Applying Teaching Agents in Project-Based Learning

Applying Teaching Agents in Project-Based Learning

by
With the rapid development of generative artificial intelligence, large model-based agents have gradually acquired capabilities such as multimodal perception, retrieval-enhanced generation, reasoning and planning, interaction, and evolution. How can teaching agents integrate into educational application scenarios? How can collaborative human-machine approaches be used to complete interdisciplinary thematic learning tasks? Let’s take a lookat the views of Associate Professor Lu Yu from the Future Education High-Precision Innovation Center at Beijing Normal University —
Applying Teaching Agents in Project-Based Learning
Project-based learning is an effective teaching model for cultivating students’ core competencies and higher-order abilities. In a typical project-based learning process, learners often need continuous support from teachers and peers to complete project outputs. Teaching agents can set project-based learning tasks and provide specific task planning for the completion of project outputs,support memory and reflection related to project-based learning content, and offer various capabilities such as multimodal project resource generation, retrieval-enhanced generative learning scaffolds, high-quality code generation, and feedback. At the same time, they also support human-machine interaction and multi-agent interaction modes.
As shown in the figure, teaching agents can assume the roles of “Teaching Assistant Agents” and “Peer Agents” in different project phases, each with distinct task settings, expansion capabilities, and individual memories, thereby demonstrating differences in abilities and functions to provide various interactive support for learners. We will illustrate the role of teaching agents in each phase of project-based learning using the common interdisciplinary theme of “waste classification” in information technology or artificial intelligence courses.
Applying Teaching Agents in Project-Based Learning
(1) Personalized Problem Proposal
Project-based learning requires driving problems to be proposed based on real situations, allowing learners to genuinely feel the urgency and feasibility of solving these problems, thereby stimulating their intrinsic motivation to explore deeply and complete the project. Therefore, during the problem proposal phase, the “Teaching Assistant Agent” can first establish a driving problem guidance framework based on the predefined learning scenario. On this basis, the “Teaching Assistant Agent” can engage in multimodal online discussions with learners and adopt personalized dialogue paths and interaction strategies according to learners’ characteristics and learning intentions, ultimately guiding learners to autonomously propose the driving problem for the project. The “Teaching Assistant Agent” can utilize the agent modules in open-source technology frameworks to achieve this primary function by setting the predefined guiding framework as the target question for each round of dialogue and making learners a necessary consulting tool in each round of task planning, thereby actively asking learners questions and initiating discussions.
For example, regarding the environmental theme of “waste classification,” the “Teaching Assistant Agent” can create real scenarios for learners, utilizing the external multimodal large model’s image generation capabilities to depict situational images such as “ocean garbage vortex” and “non-biodegradable plastic waste.” Meanwhile, the “Teaching Assistant Agent” can engage in online discussions about the urgency of waste management with learners based on its dialogue capabilities derived from large models. Combining learners’ specific feedback, the “Teaching Assistant Agent” can continue to propose possible necessary steps and methods for waste management, thus guiding students to think independently and clarify specific project activities to undertake, such as “how to promote the concept of waste classification” or “how to create a smart trash can for waste classification.”
(2) Collaborative Design of Project Plans
To address the personalized driving problems proposed by learners, teaching agents can establish dynamic discussion groups between learners and agents, helping learners to determine specific solutions based on their educational task planning capabilities and decompose the plan. Group discussions can adopt either “Agent-led” or “Learner-led” modes based on project goals and learners’ styles. In the “Agent-led” mode, teaching agents can utilize open-source technology frameworks to construct multiple “Peer Agents,” simulating and playing different roles of human group members in the project-based learning process, facilitating multi-role interactions between human learners and multiple “Peer Agents.” In this process, the “Teaching Assistant Agent” is primarily responsible for selecting the speaker (either human learners or “Peer Agents”) for each round based on the group’s dialogue history and project goals, broadcasting the spoken content to all group members, thereby achieving collaborative design of the project implementation plan through multiple rounds of speaking and information transmission. In the “Learner-led” mode, learners can directly choose to engage in dialogue with different “Peer Agents.”
Specifically, in the “Agent-led” mode, to solve the driving problem of “how to promote the concept of waste classification,” the teaching agent can first utilize its task planning capabilities to decompose the solution to the driving problem into several executable sub-tasks, such as “understanding waste classification rules,” “collecting various typical waste examples,” and “creating promotional materials and carriers.” Based on the planned sub-tasks, the teaching agent can construct multiple “Peer Agents” to discuss specific project plans with learners, providing strategic scaffolding and understanding learners’ opinions in real time, gradually guiding learners to collaboratively complete the design of the project plan. For example, concerning the sub-task of “creating promotional materials and carriers,” multiple “Peer Agents” in group discussions can propose solutions in different promotional formats such as posters, web pages, or WeChat mini-programs. If learners propose to support the web format based on their interests and expertise, the “Teaching Assistant Agent” will select a “Peer Agent” with relevant capabilities to speak, engaging in the design and discussion of a “waste classification promotional website,” helping learners clarify how to design and build the promotional website. Subsequently, the “Teaching Assistant Agent” can broadcast the obtained plan within the group and select other “Peer Agents” to refine suggestions, such as first clarifying the rules of “waste classification” and displaying them prominently on the website.
