The machining processes on the advanced machining workshop floor are becoming more sophisticated with the interdependent intrinsic processes, generation of ever-increasing in-process data and machining domain knowledge. To manage and utilize those above effectively, an industrial dataspace for machining workshop (IDMW) is presented with a three-layer framework. The IDMW architecture is Schema Centralized-Data Distributed, which relies on Process-Workpiece-Centric knowledge schema description and data storage in decentralized data silos. Subsequently, the pre-processing method for the data silos driven by RFID event graphical deduction model is elaborated to associate decentralized data with knowledge schema. Furthermore, through two industrial case studies, it is found that IDMW is effective in managing heterogeneous data, interconnecting the resource entities, handling domain knowledge, and thereby enabling machining operations control on the machining workshop floor particularly.