CN118211941A - Automatic community work order circulation method and system based on RPA - Google Patents
Automatic community work order circulation method and system based on RPA Download PDFInfo
- Publication number
- CN118211941A CN118211941A CN202410629691.5A CN202410629691A CN118211941A CN 118211941 A CN118211941 A CN 118211941A CN 202410629691 A CN202410629691 A CN 202410629691A CN 118211941 A CN118211941 A CN 118211941A
- Authority
- CN
- China
- Prior art keywords
- data
- rpa
- work order
- image
- data packet
- 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
- 238000000034 method Methods 0.000 title claims abstract description 38
- 238000012545 processing Methods 0.000 claims abstract description 34
- 238000013475 authorization Methods 0.000 claims abstract description 28
- 238000011049 filling Methods 0.000 claims abstract description 22
- 238000000586 desensitisation Methods 0.000 claims abstract description 7
- 238000000605 extraction Methods 0.000 claims abstract description 7
- 230000011218 segmentation Effects 0.000 claims description 15
- 238000012549 training Methods 0.000 claims description 15
- 238000012015 optical character recognition Methods 0.000 claims description 12
- 238000013441 quality evaluation Methods 0.000 claims description 11
- 238000013500 data storage Methods 0.000 claims description 10
- 230000004927 fusion Effects 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 230000007704 transition Effects 0.000 claims description 6
- 230000002194 synthesizing effect Effects 0.000 claims description 4
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 238000003491 array Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- 238000009434 installation Methods 0.000 claims description 3
- 238000012800 visualization Methods 0.000 claims description 3
- 238000012216 screening Methods 0.000 claims description 2
- 230000000007 visual effect Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000004888 barrier function Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides an automatic community work order circulation method and system based on RPA, and relates to the technical field of work order processing based on Internet. The invention configures the data log to form an RPA task description file; performing operation scheduling on the RPA task description file, providing an RPA task automation execution context environment, and recording an RPA task operation log; automatically acquiring authorization page data of a third-party service system, performing character recognition on authorization images and PDF file data, performing semantic extraction on text characters, and finally constructing a complete structured data packet; and automatically classifying pictures and characters in the complete structured data packet, automatically filling text data and uploading file data on an authorization page, and completing data reporting. The invention can realize intelligent collection of the business work order data of the intranet system, and automatically completes circulation to the community system after carrying out safe desensitization treatment on the work order data.
Description
Technical Field
The invention relates to the technical field of work order processing based on the Internet, in particular to an automatic community work order circulation method and system based on RPA.
Background
The intelligent community treatment belongs to an important component of the gridding treatment, wherein various business work orders, such as population information update, community event reporting treatment, community interview and the like, need to be circulated in the intelligent community and the gridding treatment system. However, due to network security limitation and old system transformation limitation, the intelligent community system cannot directly exchange data with the gridding treatment platform, so that basic grid staff needs to repeatedly input and process data in a plurality of systems, the workload of basic staff is greatly increased, and the accuracy and timeliness of the data are reduced.
Disclosure of Invention
The invention aims to: an automatic community work order circulation method based on RPA is provided, and a system for realizing the method is further provided, so that the problems in the prior art are solved.
In a first aspect, an automatic community work order circulation method based on RPA is provided, and the method includes the following steps:
S1, capturing and identifying service work order information to form a first data packet;
S2, analyzing the first data packet, selecting a service category to which the first data packet belongs, and automatically filling and popping up a service order;
S3, recording and storing the historical steps of the current service order and the data information in a preset storage space in an increment mode of a data log;
s4, creating an encrypted blockchain channel for a data log to which the current service order belongs;
s5, configuring the data log in the block chain channel through a visual operation interface selection mode type to form an RPA task description file;
S6, performing operation scheduling on the RPA task description file, providing an RPA task automation execution context environment, and recording an RPA task operation log;
s7, automatically logging in a third-party service system through preset authorization information, acquiring RPA automatic operation authorization and automatically acquiring service order data;
S8, automatically acquiring authorization page data of the third-party service system, performing text recognition on authorization images and PDF file data in the authorization page data of the third-party service system, performing semantic extraction on text, and finally constructing a complete structured data packet;
S9, automatically classifying pictures and characters in the complete structured data packet, wherein the character classification comprises title details and contents corresponding to the title details; and after checking the corresponding names of the report forms according to the title details and the contents of the corresponding details, automatically filling text data and uploading file data on an authorized page to finish data reporting.
