The paper by Martí A. Hearst explores the potential and challenges of text data mining (TDM), a field that has gained little traction despite its vast possibilities. Hearst distinguishes TDM from information access and computational linguistics, emphasizing that TDM involves discovering new information and patterns within large text collections, rather than simply retrieving or improving existing knowledge. He discusses the limitations of current approaches and highlights the need for exploratory data analysis in text data mining. Hearst provides examples of real TDM efforts, such as using text metadata to detect trends and patterns, and describes the LINDI project, which aims to support researchers in discovering new information through automated text manipulation and human-guided decision-making. The paper concludes by advocating for a blend of computationally driven and user-guided analysis to unlock the full potential of text data mining.The paper by Martí A. Hearst explores the potential and challenges of text data mining (TDM), a field that has gained little traction despite its vast possibilities. Hearst distinguishes TDM from information access and computational linguistics, emphasizing that TDM involves discovering new information and patterns within large text collections, rather than simply retrieving or improving existing knowledge. He discusses the limitations of current approaches and highlights the need for exploratory data analysis in text data mining. Hearst provides examples of real TDM efforts, such as using text metadata to detect trends and patterns, and describes the LINDI project, which aims to support researchers in discovering new information through automated text manipulation and human-guided decision-making. The paper concludes by advocating for a blend of computationally driven and user-guided analysis to unlock the full potential of text data mining.