CN114064920A - Unstructured document supervision method based on artificial intelligence and storage medium - Google Patents
Unstructured document supervision method based on artificial intelligence and storage medium Download PDFInfo
- Publication number
- CN114064920A CN114064920A CN202111344801.6A CN202111344801A CN114064920A CN 114064920 A CN114064920 A CN 114064920A CN 202111344801 A CN202111344801 A CN 202111344801A CN 114064920 A CN114064920 A CN 114064920A
- Authority
- CN
- China
- Prior art keywords
- supervision
- entity
- document
- knowledge
- entity relationship
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0635—Risk analysis of enterprise or organisation activities
Abstract
The invention relates to an unstructured document supervision method based on artificial intelligence and a storage medium, wherein the storage medium comprises the following steps: acquiring a supervision document from an enterprise or project as a material for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, and outputting an entity and entity relationship extraction combined model through model training; packaging the entity and entity relationship extraction combined model obtained by training into entity and entity relationship identification service; in the supervision process, continuously inputting supervision documents for entity and entity relationship identification service to extract information and construct a supervision knowledge map; and (4) supervising the non-structural document by a supervision knowledge graph. The original manual carding rule is replaced, the effect is more obvious, and the coverage is more comprehensive.
Description
Technical Field
The invention relates to the technical field of document supervision, in particular to an unstructured document supervision method and a storage medium based on artificial intelligence.
Background
Various engineering project process documents, financial files, system specification documents and the like exist in the enterprise operation or project management process, information such as careless and careless project process, link loss, financial fund confusion, system execution failure and the like is mined from the engineering project process documents, and centralized discovery, supervision and risk assessment are carried out; in the past, the supervision mode of manually reading and analyzing and finding the problem of document information from a large amount of documents needs to invest a large amount of manpower, and has the problems of low efficiency, more omission, supervision lag and the like. In the existing technical solution of document supervision, for example, a document monitoring and management system based on comprehensive security audit disclosed in application No. cn202111021148.x, for example, an electronic document classification supervision system based on a cloud platform in application No. CN202110759545.0, although a document can be monitored or supervised, the problem of finding potential relationships of multiple documents is not solved, and document content entities or entity relationships are supervised, and meanwhile, data supplementation all adopts a manual entry mode, so that labor cost is high.
Disclosure of Invention
Therefore, it is necessary to provide an unstructured document supervision method and a storage medium based on artificial intelligence, which solve the supervision problems that a large amount of manpower is required for supervision of various documents in the existing enterprise operation or project management process, the efficiency is low, many omissions are made, the supervision is delayed, and potential relationships among multiple documents cannot be found.
To achieve the above object, the inventor provides an unstructured document supervision method based on artificial intelligence, comprising the following steps:
acquiring a supervision document from an enterprise or project as a material for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, and outputting an entity and entity relationship extraction combined model through model training;
packaging the entity and entity relationship extraction combined model obtained by training into entity and entity relationship identification service;
in the supervision process, continuously inputting supervision documents for entity and entity relationship identification service to extract information and construct a supervision knowledge map;
and (4) supervising the non-structural document by a supervision knowledge graph.
Further optimization, the step of "supervising the non-structural document by the supervising knowledge graph" specifically comprises the following steps:
and evaluating the risks of the enterprises or projects through the supervision knowledge graph and the supervision rule base.
And further optimizing, wherein the supervision rule base is preset through a platform and is subsequently and continuously recorded and perfected.
Further optimization, the step of "supervising the non-structural document by the supervising knowledge graph" specifically comprises the following steps:
and tracing and analyzing the source of the supervision problem by the supervision knowledge map to locate the root source or the influence factor.
Further optimizing, the step of tracing, analyzing and positioning the root or influence factor of the supervision problem by the supervision knowledge map further comprises the following steps:
and presenting the influence range of the supervision problem according to a visualization technology.
Yet another embodiment is provided, a storage medium having a computer program stored therein, the computer program when executed by a processor performing the steps of:
acquiring a supervision document from an enterprise or project as a material for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, and outputting an entity and entity relationship extraction combined model through model training;
packaging the entity and entity relationship extraction combined model obtained by training into entity and entity relationship identification service;
in the supervision process, continuously inputting supervision documents for entity and entity relationship identification service to extract information and construct a supervision knowledge map;
and (4) supervising the non-structural document by a supervision knowledge graph.
Further optimization, the step of "supervising the non-structural document by supervising the knowledge graph" specifically comprises the following steps:
and evaluating the risks of the enterprises or projects through the supervision knowledge graph and the supervision rule base.
