The linear and text-laden design of PowerPoint presentations does not align with the brain’s natural way of processing information through interconnected concepts. In an effort to foster higher-order thinking and improve comprehension, this research introduces BoomMapper, an AI-powered system designed to automatically convert PowerPoint slides into hierarchical mind maps, providing a visually engaging and intuitive alternative for knowledge representation. BoomMapper leverages Named Entity Recognition (NER), TF-IDF ranking, and n-gram extraction, as well as models such as BERT, RoBERTa, SpaCy, and OpenAI GPT-3.5-turbo for content extraction and structuring. The study involved an expert evaluation with 2 participants and user acceptance testing (UAT) with 32 users to assess the system’s effectiveness. The system’s evaluation centered on expert validation to assess the quality, accuracy, and coherence of the generated mind maps. Experts reviewed the outputs and identified OpenAI GPT-3.5-turbo as the most effective model for generating accurate and coherent mind maps. The evaluation highlighted BoomMapper’s practicality and potential for real-world applications while suggesting future enhancements, such as multilingual support and media integration, to further improve its usability and impact. This study underscores the transformative potential of AI in enhancing learning and knowledge representation. By automating mind map generation, BoomMapper offers a scalable solution to bridge gaps in traditional presentation methods.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

BoomMapper: An Innovative Web-Based System for Transforming PowerPoint Presentations into Mind Maps

  • Coey Low,
  • Vinothini Kasinathan,
  • Aida Mustapha

摘要

The linear and text-laden design of PowerPoint presentations does not align with the brain’s natural way of processing information through interconnected concepts. In an effort to foster higher-order thinking and improve comprehension, this research introduces BoomMapper, an AI-powered system designed to automatically convert PowerPoint slides into hierarchical mind maps, providing a visually engaging and intuitive alternative for knowledge representation. BoomMapper leverages Named Entity Recognition (NER), TF-IDF ranking, and n-gram extraction, as well as models such as BERT, RoBERTa, SpaCy, and OpenAI GPT-3.5-turbo for content extraction and structuring. The study involved an expert evaluation with 2 participants and user acceptance testing (UAT) with 32 users to assess the system’s effectiveness. The system’s evaluation centered on expert validation to assess the quality, accuracy, and coherence of the generated mind maps. Experts reviewed the outputs and identified OpenAI GPT-3.5-turbo as the most effective model for generating accurate and coherent mind maps. The evaluation highlighted BoomMapper’s practicality and potential for real-world applications while suggesting future enhancements, such as multilingual support and media integration, to further improve its usability and impact. This study underscores the transformative potential of AI in enhancing learning and knowledge representation. By automating mind map generation, BoomMapper offers a scalable solution to bridge gaps in traditional presentation methods.