Chinese negative information recognition based on deep learning

Abstract Chinese negative information recognition is the key technology of natural language processing. Its core is to recognize the negative relationship in the text according to the negative information theory. The recognition and classification of negative information is an important research hotspot of natural language processing. It is of great significance to the establishment of big data around Chinese semantic recognition, such as information retrieval and text mining. It is also a key step to promote Chinese language recognition. Firstly, this paper studies the expression forms of Chinese negative information in the scene of life, and points out that negative information recognition mainly includes negative trigger word recognition, negative coverage recognition and negative focus recognition; Then, it summarizes the current basic knowledge of deep learning and Chinese negative information recognition, analyzes the working principle of CRF and recurrent neural network RNN model, and analyzes the basic theories of Chinese information recognition, such as Chinese corpus data and text preprocessing technology; Finally, the combination of bidirectional long-term and short-term memory network (bltstmn) and traditional CRF model is used to study Chinese negative information recognition. Theoretical research shows that the combination of bltstmn-crf has better applicability to Chinese negative information recognition, and has better improvement in recognition accuracy and fast recognition. In this paper, the recognition of Chinese negative information based on bltstmn-crf is a form of applying deep learning to natural language processing, which can provide a thinking reference for other deep learning networks in natural language processing. In the follow-up, we will further study the application of deep learning to natural language processing.