Enhancing Data Analytics and Visualization in Engineering Education: A Transformative Approach Through LIDA GPT and Experimental Prompt Optimization Methods

Data visualization is a vital skill for data scientists, enabling the interpretation of complex data [1]. Educators must have a solid grasp of the subject to teach effectively [2]. For engineering students, mastering data visualization is both challenging and rewarding, with tools like <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{R}$</tex> and Python being essential [1] [2] [4] [5]. Creative data visualization enhances understanding and emotional engagement [1]. Traditional analysis can be complex for non-experts, making visualization a more accessible approach [1]. Integrating advanced LLMs with specialized frameworks is more effective than traditional methods [2]. Tools like LIDA can mitigate issues with general-purpose AI, though some API knowledge is needed [2] [3] [8]. Curricula should focus on developing creative data visualization skills and integrating traditional methods with AI tools like LIDA with GPT[3] [8]. Our goal is to evaluate the effectiveness of utilizing LIDA GPT for data analytics and visualization within engineering education[3]. The chat interface can be employed to detect and rectify errors, which enhances the quality of the visualizations over time. However, it is important to note that the code generated by LIDA GPT is not always flawless and may require additional refinement to achieve the final visualization. While LIDA GPT can efficiently process a single CSV file, the combination of multiple CSV files must be performed manually[7]. LIDA's performance relies on its training data's visualization knowledge[8] furthermore, conducting hypothesis testing with this tool can be challenging, as it requires more detailed information or fine tuning to yield accurate results[7] [8]. The paper employs a case study combined with regression analysis on CSV data[6] to examine the relationships between various skills.LIDA GPT can complement traditional teaching methods to improve creative data visualization skills, aiding in programming and data science learning through prompt optimizations.