Interactive Teaching and Learning: Knowledge Graph Construction and Simulation Platform

To accelerate the integration of scientific research achievements into the classroom, explore the “Intelligent +” education model, and expand the depth and breadth of knowledge graph experimental teaching, Professor Lu Yonghe from the School of Information Management at Sun Yat-sen University and his teaching service team designed the Knowledge Graph Construction and Application Simulation Experimental Platform. This platform utilizes big data and artificial intelligence technology, covering the complete experimental process of planning, designing, constructing, optimizing, and applying knowledge graphs. It supports graphical simulation of entities and relationships between entities, helping students learn how computers simulate human cognitive processes.

The simulation experiment process covers the complete experimental workflow of planning, designing, constructing, optimizing, and applying knowledge graphs, and supports graphical simulation of entities and the relationships between entities. The simulation experimental platform integrates knowledge modeling, entity and relationship extraction, entity fusion, and knowledge storage, encompassing functions such as knowledge graph construction and application, featuring high concurrent access volume and fast response speed.

Experimental Introduction

Experimental Teaching Objectives

(1) Effectively address the abstraction of knowledge organization, management, and application that poses challenges to information management teaching.

(2) Accelerate the integration of scientific research achievements into the classroom and expand the depth and breadth of knowledge graph experimental teaching.

(3) Create a knowledge graph simulation experiment that meets the “high-order, innovative, challenging” standards, referred to as the “two qualities and one degree” gold course standard.

Experimental Principles

The principles of this experiment include knowledge modeling, entity and relationship extraction, entity fusion, knowledge storage, perspective analysis, dimensional analysis, and dimensional association analysis.

Experimental Teaching Process and Methods

This project adopts an “experiential + task-based” teaching method, using the construction and application of a knowledge graph for Chinese poetry as a guiding teaching example. Before the experiment, students receive systematic theoretical course instruction; during the experiment, students operate according to the experimental teaching guidance on the simulation software, completing tasks related to knowledge graph construction and application according to the experimental steps; after the experiment, students submit independently completed experimental results, write experimental reports, and reinforce the content of knowledge graph construction and application in experimental teaching.

Experimental Steps

Knowledge Graph Creation

(1) Create a new experimental project: Students can select a field of interest (such as the classification of ancient poetry) and add a project in the system, which will generate a project belonging to that student;

(2) Prepare experimental data: Collect available corpora, determine usable materials through comparative analysis, such as gathering relevant materials for the classification of ancient poetry from the internet (which can be prepared in advance before the experimental course), for example, an Excel file, and import it into the project;

(3) Vocabulary and Term Analysis: After selecting vocabulary to build a vocabulary list, perform term selection and analysis, and analyze terms that need to be supplemented. For example, extract relevant key terms from the materials obtained in the previous step, and supplement them if they are incomplete;

(4) Define the graph: Students can add, delete, modify, and query nodes in the system, create entities after setting node attributes;

(5) Publish the knowledge graph model: Click the publish button in the system; after the success message is displayed, the knowledge graph becomes effective;

Knowledge Graph Application

(1) Perform knowledge graph analysis tasks: Options include hot word analysis, clustering analysis, and perspective analysis. Here, perspective analysis is selected;

(2) Dimensional Analysis: The frequency of occurrence of various terms ranked in the vocabulary list can be seen;

(3) Dimensional Association Analysis: Analyze the relationships between different dimensions, such as the emotions a poet has towards a particular object;

(4) Save and submit experimental results: Students save their constructed knowledge graph models and analysis results to their accounts and submit them, filling out and submitting experimental reports based on the experimental situation.

The Knowledge Graph Construction Research Platform and Knowledge Simulation Experimental Platform at Sun Yat-sen University have now become a provincial experimental platform, with a view count of 13,200 and nearly 300 experimental instances.

Some research achievements are displayed as follows:

(1) Research Achievement – The Kidney Disease Nutrition Knowledge Graph constructed using the knowledge graph construction and application simulation platform.

(2) Teaching Case – Knowledge Graph of Olympic Games Winners

(3) Software Copyright – Knowledge Graph Construction and Application Simulation Platform V1.0

(4) Related papers published based on knowledge graphs

① Constructing a Diabetes Knowledge Graph Based on Deep Learning

② Constructing and Applying a Knowledge Organization Model for Oral Archives Based on Knowledge Graphs

The Knowledge Graph Simulation Experimental Platform is a practical exploration of the Intelligent + education model, intuitively and visually demonstrating the construction and application processes of knowledge graphs, addressing the challenges of knowledge graph experimental teaching, and expanding the breadth and depth of teaching.