Detailed Construction and Application of Course Knowledge Graph

The knowledge graph (Knowledge Graph), also known as knowledge domain visualization or knowledge domain mapping, is a series of various graphics that display the process and structural relationships of knowledge development. It uses visualization technology to describe knowledge resources and their carriers, mining, analyzing, constructing, drawing, and displaying knowledge and their interconnections. The knowledge graph integrates theories and methods from disciplines such as artificial intelligence, applied mathematics, graphics, and information visualization technology, using visual graphs to illustrate the core structure, development history, frontier fields, and overall knowledge architecture of disciplines, achieving the goal of multidisciplinary integration.

The construction of knowledge graphs lays the foundation for the current evolution of artificial intelligence from perceptual intelligence to cognitive intelligence. With technological advancements, artificial intelligence has reached or surpassed human levels in perception fields such as “listening, speaking, and seeing”; however, it is still in the early stages of cognitive intelligence, which requires external knowledge, logical reasoning, or domain transfer. Cognitive intelligence will draw inspiration from cognitive psychology, brain science, and human social history, utilizing knowledge graphs, causal reasoning, continuous learning, and other technologies to establish effective mechanisms for stable acquisition and expression of knowledge, allowing machines to understand and utilize knowledge, thus achieving a key breakthrough from perceptual intelligence to cognitive intelligence.

A course knowledge graph utilizes subject experts and AI technology to manually or automatically organize course content, extract course knowledge points, and establish a relational network structure between knowledge points, associating knowledge points with course resources to form a structured and visualized result of course content.

The knowledge graph serves as a refinement and summary of course content, with broad applications. Particularly in online education, the combination of knowledge graphs and AI recommendation algorithms can provide personalized education based on each student’s mastery of knowledge points, plan learning paths, and display learning conditions, thus achieving tailored teaching for every student. Specifically, it includes several typical application scenarios:

– Course knowledge point graph display

– Intelligent adaptive learning: learning path display, learning resource recommendation, learning condition report

– Study assistance robots (automated Q&A for courses)

– Intelligent lesson preparation

The course knowledge graph is a product of the integration of AI technology and education, and its various applications are the main battlefield of AI + education, especially AI + online education. The vast online education resources combined with intelligent adaptive learning and recommendation based on knowledge graphs represent a significant disruption to the traditional educational model of “one teacher for many students” that has lasted for thousands of years.

Aopeng Education is a leading company in online degree education, with thousands of course resources covering dozens of subjects and hundreds of majors. There is a pressing need for the construction of course knowledge graphs, as well as unique data advantages such as courseware and corpus resources required for building knowledge graphs.

Since 2019, the engineering center has collaborated with the product center, student user division, and industry-education integration to explore various attempts at constructing and applying course knowledge graphs. The construction of knowledge graphs and related applications for courses/majors has been incrementally increased in phases and batches.

Through continuous exploration, experimentation, and preliminary research by the AI innovation group of the engineering center on relevant technologies and algorithms, an effective method has gradually been developed for semi-automated and semi-manual (90/10) extraction of course knowledge points, recognition of overall relationships between knowledge points, and association of knowledge points with resources (videos, documents, test questions).

The methods mainly include: video content analysis (OCR), audio content analysis (speech recognition ASR), syllabus analysis (web scraping, data cleaning), knowledge point extraction (natural language processing), recognition of overall relationships between knowledge points (natural language processing), and association of knowledge points with resources.

Depending on the different styles of each course and resource types, knowledge points are automatically extracted and recognized by AI technology, combined with manual correction, forming a streamlined process that allows the construction of a course knowledge graph to be completed within 1-2 person-days. Feedback from teachers in the product center indicated that constructing a course knowledge graph manually last year required half a month to a month. This shows that the automated extraction method greatly improves the efficiency of constructing course knowledge graphs.

Course Knowledge Graph Construction Process

By comparing the knowledge points extracted manually by teachers and those extracted automatically by AI for the same course (“Introduction to the Basic Principles of Marxism”), it can be seen that the knowledge points extracted by AI are more granular, with the number of knowledge points extracted being four times that of manual extraction.

Comparison of Knowledge Points Extracted Manually (Left) and by AI (Right) for the Same Video Segment

Hierarchical Structure of Knowledge Points Extracted by AI in the Above Video Segment

AI-Generated Knowledge Graph of the Fujian Normal University Version of “Introduction to the Basic Principles of Marxism”

The total duration of the course is 14 hours and 29 minutes, with a total of 466 knowledge points extracted by AI.

AI-Generated Knowledge Graph of the Fujian Normal University Version of “Introduction to the Basic Principles of Marxism”

Note: In the above image, the orange node represents a specific knowledge point, the green nodes represent associated test questions, and the yellow nodes represent associated videos. The “Teach” relationship attribute between the video and knowledge points includes information such as startTime and endTime, which can be used for precise video recommendations.

As of now, Aopeng has 30 course knowledge graphs generated manually by teachers from the product center and 4 course knowledge graphs generated by the latest AI technology. Various application scenarios have been developed based on these.

* Course Learning Condition Report

* Test Question Recommendations Based on Mistakes (Error Book)

* Precise Video Recommendations Based on Mistakes

* Automated Q&A Based on FAQ (in development)

In the future, while continuously promoting the construction of more course knowledge graphs, more application scenarios will be developed based on the content of existing knowledge graphs.

The construction and application of course knowledge graphs is an ongoing process. On one hand, the accuracy of AI algorithms and the degree of automation in the pipeline can be further improved; on the other hand, the automatic discovery of more relationships between knowledge points (e.g., prerequisite relationship recognition) will lay a solid disciplinary knowledge foundation for applications such as intelligent adaptive learning planning. The introduction of external knowledge (e.g., Baidu Encyclopedia, Chinese Wikipedia) will greatly enhance the algorithmic accuracy in the construction and application of knowledge graphs. In the future, the AI team will focus on continuously improving the construction and application capabilities of knowledge graphs in the following areas:

* Introduction of external encyclopedic professional knowledge (Baidu Encyclopedia, Chinese Wikipedia)

* Recognition of more relationships between knowledge points (prerequisite relationships, equivalence relationships, etc.)

* Generation of disciplinary knowledge graphs (cross-course terminology graphs)

* Automatic recognition technology for science formulas (OCR)

* Automatic error correction for speech recognition results

* Online book knowledge graph construction platform

* Automated Q&A Based on Knowledge Graphs

* Various inference and recommendation algorithm applications based on graph machine learning

* Collaborative filtering recommendation systems (based on XAPI user behavior collection)

* Intelligent lesson preparation

* Adaptive learning systems

The development and application of artificial intelligence technology are the most important and core technological trends of the 21st century. With the continuous launch and maturation of related algorithms, artificial intelligence has been widely applied in fields such as facial recognition, intelligent recommendation, and intelligent customer service.

AI + online education based on knowledge graphs will lead to profound changes in educational models and learning modes. The AI innovation team at the engineering center will delve into this blue ocean of AI + education, continuously exploring and combining business innovation with technological innovation, using innovation to establish a unique technological moat for Aopeng online education.