In a further embodiment of the first aspect, the service work order information is divided into a written work order and/or an entered work order;
For the written worksheet: obtaining a preset number of bad handwriting images, and training the bad handwriting images to obtain an image-text recognition model; extracting service work order information in the written work order by using the image-text recognition model;
for the entered worksheet: and acquiring a complete work order picture in a long screenshot mode, and extracting business work order information in the complete work order picture.
In a further embodiment of the first aspect, the obtaining a predetermined number of scratched handwriting images, and training with the scratched handwriting images to obtain a graphic recognition model includes:
S1-1, searching a preset number of bad handwriting images based on a search engine, screening out the bad handwriting images with clear characters and no shielding, extracting key features of the bad handwriting images, wherein the key features comprise determined components and strokes, and summarizing the key features to construct a language training set;
s1-2, constructing an image-text recognition model, and training the image-text recognition model based on the language training set;
s1-3, performing OCR (optical character recognition) on a to-be-recognized illegal handwriting image with text blurring and/or text shielding by using the trained text-to-text recognition model to obtain a primary recognition result;
S1-4, extracting recognition components and/or strokes of the illegal handwriting image as source character codes of the image to be recognized, and combining the preliminary recognition results to obtain a recommended recognition result and confidence degrees of all recognition results of the image to be recognized through a picture and text recognition model corresponding to the source character codes of the current image to be recognized;
S1-5, outputting a recommended recognition result with the confidence coefficient higher than a set threshold value and a corresponding confidence coefficient as a recognition result of the image to be recognized.
In a further embodiment of the first aspect, extracting, in S1-4, the recognition components or/and strokes of the scratched handwriting image as a source codeword of the image to be recognized includes:
Taking the identification components and/or strokes of the sloppy handwriting image as a digital image, acquiring a gray level histogram of the digital image, and then carrying out binarization processing on the digital image according to a binarization threshold value;
Removing noise of the binarized digital image, and performing edge smoothing processing on the denoised digital image to obtain a source character code of the image to be identified.
In a further embodiment of the first aspect, the method further includes performing OCR text recognition on the to-be-recognized bad handwriting image with text blurring and/or text shielding by using the trained text recognition model, and after obtaining the preliminary recognition result, further including:
S1-3-1, calculating to obtain a fusion blind image quality evaluation index value based on the preliminary identification result;
S1-3-2, comparing the fusion blind image quality evaluation index value with a threshold value, wherein the fusion blind image quality evaluation index value is larger than or equal to the threshold value and is a clear target image, and executing the step S1-3-4; otherwise, executing the step S1-3-3 for blurring the target image;
S1-3-3, inputting the confirmed fuzzy target image as a GAN network model, outputting the clear target image as a network, and realizing the definition of the fuzzy target image;
s1-3-4, positioning and identifying the clear target image confirmed in the step S1-3-2 and the clear target image confirmed in the step S1-3-3.
In a further embodiment of the first aspect, the extracting service work order information in the complete work order picture includes:
Selecting the synthesized complete work order picture in a local storage medium;
Starting an OCR module to identify characters and patterns in a complete work order picture, and extracting and synthesizing complete work order information, wherein the complete work order information comprises N data item title texts, data item content texts and data item content pictures; wherein, N is a natural number greater than 1;
Desensitizing the extracted work order information according to the current filling work order type and the pre-configuration information thereof;
And caching the extracted complete work order data work order into DATAARRAY arrays.
