Application method in military equipment text entity extraction based on transfer learning

Abstract

The invention provides an application method for military equipment text entity extraction based on transfer learning, which comprises the following steps: step 1, establishing a network model for boundary extraction and text fragment classification as a skeleton model for text entity extraction, and effectively overcoming the difference of network structures caused by different types of extractedentities in different fields; 2, analyzing source domain data, constructing a source domain problem set, and realizing task adaptation; 3, realizing domain adaptation by using a mask-based language model; and step 4, applying the model completing the domain adaptation and the task adaptation to the target domain, and completing the extraction of the military equipment text information. Accordingto the method, the difference of network structures caused by different entity extraction types in different fields is effectively overcome; according to the method, the existing open source sequenceannotation data is fully utilized, the named entity recognition model is trained on the basis, learned knowledge is applied to the target field, and the data annotation work in the target field is effectively reduced.

Description

technical field [0001] The invention relates to an application method in military equipment text entity extraction based on migration learning. Background technique [0002] Named Entity Recognition (NER), also known as "proper name recognition", refers to the recognition of entities with specific meanings in a text, mainly including person names, place names, institution names, proper nouns, etc. Simply put, it is to identify the boundaries and categories of entity references in natural texts. [0003] Named entity recognition is a very important basic task in the field of natural language processing research, and it is an important cornerstone of high-level tasks such as entity relationship extraction and event extraction. [0004] The military equipment test appraisal text refers to the highest national inspection behavior that conducts a comprehensive assessment of the subjects and draws evaluation conclusions through standardized organizational forms and test activitie...

AI Technical Summary

Problems solved by technology

[0005] Due to the particularity and professionalism of this field, the open corpus is very scarce, and the labeling process also requires multiple iterations: independent labeling, cross-checking, expert review, and modified iteration methods to ensure labeling consistency. It can be seen that the entire labeling process is very complicated. and cumbersome

Images