I.-Chin Wu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
I.-Chin Wu has a total of 34 co-authors. Here are the top co-authors by number of shared publications:
- Pertti Vakkari
Faculty of Information Technology and Communication Sciences, Tampere University, Tampere, Finland
4 shared publications Yen-Yi Feng
Department of Emergency Medicine, Mackay Memorial Hospital, Taipei, Taiwan
4 shared publicationsTzu-Li Chen
National Taipei University of Technology
3 shared publicationsHsin-Kai Yu
National Taiwan Normal University
2 shared publicationsYi-An Chen
Fu-Jen Catholic University
2 shared publicationsYu-Ping Ho
Fu-Jen Catholic University
2 shared publications
Latest Publications
- 26 June 2026
- Published by Springer Nature
- p. 250-260
An Exploratory Study of Consumer Journey Segmentation Based on Mobile Shopping Behaviors
- 26 June 2026
- Published by Springer Nature
- p. 250-260
An Exploratory Study of Consumer Journey Segmentation Based on Mobile Shopping Behaviors
- Open Access
- 20 March 2026
- Published by University of Borås
- Vol. 31 (iConf) , 65-75
Leveraging search engines and LLMs for creative thinking: effects of prompting skills, and domain knowledge
Introduction. We investigated the effects of domain knowledge (DK) and prompting skills by leveraging a traditional search engine and a prompt-based LLM chatbot for a learning-oriented creative task. Method. On a simulated curated calligraphy exhibition task for which one must integrate DK and creativity to write summaries of curatorial planning, we observed behavior when using the NPM Collection, Google Search, and ChatGPT. We assessed the creative thinking demonstrated in the final summaries. User perception of tools supporting creative thinking was evaluated by the creativity support index. Analysis. We analysed information source usage behavior using descriptive statistical analysis and significance testing and used regression analysis to examine the predictive power of tool usage behavior, domain knowledge, and prompting skills for summary writing. Results. We discovered a positive predictive effect of DK and ChatGPT usage on the quality of curatorial planning summaries. Better prompting skills enhanced task performance and perceived creativity. For creative thinking, Google seemed to better foster exploration than ChatGPT. Conclusions. Prompt-based LLM chatbots support learning-oriented creative tasks. As prompting skills are essential, prompting should be integrated into information literacy education. The study provides implications for the design of search systems in human–information retrieval–LLM interaction to better support creative tasks.
- 20 March 2026
- Published by University of Borås
- Vol. 31 (iConf) , 65-75
Open AccessLeveraging search engines and LLMs for creative thinking: effects of prompting skills, and domain knowledge
Introduction. We investigated the effects of domain knowledge (DK) and prompting skills by leveraging a traditional search engine and a prompt-based LLM chatbot for a learning-oriented creative task. Method. On a simulated curated calligraphy exhibition task for which one must integrate DK and creativity to write summaries of curatorial planning, we observed behavior when using the NPM Collection, Google Search, and ChatGPT. We assessed the creative thinking demonstrated in the final summaries. User perception of tools supporting creative thinking was evaluated by the creativity support index. Analysis. We analysed information source usage behavior using descriptive statistical analysis and significance testing and used regression analysis to examine the predictive power of tool usage behavior, domain knowledge, and prompting skills for summary writing. Results. We discovered a positive predictive effect of DK and ChatGPT usage on the quality of curatorial planning summaries. Better prompting skills enhanced task performance and perceived creativity. For creative thinking, Google seemed to better foster exploration than ChatGPT. Conclusions. Prompt-based LLM chatbots support learning-oriented creative tasks. As prompting skills are essential, prompting should be integrated into information literacy education. The study provides implications for the design of search systems in human–information retrieval–LLM interaction to better support creative tasks.