In a further embodiment of the first aspect, the preconfiguration information includes a work item field name, a field content check format, a field desensitization mode, whether a field has to be filled;
The desensitization processing of the extracted work order information comprises keyword filtering, keyword replacement, content abstract encryption, content symmetric encryption, content asymmetric encryption and picture watermarking.
In a further embodiment of the first aspect, step S2 further comprises:
analyzing the first data packet to obtain analysis data; and selecting an item to which the analysis data belongs, selecting a corresponding table according to different items, and writing the analysis data into the corresponding table.
In a further embodiment of the first aspect, step S3 further includes:
S3-1, binding the project by using a background linking unit, producing a business background of the corresponding project, producing a background arrow of a business order, and transmitting the background arrow to a data storage module;
s3-2, the data storage module produces a channel under the arrow and a storage disk connected with the channel, and stores historical steps of service orders and data information.
In a further embodiment of the first aspect, step S4 further includes:
S4-1, obtaining a background arrow of a service order by an encryption processing module, and producing a private key corresponding to a first data packet;
s4-2, encrypting the first data packet into a second data packet by using the private key, and sending the second data packet to other nodes of the blockchain.
In a further embodiment of the first aspect, the means for encrypting the first data packet is:
splitting a first data packet into at least two first number domains, filling the first number domains to blank positions of two-dimensional codes generated by private keys, and filling black positions of the two-dimensional codes into association number domains according to an association number domain database;
The association number field comprises at least 24 groups of characters, and each group of character strings consists of any character of 0-9, a-Z and A-Z; namely, 62N combinations of each group of character strings are provided, wherein N is the number of characters of the group of character strings;
Encrypting each character of all character strings by using an elliptic curve function to obtain a character or character string, and sequentially stringing all the characters and character strings to form an associated character string;
Wherein the elliptic curve function is expressed as follows:
y 2＝x3+ax2 +b, wherein a and b are any values;
The association number field is any two groups of combination numbers within 0-9 which are randomly generated based on a binary algorithm.
In a further embodiment of the first aspect, step S8 further includes:
S8-1, executing word segmentation operation based on the authorized image or the words identified in the PDF file, decomposing a sentence into a plurality of paths connected with the nodes, and obtaining the transition probability among the nodes, wherein the word segmentation device with the maximum transition probability among the nodes is represented as follows:
wherein i represents one of the nodes; representing the next node connected to node i; And P represents the probability between nodes and the product of the probabilities, respectively;
s8-2, collecting an external knowledge dictionary, matching the content of the node j with the knowledge dictionary, giving a preset additional score, and embedding the additional score into a word segmentation device, wherein the expression is as follows:
In the method, in the process of the invention, An additional score for node j; if j is a person name, a department name or an arbitrary administrative division name, thenSet to 1, otherwise 0;
s8-3, marking word segmentation in a sentence by using the word segmentation device;
S8-4, installation spaCy and scikit-learn;
S8-5, importing the marked sentences into a Python library to perform semantic search, marking the marked sentences by using spaCy components, and vectorizing the marked sentences by using scikit-learn components;
s8-6, searching to obtain a plurality of document indexes, and taking the document index with the highest confidence as a reference result to obtain a semantic recognition result of the current sentence;
S8-7, repeating the steps S8-5 to S8-6 to obtain the semantic recognition result of all sentences;
And S8-8, summarizing semantic recognition results of all sentences to form a complete structured data packet.
As a second aspect of the present invention, an automatic circulation system of community worksheets is provided, where the automatic circulation system includes a data acquisition module, a service processing module, a data storage module, an encryption processing module, an RPA configuration module, an RPA scheduling module, an RPA automatic login module, an RPA data acquisition module, and an RPA data reporting module.
The data acquisition module is used for capturing the service work order information to form a first data packet.
The service processing module is used for analyzing the first data packet, selecting the service category to which the data packet belongs, and automatically filling and popping up the service order.
The data storage module is used for storing historical steps and data information of the business orders.
