On December 5, 2024, National Sun Yat-sen University's accounting course hosted a special lecture titled "Text Mining in Accounting," focusing on the application and potential of text mining in the field of accounting. Andrew Argue led the session, covering fundamental concepts, analytical techniques, and practical applications of text mining.

Andrew began by introducing the concept of text mining and its role as a component of artificial intelligence (AI). He outlined the four key steps in processing unstructured textual data, emphasising the importance of text preprocessing. Techniques such as tokenization, stop word filtering, and stemming were explained to demonstrate how valuable insights can be extracted from disorganized text, forming the basis for further analysis. The lecture also covered various methods for transforming and analyzing text, showcasing how approaches like the Bag of Words and vector space models can convert textual data for applications such as classification, clustering, and information extraction. For instance, sentiment analysis techniques were highlighted as tools for detecting tone variations in financial reports, which can aid in predicting corporate performance or identifying risk factors.

Andrew concluded by explaining how text mining can improve the readability of financial reports and assist in detecting fraud risks and abnormal disclosures. This lecture was particularly valuable for students in the College of Management, providing them with insights into analysing companies from a non-numerical perspective.