A Text Classification Method Based Automobile Data Management

The state grid has a large number of assets, the physical part of these assets involves many categories, such as power equipment, real estate, vehicles and so on. The state grid needs to record and manage these assets, which involves the process of storing equipment and assets with specific IDs. However, because the type of equipment and assets are not uniformly coded across provinces, and various text descriptions vary from one to another, making this management and recording process is very difficult and complex. FastText is a text-based classifier, its N-gram model can read the sequential information between words and is faster than most other models, so we decide to use FastText to solve this problem, that is, through the classification of text to record various equipment and assets. It is suitable for the need for automatic identification of equipment and assets in this project. The purpose of this method is to ensure the correct correspondence with the type of equipment and assets according to the standard rules of technical object types of the state grid. FastText has a good classification accuracy, and so far, the method has achieved good results in the automotive category.