The encryption processing module is used for creating an encrypted blockchain channel for the service order and encrypting channel data writing and reading of the blockchain channel.
And the RPA configuration module configures data through selecting a mode type through the image visualization operation interface to form an RPA task description file.
The RPA scheduling module is used for performing operation scheduling on the RPA task description file, providing an RPA task automation execution context environment, recording an RPA task operation log, and realizing automatic triggering, abnormal interruption and state rollback of the task.
The RPA automatic login module automatically logs in a third-party service system through preconfigured authorization information, acquires RPA automatic operation authorization and automatically acquires the service order data.
The RPA data acquisition module automatically acquires the authorization page data of the third party service system, performs text recognition on the third party authorization image and PDF file data through an OCR algorithm, performs semantic extraction on text, and finally constructs a complete structured data packet.
The RPA data filling module is used for automatically classifying pictures and characters in the collected structured data packet, the character classification comprises title details and contents corresponding to the title details, the names corresponding to the report are checked according to the title details and the contents corresponding to the title details, and then text data filling and file data uploading are automatically realized on an authorized page, so that data reporting is completed.
Compared with the prior art, the invention has the following beneficial effects: the intelligent collection method can realize intelligent collection of business work order data of the intranet system, automatically complete circulation to the community system after safety desensitization treatment is carried out on the work order data, break isolation barriers of the intranet system and the community system on the premise of guaranteeing data safety, reduce repeated recording work of basic staff, reduce data errors and circulation delay caused by repeated recording, and accelerate construction of the intelligent community system to fall to the ground.
Drawings
Fig. 1 is a schematic diagram of an automatic circulation system of community service worksheets based on RPA according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a data acquisition module according to an embodiment of the present invention.
Fig. 3 is a flowchart of a blur processing unit according to an embodiment of the present invention.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the present invention. It will be apparent, however, to one skilled in the art that the invention may be practiced without one or more of these details. In other instances, well-known features have not been described in detail in order to avoid obscuring the invention.
The embodiment discloses an automatic community service work order circulation system based on RPA, which is shown in FIG. 1, and comprises a data acquisition module, a service processing module, a data storage module, an encryption processing module, an RPA configuration module, an RPA scheduling module, an RPA automatic login module, an RPA data acquisition module and an RPA data reporting module.
The data acquisition module is used for capturing and identifying the work order information through a specific algorithm to form a first data packet.
The service processing module analyzes the acquired first data packet, selects the service category to which the data packet belongs, and automatically fills and pops up the service order.
The data storage module stores historical steps of business transaction behaviors and data information.
The encryption processing module is positioned in the data storage module, creates an encrypted blockchain channel for the current service, and encrypts channel data writing and reading of the blockchain channel.
The RPA configuration module configures the data automation processing flow through the image visualization operation interface to form an RPA task description file.
The RPA scheduling module performs operation scheduling on the RPA task description file, provides an RPA task automation execution context environment, records an RPA task operation log, and realizes automatic triggering, abnormal interruption and state rollback of the task.
The RPA automatic login module automatically logs in a third party service system through preconfigured authorization information to acquire RPA automatic operation authorization.
The RPA data acquisition module automatically acquires the authorization page data of the third party service system, performs text recognition on the third party authorization image and PDF file data through an OCR algorithm, performs semantic extraction on text, and finally constructs a complete structured data packet.
The RPA data filling module automatically classifies the collected pictures and words in the structured data packet, the word classification comprises title details and contents corresponding to the title details, and then the report corresponding names are checked according to the title details and the contents corresponding to the title details, and then text data filling and file data uploading are automatically realized on an authorized page, so that data reporting is completed.
In the above embodiment, the service work order information is divided into a writing work order and an entering work order.
For a written work order, the way to extract its information is as follows: obtaining a preset number of bad handwriting images, and training the bad handwriting images to obtain an image-text recognition model; and extracting the service work order information in the written work order by using the image-text recognition model.
