A method for automatically sorting official documents and computer equipment

Abstract

The present invention provides a method for automatically sorting official documents. According to historical official documents, a sorting rule library is established; according to historical official documents, a semantic association model file is established; when sorting official documents, the current official document is matched according to the rule library, and if the match is successful, the document is sorted according to the matching result; if the match is unsuccessful, the keywords of the current official document are extracted, and the matching result is obtained by querying the semantic association model file, and then the document is sorted according to the matching result; the present invention also provides a computer device, which solidifies the job experience and avoids work errors and reduced efficiency of document processing caused by job changes and personnel rotation.

Description

Technical Field

[0001] The invention relates to a method for automatically sorting official documents and computer equipment. Background Art

[0002] Document distribution refers to the advisory activity in which the person in charge of the office clerk or business department and relevant personnel put forward suggestions for handling the documents that need to be handled. The opinions on document distribution can provide support for leadership decision-making. The constructive opinions on how to handle documents during the distribution are the results of the systematic research and personal experience summary of the distributor, and can provide a preliminary plan for the leader's approval. Through distribution, the enthusiasm and subjective initiative of the secretarial department and its personnel can be fully utilized, their ability to comprehensively handle various affairs can be improved, and their responsibilities of assisting decision-making and serving as a staff assistant can be fulfilled. Practical opinions on distribution can facilitate the reference of leaders when approving documents, and can make the leaders' opinions on approval highly clear and specific on the basis of the proposed opinions, which helps save leaders a lot of time and energy.

[0003] However, the traditional distribution of official documents is done by manual judgment. Based on personal experience, the title, source unit, document number, text content and other information of the document are judged to decide whether the document should be sent to a specific department or leader. This is highly subjective and arbitrary. When the distribution personnel change, the relevant distribution experience is difficult to pass on, and it is easy to make mistakes in the distribution, which cannot assist the leader's decision-making. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a method for automatic sorting of official documents and a computer device, which automatically analyzes the title, source unit, document number, and text content of the official document, adds a sorting rule library, and realizes the correspondence between the official document content and the sorting department or leader. If the match cannot be found through the rule library, the historical sorting files are analyzed through machine learning technology to realize automatic recommendation of the document sorting results, thereby solving the problem of inaccurate manual judgment of the sorting results.

[0005] One of the present inventions is achieved in this way: a method for automatically sorting official documents, comprising:

[0006] Step 1: Establish a rule base for division of duties based on historical documents;

[0007] Step 2: Create a semantic association model file based on historical documents;

[0008] Step 3: When sorting documents, the current document is matched according to the rule library. If the match is successful, the document is sorted according to the matching result. If the match is unsuccessful, the keywords of the current document are extracted, and the matching result is obtained by querying the semantic association model file, and then the document is sorted according to the matching result.

[0009] Furthermore, the step 1 is further specified as follows: based on historical official documents, the relationship between the title of the official document, the source unit, the document number, and the opinion content and the leaders and departments is established to complete the establishment of a branch rule library.

[0010] Furthermore, the step 2 is further specified as follows: according to the historical documents, keywords are obtained by processing through HANLP natural language technology, and then the keywords are processed by the semantic analysis engine WORD2VEC to obtain a semantic association model file.

[0011] Furthermore, the step 2 is further specified as follows: extracting corpus from the historical documents, processing the historical documents using HANLP natural language technology based on the extracted corpus to obtain keywords, and then using the semantic analysis engine WORD2VEC to process the keywords to obtain a semantic association model file.

[0012] Furthermore, the step 3 of extracting the keywords of the current document is further specifically as follows: extracting the keywords of the current document according to HANLP natural language processing technology.

[0013] The second aspect of the present invention is implemented as follows: a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the program:

[0014] Step 1: Establish a rule base for division of duties based on historical documents;

[0015] Step 2: Create a semantic association model file based on historical documents;

[0016] Step 3: When sorting documents, the current document is matched according to the rule library. If the match is successful, the document is sorted according to the matching result. If the match is unsuccessful, the keywords of the current document are extracted, and the matching result is obtained by querying the semantic association model file, and then the document is sorted according to the matching result.

[0017] Furthermore, the step 1 is further specified as follows: based on historical official documents, the relationship between the title of the official document, the source unit, the document number, and the opinion content and the leaders and departments is established to complete the establishment of a branch rule library.

[0018] Furthermore, the step 2 is further specified as follows: according to the historical documents, keywords are obtained by processing through HANLP natural language technology, and then the keywords are processed by the semantic analysis engine WORD2VEC to obtain a semantic association model file.

[0019] Furthermore, the step 2 is further specified as follows: extracting corpus from the historical documents, processing the historical documents using HANLP natural language technology based on the extracted corpus to obtain keywords, and then using the semantic analysis engine WORD2VEC to process the keywords to obtain a semantic association model file.

[0020] Furthermore, the step 3 of extracting the keywords of the current document is further specifically as follows: extracting the keywords of the current document according to HANLP natural language processing technology.