And further optimizing, wherein the supervision rule base is preset through a platform and is subsequently and continuously recorded and perfected.
Further optimization, the step of "supervising the non-structural document by supervising the knowledge graph" specifically comprises the following steps:
and tracing and analyzing the source of the supervision problem by the supervision knowledge map to locate the root source or the influence factor.
Further optimizing, the step of tracing, analyzing and positioning the root or influence factor of the supervision problem by the supervision knowledge map further comprises the following steps:
and presenting the influence range of the supervision problem according to a visualization technology.
Different from the prior art, the technical scheme constructs the supervision knowledge map by analyzing the potential entities and the entity relations in the massive documents, replaces the original manual carding work, has obvious advantages and is more concentrated on the identification of the entity relations in the same or different documents; meanwhile, artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training is adopted, the model extracted from the artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training is more comprehensive in extraction and identification of actual input documents, and the artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training can be matched with model training to be rich continuously, replaces the original manual carding rule, is more remarkable in effect, and is more comprehensive in coverage. And by combining knowledge graph correlation techniques, the extracted entities and relationships can meet flexible and multi-aspect analysis requirements in actual supervision application analysis.
Drawings
FIG. 1 is a schematic flow chart of an artificial intelligence based method for managing unstructured documents according to an embodiment;
FIG. 2 is a schematic flow chart of an unstructured document supervision method based on artificial intelligence according to an embodiment.
Detailed Description
To explain technical contents, structural features, and objects and effects of the technical solutions in detail, the following detailed description is given with reference to the accompanying drawings in conjunction with the embodiments.
Referring to fig. 1-2, the present embodiment provides an unstructured document supervision method based on artificial intelligence, which includes the following steps:
step S110: acquiring a supervision document from an enterprise or project as a material for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, and outputting an entity and entity relationship extraction combined model through model training;
step S120: packaging the entity and entity relationship extraction combined model obtained by training into entity and entity relationship identification service;
step S130: in the supervision process, continuously inputting supervision documents for entity and entity relationship identification service to extract information and construct a supervision knowledge map;
step S140: and (4) supervising the non-structural document by a supervision knowledge graph.
The method comprises the steps of acquiring supervision documents such as project process documents, financial documents, system documents, contract documents and the like from enterprises or projects to serve as materials for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, outputting a model through model training, packaging the trained model into entity and entity relationship identification service, continuously inputting related supervision documents in the supervision process for entity and entity relationship service information extraction, extracting supervision entities, supervision elements and potential relationships in the supervision entities and entity relationships, constructing a supervision knowledge graph to provide a data base for subsequent application and application analysis, and supervising non-structural documents through the supervision knowledge graph. The supervision knowledge map is constructed by analyzing the potential entities and the entity relations in the massive documents, the original manual combing work is replaced, the advantages are obvious, and the supervision knowledge map is more concentrated on the identification of the entity relations in the same or different documents; meanwhile, artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training is adopted, the model extracted from the artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training is more comprehensive in extraction and identification of actual input documents, and the artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training can be matched with model training to be rich continuously, replaces the original manual carding rule, is more remarkable in effect, and is more comprehensive in coverage. And by combining knowledge graph correlation techniques, the extracted entities and relationships can meet flexible and multi-aspect analysis requirements in actual supervision application analysis.
In this example, the "supervising unstructured documents via the supervising knowledge graph" may specifically include the following step S141:
and evaluating the risks of the enterprises or projects through the supervision knowledge graph and the supervision rule base.
The method comprises the steps of establishing a complete set of complete supervision rule base, wherein the supervision rule base is preset through a platform and is continuously input and perfected subsequently, and risks of enterprises or projects can be evaluated through the combination of a constructed supervision knowledge map and the supervision rule base, so that the original manual carding work is replaced.
In this embodiment, tracing of the supervision problem may be implemented, where the "supervise the non-structural document by supervising the knowledge graph" specifically includes the following step S142:
and tracing and analyzing the source of the supervision problem by the supervision knowledge map to locate the root source or the influence factor.
And constructing a supervision knowledge graph, visually presenting each link problem in enterprise or project supervision, and analyzing the root and influence of the link problem on branch pipe presentation through the traceability analysis of the supervision knowledge graph. Wherein, the step of tracing, analyzing and positioning the root or the influence factor of the supervision problem by the supervision knowledge map further comprises the following steps:
and presenting the influence range of the supervision problem according to a visualization technology.