The following discloses a feasible process for extracting service work order information in the written work order by using an image-text recognition model:
(1) Processing the processed target image by using an image quality evaluation method to obtain a fusion blind image quality evaluation index value;
(2) Comparing the obtained fusion blind image quality evaluation index value with a threshold value, wherein the fusion blind image quality evaluation index value is larger than or equal to the threshold value and is a clear target image, and executing (4), otherwise, executing (3) the fusion blind image quality evaluation index value is a fuzzy target image;
(3) Inputting the confirmed fuzzy target image as a GAN network model, outputting the clear target image as a network, and realizing the definition of the fuzzy target image;
(4) And (3) locating and identifying the clear target image confirmed in the step (2) and the clear target image confirmed in the step (3).
As can be seen from fig. 2 and 3, the data acquisition module includes a fuzzy processing unit, and performs model training on the obtained scratched handwriting by using a fuzzy algorithm, which includes the steps of:
(1) Extracting key parts of the bad handwriting based on the bad handwriting searched by the search engine, wherein the key parts comprise determined components and strokes as a constructed language training set;
(2) Constructing a text recognition model, and carrying out model training based on a language training set to obtain a graph-text recognition model;
(3) Performing OCR character recognition on an image to be recognized with character blurring and character shielding to obtain a primary recognition result;
(4) Extracting a sloppy handwriting recognition component or/and strokes as a source character code of an image to be recognized, and combining a preliminary recognition result to obtain a recommended recognition result and confidence degrees of all recognition results of the image to be recognized through a picture-text recognition model corresponding to the source character code of the current image to be recognized;
(5) And outputting a recommended recognition result with the confidence coefficient higher than the set threshold value and the corresponding confidence coefficient as a recognition result of the image to be recognized.
It should be noted that the fuzzy processing unit constructs the image-text recognition model based on the deep neural network.
In addition, the data acquisition module comprises a handwriting processing unit, a gray level histogram of the digital image is made based on the captured picture data, and then the digital image is binarized according to a binarization threshold value;
And removing the binarized digital image noise, and finally smoothing the edge of the denoised digital image to obtain the source character code of the image to be identified.
For the entered worksheet: acquiring a complete work order picture in a long screenshot mode, extracting business work order information in the complete work order picture, and providing a feasible embodiment:
executing the screenshot, and storing the picture in a local storage medium;
Judging whether the work order page is scrolled to the bottom, if so, entering the next step, otherwise, repeatedly executing the screenshot;
When the work order page is rolled to the bottom, performing jigsaw operation on a plurality of screenshots cached in the local medium, synthesizing a plurality of screenshots into a complete work order picture, and caching the complete work order picture in the local medium;
Selecting the synthesized complete work order picture in a local storage medium;
Starting an OCR module to identify characters and patterns in a complete work order picture, and extracting and synthesizing complete work order information, wherein the complete work order information comprises N data item title texts, data item content texts and data item content pictures; wherein, N is a natural number greater than 1;
Desensitizing the extracted work order information according to the current filling work order type and the pre-configuration information thereof;
And caching the extracted complete work order data work order into DATAARRAY arrays.
The preconfiguration information comprises a work item field name, a field content check format, a field desensitization mode and whether a field is needed to be filled or not; the method comprises the steps of desensitizing the extracted work order information, wherein the desensitizing treatment comprises keyword filtering, keyword replacement, content abstract encryption, content symmetric encryption, content asymmetric encryption and picture watermarking.
Further, the business processing module comprises a judging unit, a corresponding table is selected based on verified item blending, and the identified data is correspondingly written according to the item.
And secondly, the business processing module comprises a background linking unit which is bound with the project and used for producing a business background corresponding to the project, and a background arrow for producing the current business is transmitted to the encryption processing module, and the encryption processing module is used for producing a channel under the current arrow and a storage disk connected with the channel based on the current arrow.
