Research On Named Entity Recognition Of Technology Project Notification

Science and technology project notification is issued by the government departments at all levels of the relevant scientific and technological projects to declare the work of the normative documents,but also the various scientific and technological projects to declare the basis for reporting units.At present,China attaches great importance to scientific and technological research,resulting in a large number of scientific and technological projects to notify,and its coverage is also varied.For science and technology project filers,there are two aspects of the problem:on the one hand,how to choose our own needs from a large number of scientific and technological projects,on the other hand,how a large number of scientific and technological projects to declare a quick archive to improve work efficiency.In order to solve these two problems,this paper puts forward the research content of named entity recognition.Named entity recognition is one of the important research contents extracted from the subject of information science.It is an important method to identify specific semantic entities(names,place names and various proper nouns)in the text.In information extraction,information retrieval,machine translation and other fields play a huge role.In this paper,we use the naming entity recognition method based on machine learning to identify the entities that are involved in the scientific research field.In this study,a large number of scientific and technological projects to declare notification of machine learning and automated analysis,to build a notice for the scientific and technological projects to declare the naming entity identification tool,and the proposed entity extraction model put into use,in order to help the cause of science and technology progress and development.This paper mainly has the following two aspects:1.The naming entity extraction model based on statistical named entity recognition method and conditional random field model is established.The average accuracy rate of the model is 89.84%,the average recall rate is 90.11%,and the performance is good,which lays a solid foundation for the follow-up model-based application.2.Based on the above work,we build the automated entity extraction tool,and use the tool to carry out the named entity recognition on the research field of the large number of scientific and technological projects.The research of this paper realizes the automatic interpretation of the notification of the declaration of science and technology projects,but there is still a lot of room for development in this field,and we need our continuous research and exploration.