Named Entity Recognition Model of Power Equipment Based on Multi-feature Fusion

Abstract

AbstractExtracting useful information from a large number of text files in the power field is of great significance to power informatization, and the identification of power equipment entities is a key part. Aiming at the difficulties of entity recognition of power equipment in the field of Chinese electric power, such as complex entity names and difficult identification of rare entities, this paper proposes a Chinese named entity recognition model based on multi-feature fusion. From the knowledge of the electric power field (concise dictionary of electric technical terms, English dictionary of electric power terms, etc.), a large number of electric power professional terms are sorted out to construct the electric power field dictionary, and then text segmentation and part-of-speech tagging are carried out under the guidance of it. Integrate various features of characters, words and word categories into input vectors and input them into the BiLSTM-CRF model for sequence labeling. The experimental results show that the entity recognition model proposed in this paper improves the recognition effect of Chinese named entities in the field of power equipment.KeywordsPower equipmentChinese named entity recognitionDomain dictionaryDeep learning