Furthermore, the encryption processing module generates a private key generating module under the first data based on the acquired arrow, generates a private key corresponding to the first data based on the private key generating module, encrypts the first data into second data and sends the second data to other nodes of the blockchain.
The first data is encrypted in the following manner: splitting the first data into at least two first number domains, filling the first number domains to blank positions of the two-dimensional code generated by the private key, and filling the black positions of the two-dimensional code into an association number domain according to an association number domain database.
The association number field comprises at least 24 groups of characters, each group of character strings consists of any character of 0-9, a-Z and A-Z, namely 62N kinds of combination modes of each group of character strings are adopted, wherein N is the number of characters of the group of character strings, and an elliptic curve function is utilized: y 2＝x3+ax2 +b, wherein a and b are arbitrary values, each character of the M groups of character strings is obtained and encrypted by using an elliptic curve function to obtain a character or character string, and then all the characters and character strings are sequentially connected in series to form an associated character string.
It should be noted that the number field is any two sets of combined numbers within 0-9 randomly generated based on a binary algorithm.
In the technology, the same batch of processing of the sweep data of the same batch is created by adopting the blockchain, so that the server pressure in the data processing process is relieved, the collected picture is processed in a handwriting way, the word sequence similar words are selected to be generated, and the word sequence similar words are selected automatically, so that the effect of data processing is optimized, and the probability of occurrence of unrecognizable record is reduced.
Furthermore, the authorization page data of the third party service system is automatically acquired, character recognition is carried out on the authorization image and PDF file data in the authorization page data of the third party service system, semantic extraction is carried out on text characters, and finally a complete structured data packet is constructed, wherein the specific process is as follows:
(1) Based on the authorized image or the words identified in the PDF file, word segmentation operation is executed, a sentence is decomposed into paths connected with a plurality of nodes, the transition probability among the nodes is obtained, and the word segmentation device with the maximum transition probability among the nodes is used, wherein the expression is as follows:
wherein i represents one of the nodes; representing the next node connected to node i; and P represents the probability between nodes and the product of the probabilities, respectively.
(2) Collecting an external knowledge dictionary, matching the content of the node j with the knowledge dictionary, giving a preset additional score, and embedding the additional score into a word segmentation device, wherein the expression is as follows:
In the method, in the process of the invention, An additional score for node j; if j is a person name, a department name or an arbitrary administrative division name, thenSet to 1, otherwise 0.
(3) And marking the word segmentation in a sentence by using the word segmentation device.
(4) Python libraries, including installations spaCy and scikit-learn; importing the marked sentences into a Python library to execute semantic search, marking the marked sentences by using spaCy components, and vectorizing the marked sentences by using scikit-learn components.
(5) And searching to obtain a plurality of document indexes, and taking the document index with the highest confidence as a reference result to obtain the semantic recognition result of the current sentence.
(6) Repeating the above processes to obtain the semantic recognition results of all sentences, and summarizing the semantic recognition results of all sentences to form a complete structured data packet.
In the automatic community work order circulation method based on the RPA disclosed by the embodiment, in the actual application process, the running logic of the method can be written into a set of executable instructions, and the executable instructions are written into a storage medium and run on the electronic equipment. More specific examples of the computer-readable storage medium mentioned in the present embodiment may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. The computer readable storage medium may include a data signal propagated in baseband or as part of a carrier wave, with readable program code embodied therein. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination of the foregoing. A readable signal medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
As described above, although the present invention has been shown and described with reference to certain preferred embodiments, it is not to be construed as limiting the invention itself. Various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (11)
1. The automatic community work order circulation method based on the RPA is characterized by comprising the following steps:
S1, capturing and identifying service work order information to form a first data packet;
S2, analyzing the first data packet, selecting a service category to which the first data packet belongs, and automatically filling and popping up a service order;
S3, recording and storing the historical steps of the current service order and the data information in a preset storage space in an increment mode of a data log;
s4, creating an encrypted blockchain channel for a data log to which the current service order belongs;
s5, configuring the data log in the block chain channel through a visual operation interface selection mode type to form an RPA task description file;
S6, performing operation scheduling on the RPA task description file, providing an RPA task automation execution context environment, and recording an RPA task operation log;
s7, automatically logging in a third-party service system through preset authorization information, acquiring RPA automatic operation authorization and automatically acquiring service order data;
S8, automatically acquiring authorization page data of the third-party service system, identifying characters of an authorization image or a PDF file in the authorization page data of the third-party service system, extracting semantics of the identified text characters, and finally constructing a complete structured data packet;
S9, automatically classifying pictures and characters in the complete structured data packet, wherein the character classification comprises title details and contents corresponding to the title details; and after checking the corresponding names of the report forms according to the title details and the contents of the corresponding details, automatically filling text data and uploading file data on an authorized page to finish data reporting.
