In response to the optimization problem of text entity information extraction,this article takes the leasing industry as the research object.Firstly,crawler technology is used to crawl the information published by customers,and the BiLSTM-CRF algorithm is used to extract and process the entity information.The processed information is stored in a database,and a data layer for the App's data source is constructed.Then,an App application layer is developed based on the data from the data layer.The developed App application layer module includes a user authentication module and a homepage module.The BiLSTM-CRF model outperforms LSTM and BiLSTM in entity boundary recognition,with model accuracy,recall,and F1 values reaching 96.58%,88.94%,and 92.60%,respectively.
维普
万方数据