INTEGRATING CHATGPT INTO BIG DATA EDUCATION: A STUDENT-CENTERED TEACHING FRAMEWORK

SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 4 ISSUE 12 DECEMBER 2025 ISSN: 2181-3337 | SCIENTISTS.UZ 150 INTEGRATING CHATGPT INTO BIG DATA EDUCATION: A STUDENT-CENTERED TEACHING FRAMEWORK Seungjik Lee Professor, Artificial Intelligence, Tashkent University of Information Technologiesnamed after Muhammad al-Khwarizmi https://doi.org/10.5281/zenodo.18015676 Abstract. This study presents a structured, student-centered pedagogical framework for teaching Big Data concepts using ChatGPT as an integrated educational assistant. A semesterlong implementation in an undergraduate Big Data course at TUIT demonstrated that guided ChatGPT usage, coupled with incremental project-based tasks, collaborative discussion, and reflective evaluation, improves students’ autonomy, engagement, and technical competence. Findings suggest that this AI-supported model fosters independent learning while maintaining critical evaluation skills and peer collaboration, providing a replicable approach for integrating generative AI into higher education curricula. Keywords: big data education, ChatGPT, Student-Centered Learning, Project-Based Learning, AI in Education. I. Introduction The demand for Big Data skills has increased significantly due to the growth of dataintensive technologies and AI applications. Traditional lecture-based models often fail to cultivate applied learning and self-directed problem-solving, which are essential for mastering tools such as Python, SQL, and data visualization frameworks. Generative AI models like ChatGPT provide immediate guidance, code generation, and conceptual explanations, enabling students to explore complex topics independently. This study introduces a student-centered teaching framework integrating ChatGPT into a Big Data curriculum. The framework emphasizes small, incremental tasks, guided AI-assisted problem solving, group discussion, and concise summarization. The objectives are to enhance students’ autonomy, technical competence, collaborative skills, and confidence while maintaining rigorous evaluation of AI outputs. Research questions include: 1. How does ChatGPT-assisted learning affect students’ independent problemsolving? 2. How does group discussion and summarization improve comprehension and collaboration? 3. What impact does AI-assisted instruction have on technical competence and project outcomes? II. Literature Review Prior research demonstrates the benefits and challenges of integrating AI in education. Hao et al. [1] show that large language models enhance programming comprehension through immediate feedback and error correction. Project-based learning frameworks improve student engagement and knowledge retention [2]. Recent studies in Big Data education highlight students’ difficulties with fragmented curricula spanning programming, statistics, and databases [3-4]. SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 4 ISSUE 12 DECEMBER 2025 ISSN: 2181-3337 | SCIENTISTS.UZ 151 Although experimental papers on AI and IoT have been published [5], there are not many papers on student-centered education using AI apps. Generative AI tools such as ChatGPT can scaffold learning by providing tailored explanations, example code, and iterative guidance. Concerns exist regarding overreliance on AI outputs, misinformation, and reduced creative problem-solving. The proposed framework addresses these issues by requiring students to critically evaluate AI output, engage in group discussions, and summarize key findings. III. Methodology A. Course Context The framework was implemented in an undergraduate Big Data course at TUIT spanning one semester with approximately 25 students, organized into groups of 3–4. The course consisted of seven sequential modules: understanding Big Data concepts, Python programming, Excel-style table creation, data manipulation, SQLite introduction, SQL querying, and data visualization. B. Python-Based Table Creation and Integration Each student created an individual Excel-style table using pandas and openpyxl. Students then collaborated to merge individual tables into a single group dataset using pandas.concat() and pandas.merge(). This stage emphasized schema alignment, handling missing values, and collaborative decision-making. Sample Python Table Creation: import pandas as pd data = {'Student': ['Alice', 'Bob'], 'Score': [85, 90]} df = pd.DataFrame(data) df.to_excel('student_scores.xlsx', index=False) Table Integration Example: df1 = pd.read_excel('student_scores.xlsx') df2 = pd.read_excel('student_projects.xlsx') merged_df = pd.concat([df1, df2]) merged_df.to_excel('group_merged.xlsx', index=False) C. Data Manipulation in Python Students performed Python-based manipulations simulating SQL operations: adding rows/columns, updating records, filtering, and deleting entries. Example: merged_df.loc[merged_df['Student'] == 'Alice', 'Score'] = 95 merged_df.loc[len(merged_df)] = ['Charlie', 88] merged_df = merged_df[merged_df['Score'] >= 90] D. SQLite and SQL Queries Students gradually learned the fundamentals of database theory and SQL, from commands like SELECT to joins, subqueries, and data modification, by importing tables and executing queries with SQLite. Example: import sqlite3 conn = sqlite3.connect('bigdata.db') merged_df.to_sql('scores', conn, if_exists='replace', index=False) query = "SELECT Student, Score FROM scores WHERE Score > 90" high_scores = pd.read_sql(query, conn) SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 4 ISSUE 12 DECEMBER 2025 ISSN: 2181-3337 | SCIENTISTS.UZ 152 Table 1. Stepwise SQL Learning: Stage Topics & Commands Learning Outcome 1 SELECT * FROM sales; Basic retrieval 2 ORDER BY Sorting 3 Expressions Computed columns 4 Aggregation & GROUP BY Summary 5 Joins Multi-table queries 6 Subqueries & views Modular queries 7 INSERT, UPDATE, DELETE Data modification E. Data Visualization Students used matplotlib for bar charts, line graphs, histograms, and scatter plots. Visualizations emphasized clarity, labeling, and aesthetics. Example: import matplotlib.pyplot as plt plt.bar(merged_df['Student'], merged_df['Score']) plt.xlabel('Student') plt.ylabel('Score') plt.title('Student Scores') plt.show() F. Final Project Each student created a dataset from a different domain. Groups merged these tables, resolved schema conflicts, and produced a comprehensive multi-domain dataset. G. Final Presentation and Peer Evaluation Each group presented visualizations on a single monitor. A representative from each group presented their integrated decisions and insights to the other group. The other group used a structured form to evaluate and provide constructive feedback. H. Workflow Figure Figure 1: Workflow of ChatGPT-Assisted Student-Centered Learning SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL VOLUME 4 ISSUE 12 DECEMBER 2025 ISSN: 2181-3337 | SCIENTISTS.UZ 153 Figure 1 illustrates the process of this workflow. Python Table Creation → Table Integration → Data Manipulation → SQLite Import & SQL Queries → Data Visualization → Final Project & Presentation → Peer Evaluation IV. Results Through this framework, students develop independent problem-solving skills. They also discuss their research findings with other students, draw final conclusions, and present them to other groups. The framework improved:  Independent Problem-Solving  Collaboration  Technical Competence  Visualization Skills  Critical Evaluation Table 2: Sample Evaluation Metrics Metric Description Score Range Score Independence Task completion without instructor 0–10 Collaboration Peer discussion quality 0–10 Technical Competence Code, queries, visualization accuracy 0–10 Engagement Participation 0–10 V. Discussion The AI-based project-based framework proposed in this study helped cultivate students' autonomy, collaboration, and technical skills. Students organically applied Python, SQL, and visualization techniques, and improved their analytical thinking skills through critical evaluation of AI results. VI. Conclusion The student-centered model that integrates ChatGPT into big data education has been shown to improve self-directed learning, technical proficiency, and communication skills. The educational effectiveness of this framework is superior to traditional teaching methods. Future research suggests that it could be applied to other subjects, such as AI, IoT, and image processing, with scalability and quantitative evaluation. We will continue to explore applications in these areas. REFERENCES 1. [Hao, J., et al. “AI-Assisted Programming Education: A Study of LLMs in the Classroom.” J. Comput. Sci. Educ., 2023. 2. Thomas, J. “Project-Based Learning in Higher Education.” Learn. Res. Rev., 2020. 3. Kim, H. “Challenges in Big Data Pedagogy.” Int. J. Data Educ., 2024. 4. Rahman, A., Liu, S. “Generative AI in Data Science Learning.” Data Pedagogy Quart., 2023. 5. Seungjik Lee, “AI AND IoT PRACTICAL EDUCATION”, SCIENCE AND INNOVATION INTERNATIONAL SCIENTIFIC JOURNAL, VOLUME 3 ISSUE 5, MAY 2024 6. OpenAI. “ChatGPT Technical Report,” 2023.