Lexical Profile of AI-Generated Content in Higher Education

This book investigates vocabulary use in GenAI texts to understand the nature of the language being generated as potential reading material in university settings. Twelve AI-generated samples of writing covering arts, commerce, law and science are analysed for vocabulary use and compared with published research articles on the same topic.

This book investigates vocabulary use in texts generated by Artificial Intelligence. The growing use of AI tools such as ChatGPT and DeepSeek in University settings indicates the importance of evaluating the products of such technology as potential reading material. In the study reported, twelve AI-generated samples of writing covering arts, commerce, law and science are analysed for vocabulary use, and compared with four research articles published in journals on the same topics. The vocabulary profiling covers high-frequency words, academic and technical vocabulary, most frequently used content words and low-frequency words. Potential future developments in AI are explored.

Autorentext
David Hirsh is Associate Professor at the University of Sydney. His research explores vocabulary use, bilingual education and language revitalisation.

Inhalt

List of Tables - List of Figures - 1. Introduction - 1.1. The GenAI Era - 1.2. GenAI in Higher Education - 1.3. Vocabulary Profiling - 1.4. Vocabulary Lists - 1.5. Anticipations of the Future - 2. GenAI in Higher Education - 2.1. Artificial Intelligence - 2.2. Generative Artificial Intelligence - 2.2.1. Machine Learning - 2.2.2. Natural Language Processing - 2.2.3 Large Language Models - 2.2.4. Datasets - 2.2.5. Algorithms - 2.3. GenAI Tools in Higher Education - 3. Vocabulary Profiling - 3.1. Lexical Nature of Academic Writing - 3.1.1. High Frequency Words - 3.1.2. Academic Words - 3.1.3. Technical Words - 3.1.4. Proper Nouns - 3.1.5. Low Frequency Words - 3.2. K1, K2 and AWL Word Lists - 3.3. BNC-COCA Word Lists - 4. The Current Study - 4.1. Selection of GenAI Tools - 4.2. Selection of Prompts for GenAI Text Generation - 4.3. Analysis of Texts - 5. Analysis 1: K1, K2 and AWL Words - 5.1. Arts Texts - 5.1.1. Non-GenAI - 5.1.2. GenAI Scite - 5.1.3. GenAI Jenni - 5.1.4. GenAI Yomu - 5.1.5. Comparison of Arts Texts - 5.2. Commerce Texts - 5.2.1. Non-GenAI -5.2.2. GenAI Scite - 5.2.3. GenAI Jenni - 5.2.4. GenAI Yomu - 5.2.5. Comparison of Commerce Texts - 5.3. Law Texts - 5.3.1. Non-GenAI - 5.3.2. Non-GenAI Scite - 5.3.3. Non-GenAI Jenni - 5.3.4. Non-GenAI Yomu - 5.3.5. Comparison of Law Texts - 5.4. Science Texts - 5.4.1. Non-GenAI - 5.4.2. Non-GenAI Scite - 5.4.3. Non-GenAI Jenni - 5.4.4. Non-GenAI Yomu - 5.4.5. Comparison of Science Texts - 6. Analysis 2: K1 to K10 Words - 6.1. Arts Texts - 6.2. Commerce Texts - 6.3. Law Texts - 6.4. Science Texts - 7. Analysis 3: 20 Most Frequently Occurring Content Words - 7.1. Arts Texts - 7.2. Commerce Texts -7.3. Law Texts - 7.4. Science Texts - 8. Analysis 4: Low Frequency Word Use - 8.1. Arts Texts - 8.2. Commerce Texts - 8.3. Law Texts - 8.4. Science Texts - 9. Conclusion - 9.1. Use of GSL K1, GSL K2 and AWL Words - 9.2. Use of BNC-COCA K1 to K10 Words - 9.3. 20 Most Frequently Occurring Content Words - 9.4. Use of Low Frequency Words - 9.5. Final Comments - References - Appendices - Index