Research On Named Entity Recognition And Disambiguation For Short Text

With the rapid development of information technology,there is more and more text information on the internet,and people's demand for natural language processing is also growing.In recent years,with the rapid spread of mobile phones and other devices and the rapid development of mobile networks,people frequently use short text information exchange tools such as Weibo.How to extract valuable information from short text information is an urgent problem to be solved.Named entity recognition and disambiguation are the basis of a series of text processing techniques.This paper revolves around how to improve the research on named entity recognition and disambiguation for short text,and does the following research work:(1)Based on the self-attention deep learning method,a SelfAttention-BiLSTM-CRF neural network model was constructed to identify short-text named entities.A self-attention mechanism is introduced to make the model more capable of capturing context-related internal information without correlating external data,to better understand the sentence structure and fully exploit the features contained in the text itself.This method does not require manual design of features,and can automatically complete the end-to-end named entity recognition in the text.Experimental results show that this model has a high recognition accuracy.(2)Propose a short text entity disambiguation method based on multi-feature fusion.According to the reference of the entity to be disambiguated,the Baidu Encyclopedia is used to obtain the set of candidate entities,and extractis multiple features of the object to be disambiguated entity and the candidate entity for comparison,adopts a linear weighting method for fusion,and finds the entity with the highest score as the link target from the candidate entity set,complete disambiguation.The experimental results show that this method has a good effect on the entity disambiguation of short text.