- 11 June 2025
- Published by SAGE Publications in Intelligent Data Analysis: An International Journal
- Vol. 30 (2) , 451-477
Early prediction of hospital admission in the emergency department by generating expanded chief complaints
Early hospital admission prediction at the triage stage is an important and challenging task for emergency departments (EDs), aimed at effectively managing and utilizing limited medical resources for critical patients. A retrospective study was conducted at MacKay Memorial Hospital (MMH) from 2011 to 2018, including 1,061,760 records of valid patients, using logistic regression (LR), eXtreme Gradient Boosting (XGBoost), Word2Vec, and bidirectional encoder representations from transformers (BERT). The chief complaints (CCs) and limited structured variables collected at triage are considered predictor variables. The results show that XGBoost achieves better prediction than LR with patient structured variables and better prediction than Word2Vec with patient CCs in terms of AUC and F-measure. We further propose the novel concept of generating expanded CCs as BERT input by integrating the original CCs with selected structured variables using XGBoost to predict the probability of patient admission. Among the structured variables, triage category, mode of arrival, age, arrival time, and fever status are the most important. This study demonstrates BERT's (in particular, BERT-ROS with 5 variables) superior prediction capability compared to other models by considering only patient CCs or expanded CCs in terms of AUC and F-measure. Moreover, given the low admission rates in Taiwan's EDs, this study employs imbalanced data processing to show that the proposed method enhances the predictive capability of hospitalization. These experimental results provide a reference model with associated variables for developing a hospital admission tool at triage, identifying the risk of stratification of critical patients.
- 11 June 2025
- Published by SAGE Publications in Intelligent Data Analysis: An International Journal
- Vol. 30 (2) , 451-477
Early prediction of hospital admission in the emergency department by generating expanded chief complaints
Early hospital admission prediction at the triage stage is an important and challenging task for emergency departments (EDs), aimed at effectively managing and utilizing limited medical resources for critical patients. A retrospective study was conducted at MacKay Memorial Hospital (MMH) from 2011 to 2018, including 1,061,760 records of valid patients, using logistic regression (LR), eXtreme Gradient Boosting (XGBoost), Word2Vec, and bidirectional encoder representations from transformers (BERT). The chief complaints (CCs) and limited structured variables collected at triage are considered predictor variables. The results show that XGBoost achieves better prediction than LR with patient structured variables and better prediction than Word2Vec with patient CCs in terms of AUC and F-measure. We further propose the novel concept of generating expanded CCs as BERT input by integrating the original CCs with selected structured variables using XGBoost to predict the probability of patient admission. Among the structured variables, triage category, mode of arrival, age, arrival time, and fever status are the most important. This study demonstrates BERT's (in particular, BERT-ROS with 5 variables) superior prediction capability compared to other models by considering only patient CCs or expanded CCs in terms of AUC and F-measure. Moreover, given the low admission rates in Taiwan's EDs, this study employs imbalanced data processing to show that the proposed method enhances the predictive capability of hospitalization. These experimental results provide a reference model with associated variables for developing a hospital admission tool at triage, identifying the risk of stratification of critical patients.
- 26 May 2025
- Published by Springer Nature
- p. 365-377
Exploring How Muddled Information Needs Influence Learning Outcomes
- 26 May 2025
- Published by Springer Nature
- p. 365-377
Exploring How Muddled Information Needs Influence Learning Outcomes
- 12 September 2023
- Published by Emerald Publishing in Journal of Documentation
- Vol. 80 (2) , 298-319
An exploration of search-as-learning in digital archives of an online museum
Purpose: Recent studies on search-as-learning (SAL) have recognized the significance of identifying users' learning needs as they evolve for acquiring knowledge during the search process. In this study, the authors clarify the extent to which search behaviors reflect the learning outcome and foster the users' knowledge of Chinese art.Design/methodology/approach: The authors conducted an exploratory-sequential mixed-methods approach using simulated work task situations to collect empirical data. The authors used two types of simulated learning tasks for topics related to painting and antique knowledge. A lot of 25 users participated in this evaluation of digital archives (DAs) at the National Palace Museum (NPM) in Taiwan. For each set of topics, a close-ended task related to lower-level learning goals and an open-ended task related to higher-level learning goals.Findings: The learning criteria reflect changes in the users' knowledge structure, revealing the SAL process. Furthermore, users achieved better task performance on the higher-level creative-learning task, which suggests that they met more learning criteria, exhibited a greater variety of search patterns when exploring the topics via interaction with various sources. Finally, there is a close relationship between creative-learning tasks, prior knowledge, keyword search actions and learning outcomes.Originality/value: The authors discuss implications with respect to the design of DAs in practice and contributions to the body of SAL knowledge in DAs of online museums. For future reference, the authors provide implications for the development of learning measures from the perspective of user search behavior with associated learning outcomes in the context of DAs.