[0021] The present invention has the following advantages: a method for automatically sorting official documents and a computer device, which adopts this scheme to carry out the preparation and sorting of received documents, thereby improving the quality of information services and effectively improving the efficiency of the preparation and sorting of received documents by 30%. At the same time, it solidifies the job experience and avoids work errors and reduced efficiency of official document processing caused by job changes and personnel rotation. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The present invention will be further described below in conjunction with embodiments with reference to the accompanying drawings.

[0023] Figure 1 The present invention is a flowchart for executing the method.

[0024] Figure 2 It is a structural block diagram of a specific implementation method of the present invention. DETAILED DESCRIPTION

[0025] like Figure 1 As shown, the method for automatically sorting official documents of the present invention comprises:

[0026] Step 1: Based on historical official documents, establish the relationship between the title, source unit, document number, opinion content, and leadership and department, and complete the establishment of the sub-rule library;

[0027] Step 2: extracting corpus from the historical documents, processing the historical documents by HANLP natural language technology according to the extracted corpus to obtain keywords, and then using semantic analysis engine WORD2VEC to process the keywords to obtain a semantic association model file;

[0028] Step 3: When sorting documents, the current document is matched according to the rule base. If the match is successful, the document is sorted according to the matching result. If the match is unsuccessful, the keywords of the current document are extracted according to the HANLP natural language processing technology, and the matching result is obtained by querying the semantic association model file, and then the document is sorted according to the matching result.

[0029] like Figure 1As shown, the computer device of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the following steps are implemented:

[0030] Step 1: Based on historical official documents, establish the relationship between the title, source unit, document number, opinion content, and leadership and department, and complete the establishment of the sub-rule library;

[0031] Step 2: extracting corpus from the historical documents, processing the historical documents by HANLP natural language technology according to the extracted corpus to obtain keywords, and then using semantic analysis engine WORD2VEC to process the keywords to obtain a semantic association model file;

[0032] Step 3: When sorting documents, the current document is matched according to the rule base. If the match is successful, the document is sorted according to the matching result. If the match is unsuccessful, the keywords of the current document are extracted according to the HANLP natural language processing technology, and the matching result is obtained by querying the semantic association model file, and then the document is sorted according to the matching result.

[0033] A specific implementation mode of the present invention:

[0034] like Figure 2 As shown, when used in the power sector, the system mainly solves the problem of automatic recommendation of document distribution results, and gives priority to recommending through the distribution rule base. When the rule base cannot match the result, the machine learning recommendation technology is used. Specific operation steps:

[0035] (1) Establish a rule library for handling different tasks, and establish the relationship between the title, source unit, document number, opinion content, etc. of official documents and their leaders and departments.

[0036] (2) Corpus is extracted from the massive unstructured data of historical documents. The main text of each historical document is analyzed and processed. Through HANLP natural language processing, keywords are extracted after stop word removal, duplicate filtering, Chinese word segmentation, part-of-speech tagging, etc., to form segmented texts. Natural language processing technology is used to perform concept discovery, entity extraction, and domain segmentation on the segmented texts. Then, the semantic analysis engine WORD2VEC is used to perform association calculation and relationship extraction on the segmented texts, and finally a semantic association model file suitable for the office business field is formed.

[0037] (3) When handling official documents, the rule base is used for matching first. If a department or leader cannot be matched, machine learning recommendation is used.

[0038] (4) Extract the title, source unit, text content and other data of the currently registered official document, perform word segmentation and extraction, and submit it to the synonym service. By querying the semantic model file, the synonyms with high relevance and the part of speech of the organization name are returned. The organization names with high frequency of occurrence are ranked and calculated, and finally the recommended receiving departments are analyzed.

[0039] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A method for automatically sorting official documents, characterized in that: include: Step 1: Based on historical official documents, establish the relationship between the title, source unit, document number, opinion content, and leadership and department, and complete the establishment of the sub-rule library; Step 2: extracting corpus from the body of the historical document, processing the body of the historical document using HANLP natural language technology based on the extracted corpus to obtain keywords, and then using semantic analysis engine WORD2VEC to process the keywords to obtain a semantic association model file; Step 3. When sorting official documents, the current official document is matched according to the rule library. If the match is successful, the documents are sorted according to the matching results. If the match is unsuccessful, the keywords of the current official document are extracted according to the HANLP natural language processing technology, and the synonyms of the organization name with the set relevance and part of speech are obtained by querying the semantic association model file. The organization names are sorted and calculated according to the frequency of occurrence, and finally the matching results are obtained, and then the documents are sorted according to the matching results.

2. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the following steps are implemented: Step 1: Based on historical official documents, establish the relationship between the title, source unit, document number, opinion content, and leadership and department, and complete the establishment of the sub-rule library; Step 2: extracting corpus from the body of the historical document, processing the body of the historical document using HANLP natural language technology based on the extracted corpus to obtain keywords, and then using semantic analysis engine WORD2VEC to process the keywords to obtain a semantic association model file; Step 3. When sorting official documents, the current official document is matched according to the rule library. If the match is successful, the documents are sorted according to the matching results. If the match is unsuccessful, the keywords of the current official document are extracted according to the HANLP natural language processing technology, and the synonyms of the organization name with the set relevance and part of speech are obtained by querying the semantic association model file. The organization names are sorted and calculated according to the frequency of occurrence, and finally the matching results are obtained, and then the documents are sorted according to the matching results.

Images