Statistics of Textractor: A Framework for Extracting Relevant Domain Concepts from Irregular Corporate Textual Datasets