Reimagining Textbook Learning: An Interactive AI Tutor Approach Using Retrieval-Augmented Generation

ACL ARR 2025 July Submission660 Authors

Abstract: We present an Interactive AI Tutor designed to make textbook learning more accessible, adaptive, and engaging. Addressing the limitations of static educational resources, the system transforms textbook chapters into dynamic learning experiences using retrieval-augmented generation, interactive challenges, and narrative-based instruction. Initially developed using DeepSeek Coder-6.7B for question-answer-generation, later optimized with Mistral-7B, and further adapted for deployment using Falcon-RW-1B on resource-constrained platforms such as Google Colab, the system integrates LangChain pipelines, FAISS retrieval, and Google APIs. It supports modular learning modes, including storytelling, business simulations, and quizzes, with real-time progress tracking. Our findings demonstrate the feasibility of deploying lightweight yet interactive AI systems to personalize learning from open educational resources (OER) at scale.

Paper Type: Long

Research Area: Human-Centered NLP

Research Area Keywords: Intelligent tutoring systems, retrieval-augmented generation, human-centered NLP, educational technology

Contribution Types: Model analysis & interpretability, NLP engineering experiment, Approaches to low-resource settings, Publicly available software and/or pre-trained models, Data resources, Data analysis

Languages Studied: English

Previous URL: https://openreview.net/forum?id=RB3MaSgod2

Reassignment Request Area Chair: No, I want the same area chair from our previous submission (subject to their availability).

Reassignment Request Reviewers: No, I want the same set of reviewers from our previous submission (subject to their availability)

A1 Limitations Section: This paper has a limitations section.

A2 Potential Risks: Yes

A2 Elaboration: See Section 8 – discusses limitations including deployment, accessibility, and model reliability.

B Use Or Create Scientific Artifacts: Yes

B1 Cite Creators Of Artifacts: Yes

B1 Elaboration: See Section 4.1, 4.2, and References

B2 Discuss The License For Artifacts: Yes

B2 Elaboration: See References, licensing mentioned in software used

B3 Artifact Use Consistent With Intended Use: Yes

B3 Elaboration: See Section 3.1 and 4.4

B4 Data Contains Personally Identifying Info Or Offensive Content: Yes

B4 Elaboration: See Section 5 (IRB approved, no personal info collected)

B5 Documentation Of Artifacts: Yes

B5 Elaboration: Dataset and guides are described in Sections 3.1 and 3.3

B6 Statistics For Data: Yes

B6 Elaboration: Dataset stats in Section 3.1

C Computational Experiments: Yes

C1 Model Size And Budget: Yes

C1 Elaboration: See Section 4.1

C2 Experimental Setup And Hyperparameters: Yes

C2 Elaboration: See Section 4.1

C3 Descriptive Statistics: Yes

C3 Elaboration: See Section 6

C4 Parameters For Packages: Yes

C4 Elaboration: See Section 4.1–4.3

D Human Subjects Including Annotators: Yes

D1 Instructions Given To Participants: Yes

D1 Elaboration: See Section 3.3

D2 Recruitment And Payment: Yes

D2 Elaboration: See Section 5

D3 Data Consent: Yes

D3 Elaboration: See Section 5

D4 Ethics Review Board Approval: Yes

D4 Elaboration: See Section 5 (IRB2025-0182)

D5 Characteristics Of Annotators: Yes

D5 Elaboration: See Table 4

E Ai Assistants In Research Or Writing: Yes

E1 Information About Use Of Ai Assistants: Yes

E1 Elaboration: Used DeepSeek Coder for content generation — See Section 4.3

Author Submission Checklist: yes

Submission Number: 660