2. The automatic community work order circulation method based on RPA according to claim 1, wherein the service work order information is divided into a written work order and/or an input work order;
For the written worksheet: obtaining a preset number of bad handwriting images, and training the bad handwriting images to obtain an image-text recognition model; extracting service work order information in the written work order by using the image-text recognition model;
for the entered worksheet: and acquiring a complete work order picture in a long screenshot mode, and extracting business work order information in the complete work order picture.
3. The automatic circulation method of community worksheets based on RPA according to claim 2, wherein the obtaining a predetermined number of bad handwriting images, and training the bad handwriting images to obtain a graph-text recognition model, includes:
S1-1, searching a preset number of bad handwriting images based on a search engine, screening out the bad handwriting images with clear characters and no shielding, extracting key features of the bad handwriting images, wherein the key features comprise determined components and strokes, and summarizing the key features to construct a language training set;
s1-2, constructing an image-text recognition model, and training the image-text recognition model based on the language training set;
s1-3, performing OCR (optical character recognition) on a to-be-recognized illegal handwriting image with text blurring and/or text shielding by using the trained text-to-text recognition model to obtain a primary recognition result;
S1-4, extracting recognition components and/or strokes of the illegal handwriting image as source character codes of the image to be recognized, and combining the preliminary recognition results to obtain a recommended recognition result and confidence degrees of all recognition results of the image to be recognized through a picture and text recognition model corresponding to the source character codes of the current image to be recognized;
S1-5, outputting a recommended recognition result with the confidence coefficient higher than a set threshold value and a corresponding confidence coefficient as a recognition result of the image to be recognized.
4. The automatic circulation method of community worksheets based on RPA according to claim 3, wherein the step S1-4 of extracting the recognition components or/and strokes of the scratched handwriting image as the source character code of the image to be recognized comprises:
Taking the identification components and/or strokes of the sloppy handwriting image as a digital image, acquiring a gray level histogram of the digital image, and then carrying out binarization processing on the digital image according to a binarization threshold value;
Removing noise of the binarized digital image, and performing edge smoothing processing on the denoised digital image to obtain a source character code of the image to be identified.
5. The automatic circulation method of community worksheets based on RPA according to claim 3, wherein step S1-3 further comprises:
S1-3-1, calculating to obtain a fusion blind image quality evaluation index value based on the preliminary identification result;
S1-3-2, comparing the fusion blind image quality evaluation index value with a threshold value, wherein the fusion blind image quality evaluation index value is larger than or equal to the threshold value and is a clear target image, and executing the step S1-3-4; otherwise, executing the step S1-3-3 for blurring the target image;
S1-3-3, inputting the confirmed fuzzy target image as a GAN network model, outputting the clear target image as a network, and realizing the definition of the fuzzy target image;
s1-3-4, positioning and identifying the clear target image confirmed in the step S1-3-2 and the clear target image confirmed in the step S1-3-3.
