Applications of ChatGPT in Higher Education, Research and Development

ChatGPT in Text processing and analysis

 3.10. Text processing and analysis

ChatGPT is a powerful and innovative AI model that can enhance the experience and outcomes of text processors and writers who are seeking text processing and writing. By providing information, advice, feedback, and support in a conversational way, ChatGPT can make text processing and writing more accessible, efficient, and enjoyable for everyone. ChatGPT can also be used in text processing and natural language processing (NLP) tasks, such as text classification, sentiment analysis, and language translation (figure 41). ChatGPT can be fine-tuned for specific NLP tasks and generate high-quality output, based on its understanding of the language and context. This can help to improve the accuracy of text processing and NLP tasks and reduce the workload of human annotators.

Figure 41: ChatGPT's applications in text processing and analysis.

3.10.1 Text processing

ChatGPT can provide information about different data and text processing tasks, methods, tools, or resources available, such as data or text sources, formats, standards, algorithms, libraries, frameworks, etc. ChatGPT can also answer common questions and clarify doubts that data and text processors may have about these tasks, methods, tools, or resources. ChatGPT can help them find the best tasks, methods, tools, or resources for their data and text processing, suggest ways to improve their data or text quality, efficiency, or usefulness, and prepare them for feedback and evaluation. ChatGPT can provide positive reinforcement, emotional support, and inspiration to data and text processors who may face challenges, difficulties, or errors along the way.

3.10.2 Sentiment analysis

ChatGPT can be used to analyze large amounts of text data, such as social media posts, emails, and to customer feedback, and generate insights and predictions about user behavior and preferences. This can help improve the quality of business decisions and ensure a more personalized experience for users. ChatGPT can analyse large amounts of text data to obtain useful insights. It can identify patterns, themes, and trends in data and make informed decisions. This is useful for research in fields such as psychology, marketing, and social science, where researchers need to analyse large amounts of text data. ChatGPT is used for predictive modeling, classification, and clustering tasks [44]. It can learn data patterns and predict future results. This is useful in a variety of scientific fields, including finance, biology, and engineering.

3.10.3 Text classification

Text classification is a technique that assigns a text to one or more categories or labels based on its content or context [45]. Text classification can be performed by predefined rules or criteria (rule-based classification) or by using a machine learning model that learns from labeling examples (supervised classification). Text classification can be used for various purposes, such as spam filtering, detecting hate speech, organizing documents by topic, and sentiment analysis.

3.10.4 Text extraction

Text extraction is a technique that extracts specific information from a text based on certain criteria and rules [46]. Text extraction can be performed with regular expressions and patterns (pattern-based extraction) or with machine learning models (supervised extraction) to learn from label examples. Researchers can extract data using ChatGPT by asking ChatGPT to search or summarize specific information from specific data sources. Faculty can generate text for sections of IRB protocols, such as narratives, data management plans, project management, consent forms, recruitment letters, and confidentiality statements for research associates.

3.10.5 Text generation

Another important application of ChatGPT in text processing and writing is in the generation of high-quality text content [47]. ChatGPT is a chatbot that generates text and images based on user input using natural language processing and pretraining transformation models. ChatGPT can be fine-tuned on specific text generation tasks, such as article writing, email writing, and social media post writing, and generate high-quality outputs based on its understanding of the language and context. This can help reduce the workload of writers and ensure more accurate and engaging text content. ChatGPT application includes text editing, summarization, translation, and transformation. ChatGPT can be used to edit text, such as correcting grammar and spelling errors, summarize text, such as news articles and research papers, and translate text from one language to another.

3.10.6 Image captioning

Image captioning is a task that generates text descriptions of images such as objects, actions, and scenes [48]. Image captioning can be done manually by writing or speaking captions based on human observations and interpretation, or by automatically using tools and algorithms that can analyse images using computer vision and natural language processing techniques and generate captions based on certain criteria or rules. ChatGPT can generate image captions using natural language. It is useful in fields such as computer vision and robotics, where researchers need to analyse and interpret visual data.

 Overall, the use of ChatGPT in text processing and writing has the potential to revolutionize the way writers and businesses work with text. By providing a powerful natural language interface for text processing and writing data and models, ChatGPT can help to improve the efficiency of the text processing and writing support process, provide a better experience for the writers, and ensure more accurate and engaging text content. As the technology continues to develop, we can expect to see even more exciting applications of ChatGPT in text processing and writing.