Date & Time: Friday, October 3, 2025, 13:00–16:00
Venue: College of Management, National Sun Yat-sen University, Room CM3022
The final workshop of the 2025 “Text Mining and Digital Transformation Service Industry-Academia Alliance” was the Tarflow Advanced Workshop, designed to provide participants with a deeper understanding of practical processes and applications of text mining. The session combined theoretical introductions with hands-on demonstrations and was conducted both onsite and online. To enhance the learning experience, three teaching assistants (two onsite and one online) were arranged to support participants with practice exercises and troubleshooting.
The workshop opened with remarks by the Alliance’s Principal Investigator, Professor San-Yi Huang, who highlighted that this was the last workshop of 2025. The program covered data collection, preprocessing, and various analysis methods. As some new members joined the Alliance for the first time, the course began with the basics and gradually guided participants to become familiar with the Tarflow platform, ultimately applying it to their own research or project topics.
The main session was led by Alliance team member Chia-Ming Chang, Ph.D. candidate, who guided participants through three key modules: data preprocessing, sentiment analysis, and topic modeling. He first introduced the advantages of using the Tarflow text analysis workflow platform, including its drag-and-drop interface, reusable module design, and support for both Chinese and English text, which enables users to quickly adapt to practical use cases. Practical case studies and exercises were also incorporated into the class, such as cleaning and extracting topics from news datasets, helping participants understand the complete process from data acquisition to analytical output.
Toward the end of the workshop, a networking and discussion session was arranged, allowing members to share their experiences and insights. Through this workshop, participants learned how to apply text analysis tools to public opinion studies and research applications. The course received enthusiastic feedback, with active exchanges reflecting strong engagement from participants.
Looking ahead, the Alliance will continue to promote workshops and events on related topics, supporting members in applying text analysis technologies to decision-making processes and innovative services.