Yet another embodiment is provided, a storage medium having a computer program stored therein, the computer program when executed by a processor performing the steps of:
acquiring a supervision document from an enterprise or project as a material for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, and outputting an entity and entity relationship extraction combined model through model training;
packaging the entity and entity relationship extraction combined model obtained by training into entity and entity relationship identification service;
in the supervision process, continuously inputting supervision documents for entity and entity relationship identification service to extract information and construct a supervision knowledge map;
and (4) supervising the non-structural document by a supervision knowledge graph.
The method comprises the steps of acquiring supervision documents such as project process documents, financial documents, system documents, contract documents and the like from enterprises or projects to serve as materials for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, outputting a model through model training, packaging the trained model into entity and entity relationship identification service, continuously inputting related supervision documents in the supervision process for entity and entity relationship service information extraction, extracting supervision entities, supervision elements and potential relationships in the supervision entities and entity relationships, constructing a supervision knowledge graph to provide a data base for subsequent application and application analysis, and supervising non-structural documents through the supervision knowledge graph. The supervision knowledge map is constructed by analyzing the potential entities and the entity relations in the massive documents, the original manual combing work is replaced, the advantages are obvious, and the supervision knowledge map is more concentrated on the identification of the entity relations in the same or different documents; meanwhile, artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training is adopted, the model extracted from the artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training is more comprehensive in extraction and identification of actual input documents, and the artificial intelligence bidirectional LSTM + RNN entity and entity relation extraction combined model training can be matched with model training to be rich continuously, replaces the original manual carding rule, is more remarkable in effect, and is more comprehensive in coverage. And by combining knowledge graph correlation techniques, the extracted entities and relationships can meet flexible and multi-aspect analysis requirements in actual supervision application analysis.
In this example, the step of "supervising the unstructured document by the supervising knowledge graph" may specifically include the following steps:
and evaluating the risks of the enterprises or projects through the supervision knowledge graph and the supervision rule base.
The method comprises the steps of establishing a complete set of complete supervision rule base, wherein the supervision rule base is preset through a platform and is continuously input and perfected subsequently, and risks of enterprises or projects can be evaluated through the combination of a constructed supervision knowledge map and the supervision rule base, so that the original manual carding work is replaced.
In this embodiment, the tracing of the supervision problem may be implemented, and the step of "supervising the non-structural document through the supervision knowledge graph" specifically includes the following steps:
and tracing and analyzing the source of the supervision problem by the supervision knowledge map to locate the root source or the influence factor.
And constructing a supervision knowledge graph, visually presenting each link problem in enterprise or project supervision, and analyzing the root and influence of the link problem on branch pipe presentation through the traceability analysis of the supervision knowledge graph. Wherein, the step of tracing, analyzing and positioning the root or the influence factor of the supervision problem by the supervision knowledge map further comprises the following steps:
and presenting the influence range of the supervision problem according to a visualization technology.
It should be noted that, although the above embodiments have been described herein, the invention is not limited thereto. Therefore, based on the innovative concepts of the present invention, the technical solutions of the present invention can be directly or indirectly applied to other related technical fields by making changes and modifications to the embodiments described herein, or by using equivalent structures or equivalent processes performed in the content of the present specification and the attached drawings, which are included in the scope of the present invention.
Claims (10)
1. An unstructured document supervision method based on artificial intelligence is characterized by comprising the following steps:
acquiring a supervision document from an enterprise or project as a material for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, and outputting an entity and entity relationship extraction combined model through model training;
packaging the entity and entity relationship extraction combined model obtained by training into entity and entity relationship identification service;
in the supervision process, continuously inputting supervision documents for entity and entity relationship identification service to extract information and construct a supervision knowledge map;
and (4) supervising the non-structural document by a supervision knowledge graph.
2. The method for managing unstructured documents based on artificial intelligence as claimed in claim 1, wherein said step of "managing unstructured documents by managing knowledge-graph" includes the following steps:
and evaluating the risks of the enterprises or projects through the supervision knowledge graph and the supervision rule base.
3. The method for managing unstructured documents based on artificial intelligence, as recited in claim 2, wherein the management rule base is pre-set by platform and then continuously entered and perfected.
4. The method for managing unstructured documents based on artificial intelligence as claimed in claim 1, wherein said step of "managing unstructured documents by managing knowledge-graph" further comprises the following steps:
and tracing and analyzing the source of the supervision problem by the supervision knowledge map to locate the root source or the influence factor.
5. The method for managing unstructured documents based on artificial intelligence, as defined in claim 4, further comprising the following steps after said step of "locating root or influence factor for analysis of regulatory issues by regulatory knowledge-graph tracing":
and presenting the influence range of the supervision problem according to a visualization technology.
