A method and device for tracing official document information of government organizations based on convolutional neural network

Abstract

The invention provides a method and device for tracing government organization official document information based on a convolutional neural network. The method includes the steps of: establishing several top-down catalogs of government documents at all levels; obtaining the text to be traced, and using all characters of the text to be traced as target characters; Match the titles and texts of official documents of government organizations at all levels; form a traceable subset including the titles and matching paragraphs of official documents of target government organizations; display the above traceable subsets on electronic devices as a relationship diagram. Targeted solution to the document characteristics of government organization documents improves the efficiency of tracing government organization document information; adopts neural network technology, especially feed-forward convolutional neural network technology, and has a high matching accuracy for target text and official document information; The graph and node links are displayed, which is beneficial to people's browsing of files; the model algorithm is optimized to improve the accuracy of matching.

Description

technical field [0001] The invention relates to the field of electronic text processing, in particular to a method and device for archiving and retroactively processing official document information of government organizations. Background technique [0002] At present, document processing, collection and archiving, including government documents, are increasingly moving towards online automated processing under the background of "integration of industrialization and industrialization", which brings great convenience to the work of enterprises, government agencies and other units. [0003] The official documents of government organizations have their particularities. The first is the level-by-level communication from top to bottom at all levels, the second is the assistance of brother units to formulate relevant supporting official documents, and the third is the high relevance of themes. With the accumulation of a large number of official documents of government organizatio...

AI Technical Summary

Problems solved by technology

With the accumulation of a large number of official documents of government organizations, the government organization itself has archives of documents, but subsequent development of new documents, there are difficulties in tracing and archiving. If random traceability and archiving will bring great inconvenience to future use and retrieval

For enterprises that pay attention to the official documents of government organizations, they have collected a large number of official documents of government organizations related to the business development of enterprises. If they are archived out of order and lack appropriate methods of tracing official document information or electronic equipment, when it is necessary to access historical documents on related topics, it seems difficult. Stretched and unable to cope

[0004] In addition, there are related processing methods for government documents in the prior art, such as the patent document with the publication number CN108829651A, which was applied by Ping An Technology (Shenzhen) Co., Ltd. to solve the intelligent problem of automatic matching of official documents, but it is still not enough to appear When relevant information or topics are to be traced, it is convenient for people to quickly obtain the official documents of relevant historical government organizations, which is also the deficiency of the existing technology

Images