Hierarchical Attention Networks for Grid Text Classification
Yunxiang Zhang, Zhuyi Rao · 2020
With the development and popularization of power Internet of Things (IoT), IoT produces a large number of distributed data, the electronic text data and the concurrency of multiple services. To support unified management of all data on an existing basis and make the system has the ability to accurate and efficient data analysis, many researchers have devoted efforts to studying an effective algorithms for solving the problem of the electronic text retrieval, information extraction, and classification in the power grid system. However, the accuracy of these algorithms is not satisfactory. In our paper, we will use an effective deep learning model to classify the grid text. i.e., the hierarchical attention networks model. In particular, Attention mechanism is a commonly model with long-term memory mechanism in the field of NLP, which can intuitively evaluate the contribution of each word to the result in a text. Experimental results show that the Hierarchical Attention Networks model obtains the highest accuracy than the existing methods.