6. A storage medium having a computer program stored therein, the computer program when executed by a processor performing the steps of:
acquiring a supervision document from an enterprise or project as a material for bidirectional LSTM + RNN entity and entity relationship extraction combined model training, and outputting an entity and entity relationship extraction combined model through model training;
packaging the entity and entity relationship extraction combined model obtained by training into entity and entity relationship identification service;
in the supervision process, continuously inputting supervision documents for entity and entity relationship identification service to extract information and construct a supervision knowledge map;
and (4) supervising the non-structural document by a supervision knowledge graph.
7. The storage medium of claim 6, wherein the step of "supervising the unstructured document by means of a supervising knowledge graph" comprises in particular the steps of:
and evaluating the risks of the enterprises or projects through the supervision knowledge graph and the supervision rule base.
8. The storage medium of claim 7, wherein the regulatory rule base is pre-set by a platform and then continuously entered and perfected.
9. The storage medium of claim 6, wherein the step of "supervising the unstructured document by means of a supervising knowledgegraph" further comprises the steps of:
and tracing and analyzing the source of the supervision problem by the supervision knowledge map to locate the root source or the influence factor.
10. The storage medium of claim 9, wherein said step of "locating root cause or impact factor for regulatory problem analysis by regulatory knowledge-graph" further comprises the steps of:
and presenting the influence range of the supervision problem according to a visualization technology.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111344801.6A CN114064920A (en) | 2021-11-15 | 2021-11-15 | Unstructured document supervision method based on artificial intelligence and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111344801.6A CN114064920A (en) | 2021-11-15 | 2021-11-15 | Unstructured document supervision method based on artificial intelligence and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114064920A true CN114064920A (en) | 2022-02-18 |
Family
ID=80271828
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111344801.6A Pending CN114064920A (en) | 2021-11-15 | 2021-11-15 | Unstructured document supervision method based on artificial intelligence and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114064920A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230059494A1 (en) * | 2021-08-19 | 2023-02-23 | Digital Asset Capital, Inc. | Semantic map generation from natural-language text documents |
-
2021
- 2021-11-15 CN CN202111344801.6A patent/CN114064920A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20230059494A1 (en) * | 2021-08-19 | 2023-02-23 | Digital Asset Capital, Inc. | Semantic map generation from natural-language text documents |
| US20230056987A1 (en) * | 2021-08-19 | 2023-02-23 | Digital Asset Capital, Inc. | Semantic map generation using hierarchical clause structure |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110929036B (en) | Electric power marketing inspection management method, electric power marketing inspection management device, computer equipment and storage medium | |
| CN111461668A (en) | Digital auditing system and method based on process automation technology | |
| CN109784758B (en) | Engineering quality supervision early warning system and method based on BIM model | |
| CN114281877A (en) | Data management system and method | |
| CN114064920A (en) | Unstructured document supervision method based on artificial intelligence and storage medium | |
| CN113609393A (en) | Digital platform based on data service and data management | |
| CN112598142B (en) | Wind turbine maintenance working quality inspection auxiliary method and system | |
| CN105138612A (en) | Method and system for analyzing and locating reasons of data consistency difference | |
| Kropatschek et al. | Towards the representation of cross-domain quality knowledge for efficient data analytics | |
| CN112508541A (en) | Portable mobile phone acquisition software system based on training trainee information data acquisition | |
| CN112182065A (en) | Asset management system and method based on automatic acquisition and multi-source import | |
| Suleykin et al. | Associative Rules-Driven Intelligent Production Schedule Control System for Digital Manufacturing Ecosystem | |
| CN113706098B (en) | Business-based deviation reason identification method and device and electronic equipment | |
| CN111913706B (en) | Topology construction method of dispatching automation system, storage medium and computing equipment | |
| Nagy et al. | Using process mining in real-time to reduce the number of faulty products | |
| CN112650796A (en) | Automatic application data collection and storage management system | |
| Xing et al. | Study on the Impact of Big Data Technology on the Audit and its Application | |
| CN112968941B (en) | Data acquisition and man-machine collaborative annotation method based on edge calculation | |
| Oyoo | Collaboration-Based Automatic Data Validation Framework for Enterprise Asset Management | |
| Oladimeji | Identifying case studies for performance measurement implementation models using systematic review | |
| Siek | Investigating inductive miner and fuzzy miner in automated business model generation | |
| CN112799935A (en) | Intelligent quality detection method, device and platform | |
| CN117896171A (en) | Safety testing method and device | |
| CN113887862A (en) | Energy metering service data analysis method and system | |
| Venkatesh | Beyond the Four Key Metrics: Quantifying DevOps Practice |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |