convert data into the data format we need during the process of collecting data, which brings a lot of workload.

There are hundreds of data types and attributes in the school business system. We also involved more than 40 kinds of data after summarizing and organizing the data.

For example: Zhengfang educational administration system mainly includes basic information of students (native place, place of origin, etc.), academic information (completed credits, uncompleted credits), course information (students have taken courses, semester schedule, etc.), grade information, etc.

1. Smart campus big data service user types

The departments in the smart campus that have specific needs for big data, as well as the specific needs of big data applications, need relevant corresponding analysis in the construction of smart campus big data.

1. 1. school leaders

1. The overall situation of teachers and students

2. School public opinion

3. Scientific research level

4. Assets

1. 1. Hospital leadership

1. The collective/individual behavior portraits of each major and class in the college

2. The courses and scientific research of the teachers of the college

1. 1. school union

1. Judging the rationality of poor students and scholarship results, and identifying fraudulent claimants

2. Labor and capital situation of teaching staff

1. 1. Equipment Department

1. Current status of existing assets

2. Asset development trend

3. The failure rate of each brand asset

4. Asset procurement planning of each unit

1. 1. library

1. Current status of library collections

2. Current status of book borrowing

3. Trend analysis

4. Book purchasing plan

1. 1. school clinic

1. Drug revenue and expenditure ratio

2. School health records

3. Prevention and treatment of infectious diseases

4. Analyze side effects of drugs

1. 1. teacher

1. Class results

2. Class/personal behavior portrait

3. Book borrowing and pushing books

4. Scientific research

1. 1. student

1. Achievements and abilities

2. Job market and recruitment guidance

3. Book borrowing

4. Consumption and Internet

1. 1. enterprise

According to the matching degree of student behavior profile and enterprise job requirements, recommend students to the enterprise for internship or employment.

The smart campus big data platform can provide integrated big data products and services for higher education, and realize end-to-end implementation.

In terms of data sources, the smart campus big data platform can realize the acquisition of data sources for school business systems, one-card systems, online behavior systems, various hardware devices, major website forums inside and outside the school, etc. For occasions where the data interface is not standard or not developed Targeted approaches are also possible. In terms of data collection, we can use the ETL collection toolbox of the smart campus big data platform and general collection technology to provide collection services for schools, and carry out targeted data cleaning and modeling work according to school business conditions and big data needs, and formulate Complete data specification standard system. At the same time, it can also integrate various step-by-step databases, sql databases, oralce databases, etc., build a unified data warehouse for the school, and provide a unified interface for various big data application services.

The smart campus big data platform will also provide customers with a set of big data management platform for effective monitoring and management of various links, including data status, data quality and accuracy, etc., to facilitate the school's daily operation and maintenance and management.

The big data business application is based on the Hadoop ecological cluster, and the structured, semi-structured and unstructured data related to the digital campus, such as people, finances, objects, and networks, uses HDFS and Mapreduce for distributed file storage and calculation, and uses association rules to analyze , cluster analysis, correlation analysis and other analysis methods, display the results to users in various visual graphics, and provide data support for leadership decision-making. Big data business applications include the application of teaching innovation, scientific research innovation, management innovation and other innovative applications.

To get the full version of the document, please add attention and communicate with private messages!

Smart Campus Big Data Cloud Platform (1)

To get the full version of the document, please add attention and communicate with private messages!

Guess you like