A new method for handling unstructured data in the High-speed Railway Passenger Service System

As the High-speed Railway develops rapidly, operational maintenance system plays a more significant role in ensuring safety of the High-speed Railway Passenger Service System. The traditional algorithms in literature dealing with data in an operational maintenance system are usually designed only for structured data. However, large amounts of unstructured data will be produced when a High-speed Railway Passenger Service System runs in the real world. Considering that unstructured data, usually consisting of natural languages, is not in a certain form, traditional algorithms cannot be directly used for handling unstructured data. However, the unstructured data in the form of natural language contains maintenance information of great importance. To address this problem, This paper first proposes a clustering algorithm on unstructured data, and then provides a foundation for fault diagnosis on unstructured data in the operational maintenance system.