(3) Collaborative Completion of Project Outputs
Based on the designed project plan, teaching agents can construct corresponding “Peer Agents” to collaboratively complete the production of project outputs with learners. The production of project outputs first requires the collection of relevant materials and information. For example, in the sub-task of “understanding waste classification rules,” learners need to collect the latest local waste classification standards. Since waste classification standards vary and change worldwide, the “Teaching Assistant Agent” can adopt retrieval-enhanced generation (RAG) methods to provide learners with accurate content generation.
As shown in the figure, “Peer Agents” utilize various functions provided by the open-source technology framework to quickly implement the RAG process. First, the “Index Establishment” step involves the agent automatically crawling or manually filtering resources from official government environmental department websites on the internet, collecting reliable information using document loading methods, and employing text segmentation methods to break long texts into semantically relevant short sentences. On this basis, the “Question Retrieval” step selects large model-extracted text feature vectors, storing them in a vector database to build a characteristic retrieval knowledge base for “waste classification standards.” Subsequently, using a question-answer-based retrieval method, it extracts text features from user questions and retrieves information most relevant to the question from the vector database based on feature similarity. Finally, the “Content Generation” step inputs the retrieved information and user question information into a prompt template, constructing complete prompt information, and utilizing large models to ultimately generate the latest and correct waste classification rules from various regions.
Applying Teaching Agents in Project-Based Learning
Once the relevant materials are collected, the “Teaching Assistant Agent” can further assist learners in creating a “waste classification promotional website.” During this process, learners can communicate with the agent in multimodal ways, displaying hand-drawn website front-end design drafts or discussing website back-end design concepts through text. The “Teaching Assistant Agent” can invoke multiple external web design scripting language libraries to automatically generate the corresponding web code. At the same time, the teaching agent can utilize a machine language execution environment to directly execute the generated code, providing feedback on execution results and error messages to the “Teaching Assistant Agent,” guiding it to further automatically modify and improve the code. Learners can also provide feedback on modifications based on the generated pages, through screenshots or natural language, allowing the “Teaching Assistant Agent” to further adjust and optimize the website based on the feedback content.
(4) Multi-role Evaluation of Project Outputs
During the project output presentation and evaluation phase, the “Teaching Assistant Agent” and “Peer Agents” can conduct evaluations of project outputs from their respective perspectives, providing teacher evaluations and peer evaluations. The agents generate corresponding process and outcome evaluation rubrics in advance based on personalized driving problems and project plans. During the learners’ presentation of the human-machine collaboratively produced project outputs, the “Teaching Assistant Agent” and “Peer Agents” can conduct formative evaluations of the presentation content based on the different process information stored in their memory modules and the corresponding evaluation rubrics from the perspectives of teachers and external peers. For example, regarding the “waste classification promotional website” project output, the agents can provide objective evaluations based on learners’ contributions during group discussions and website production processes. Additionally, the agents can utilize their environmental interaction capabilities to interactively test the website’s design through clicking access, quantitatively assessing the project output from aspects such as the number of elements on the webpage, color choices, and multimedia usage. Furthermore, agents can use their multimodal perception abilities to input learners’ presentation content in video form, evaluating aspects such as the fluency of language, logical coherence of content, and completeness of explanations in learners’ project presentations.
Based on the evaluation information from this round of project outputs, the “Teaching Assistant Agent” and “Peer Agents” can reflectively ask questions from the perspectives of learners’ knowledge mastery, skill acquisition, and interaction effectiveness, promoting the synchronous enhancement of their educational task planning, teaching, and interaction capabilities. Thus, in the next round of project-based learning, agents can more effectively conduct project-based learning under the same theme for new groups of learners, achieving the evolution of agents’ educational capabilities.

July 16 (Tuesday) 19:00, Zhongguancun Internet Education Innovation Center and Baidu Wenxin Agents jointly created the series planning of new forces in education “AI New Wind: How Do Agents Empower Education.”

Cutting-edge insights, in-depth exchanges, industry experts will jointly analyze the latest trends in the wind, anticipate the application prospects of education, and discuss the empowering role and practical value of AI agents in the field of education, painting the infinite possibilities of artificial intelligence empowering education together.

↓↓ Click to reserve the live broadcast ↓↓
Applying Teaching Agents in Project-Based Learning

Applying Teaching Agents in Project-Based Learning

Follow the official video account
Get more educational information
Applying Teaching Agents in Project-Based Learning
Applying Teaching Agents in Project-Based Learning

Applying Teaching Agents in Project-Based Learning

Source丨Excerpt from “China Electric Education”, original title “Research on the Construction and Application of Teaching Agents Based on Large Models“
Authors丨Lu Yu (Associate Professor at the Future Education High-Precision Innovation Center, Beijing Normal University), Yu Jinglei (Beijing Normal University), Chen Penghe (Lecturer at the Future Education High-Precision Innovation Center, Beijing Normal University)
Design丨Jia Chen
Review丨Yezi Xibei

Leave a Comment