- 12 September 2023
- Published by Emerald Publishing in Journal of Documentation
- Vol. 80 (2) , 298-319
An exploration of search-as-learning in digital archives of an online museum
Purpose: Recent studies on search-as-learning (SAL) have recognized the significance of identifying users' learning needs as they evolve for acquiring knowledge during the search process. In this study, the authors clarify the extent to which search behaviors reflect the learning outcome and foster the users' knowledge of Chinese art.Design/methodology/approach: The authors conducted an exploratory-sequential mixed-methods approach using simulated work task situations to collect empirical data. The authors used two types of simulated learning tasks for topics related to painting and antique knowledge. A lot of 25 users participated in this evaluation of digital archives (DAs) at the National Palace Museum (NPM) in Taiwan. For each set of topics, a close-ended task related to lower-level learning goals and an open-ended task related to higher-level learning goals.Findings: The learning criteria reflect changes in the users' knowledge structure, revealing the SAL process. Furthermore, users achieved better task performance on the higher-level creative-learning task, which suggests that they met more learning criteria, exhibited a greater variety of search patterns when exploring the topics via interaction with various sources. Finally, there is a close relationship between creative-learning tasks, prior knowledge, keyword search actions and learning outcomes.Originality/value: The authors discuss implications with respect to the design of DAs in practice and contributions to the body of SAL knowledge in DAs of online museums. For future reference, the authors provide implications for the development of learning measures from the perspective of user search behavior with associated learning outcomes in the context of DAs.
- 17 July 2023
- Published by Springer Nature
- p. 302-313
Types of Mobile Retail Consumers’ Shopping Behaviors from the Perspective of Time
- 17 July 2023
- Published by Springer Nature
- p. 302-313
Types of Mobile Retail Consumers’ Shopping Behaviors from the Perspective of Time
- 16 June 2022
- Published by Springer Nature
- p. 233-244
Predicting Hospital Admission by Adding Chief Complaints Using Machine Learning Approach
- 16 June 2022
- Published by Springer Nature
- p. 233-244
Predicting Hospital Admission by Adding Chief Complaints Using Machine Learning Approach
- 13 October 2021
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 58 (1) , 380-389
Learning Outcomes during Information Search in Digital Archives
A museum's digital archive system gathers information about cultural heritage and makes it accessible to the public. In this study we clarify the extent to which search behaviors reflect task outcome and foster users’ knowledge of painting and calligraphy. Ten users participated who are special interest museum visitors. They joined in this evaluation of the Digital Archives of Chinese Painting and Calligraphy Search System (DA‐PCSS) of the National Palace Museum in Taiwan. Participants’ search activities and interactions with the DA‐PCSS were recorded in two simulated tasks. The results show that the high‐performance (HP) group who received high scores for their essays on the tasks formulated precise and relevant queries. Furthermore, the HP group were able to find information needed inside or outside the system to explore and synthesize the information, which was clearly reflected in their search move patterns, during the search process. Our results suggest that the criteria for learning at various stages of search suggested by Vakkari (2016) seem to validly reflect the quality of the search outcomes. In all, the results elucidate how the evaluated system supports users as they search for target items, as well as how learning occurs during the search process and in turn influences task outcomes.
- 13 October 2021
- Published by Wiley in Proceedings of the Association for Information Science and Technology
- Vol. 58 (1) , 380-389
Learning Outcomes during Information Search in Digital Archives
A museum's digital archive system gathers information about cultural heritage and makes it accessible to the public. In this study we clarify the extent to which search behaviors reflect task outcome and foster users’ knowledge of painting and calligraphy. Ten users participated who are special interest museum visitors. They joined in this evaluation of the Digital Archives of Chinese Painting and Calligraphy Search System (DA‐PCSS) of the National Palace Museum in Taiwan. Participants’ search activities and interactions with the DA‐PCSS were recorded in two simulated tasks. The results show that the high‐performance (HP) group who received high scores for their essays on the tasks formulated precise and relevant queries. Furthermore, the HP group were able to find information needed inside or outside the system to explore and synthesize the information, which was clearly reflected in their search move patterns, during the search process. Our results suggest that the criteria for learning at various stages of search suggested by Vakkari (2016) seem to validly reflect the quality of the search outcomes. In all, the results elucidate how the evaluated system supports users as they search for target items, as well as how learning occurs during the search process and in turn influences task outcomes.
- 13 March 2019
- Published by Springer Nature
- p. 554-561
Effects of the User-Need State for Online Shopping: Analyzing Search Patterns
- 13 March 2019
- Published by Springer Nature
- p. 554-561
Effects of the User-Need State for Online Shopping: Analyzing Search Patterns