6. The RPA-based community worksheet automatic circulation method of claim 2, wherein for the entered worksheets: acquiring a complete work order picture in a long screenshot mode, and extracting business work order information in the complete work order picture, wherein the method specifically comprises the following steps of:
Selecting the synthesized complete work order picture in a local storage medium;
starting an OCR module to identify characters and patterns in a complete work order picture, and extracting and synthesizing complete work order information, wherein the complete work order information comprises N data item title texts, data item content texts and data item content pictures; wherein N is a natural number greater than 1;
Desensitizing the extracted work order information according to the current filling work order type and the pre-configuration information thereof;
caching the extracted complete work order data work order into DATAARRAY arrays;
The preconfiguration information comprises a work item field name, a field content check format, a field desensitization mode and whether a field is necessary to be filled;
The desensitization processing of the extracted work order information comprises keyword filtering, keyword replacement, content abstract encryption, content symmetric encryption, content asymmetric encryption and picture watermarking.
7. The automatic circulation method of community worksheets based on RPA according to claim 1, wherein step S2 further comprises:
analyzing the first data packet to obtain analysis data; and selecting an item to which the analysis data belongs, selecting a corresponding table according to different items, and writing the analysis data into the corresponding table.
8. The automatic circulation method of community worksheets based on RPA according to claim 1, wherein step S3 further comprises:
S3-1, binding the project by using a background linking unit, producing a business background of the corresponding project, producing a background arrow of a business order, and transmitting the background arrow to a data storage module;
s3-2, the data storage module produces a channel under the arrow and a storage disk connected with the channel, and stores historical steps of service orders and data information.
9. The RPA-based community worksheet automatic circulation method of claim 7, wherein step S4 further comprises:
S4-1, obtaining a background arrow of a service order by an encryption processing module, and producing a private key corresponding to a first data packet;
s4-2, encrypting the first data packet into a second data packet by using the private key, and sending the second data packet to other nodes of the blockchain, wherein the specific process is as follows:
splitting a first data packet into at least two first number domains, filling the first number domains to blank positions of two-dimensional codes generated by private keys, and filling black positions of the two-dimensional codes into association number domains according to an association number domain database;
The association number field comprises at least 24 groups of characters, and each group of character strings consists of any character of 0-9, a-Z and A-Z; namely, 62N combinations of each group of character strings are provided, wherein N is the number of characters of the group of character strings;
Encrypting each character of all character strings by using an elliptic curve function to obtain a character or character string, and sequentially stringing all the characters and character strings to form an associated character string;
Wherein the elliptic curve function is expressed as follows:
y 2＝x3+ax2 +b, wherein a and b are any values;
The association number field is any two groups of combination numbers within 0-9 which are randomly generated based on a binary algorithm.
10. The automatic community work order circulation method based on RPA as claimed in claim 1, wherein the step S8 is characterized in that the semantic extraction is performed on the identified text words, and the complete structured data packet is finally constructed, specifically comprising:
S8-1, executing word segmentation operation based on the authorized image or the words identified in the PDF file, decomposing a sentence into a plurality of paths connected with the nodes, and obtaining the transition probability among the nodes, wherein the word segmentation device with the maximum transition probability among the nodes is represented as follows:
wherein i represents one of the nodes; Representing the next node connected to node i; /(I) And P represents the probability between nodes and the product of the probabilities, respectively;
S8-2, automatically collecting an external knowledge dictionary, matching the content of the node j with the knowledge dictionary, giving a preset additional score, and embedding the additional score into a word segmentation device, wherein the expression is as follows:
In the method, in the process of the invention, An additional score for node j; if j is a person name, a department name or an arbitrary administrative division name, then/>Set to 1, otherwise 0;
s8-3, marking word segmentation in a sentence by using the word segmentation device;
S8-4, installation spaCy and scikit-learn;
S8-5, importing the marked sentences into a Python library to perform semantic search, marking the marked sentences by using spaCy components, and vectorizing the marked sentences by using scikit-learn components;
s8-6, searching to obtain a plurality of document indexes, and taking the document index with the highest confidence as a reference result to obtain a semantic recognition result of the current sentence;
S8-7, repeating the steps S8-5 to S8-6 to obtain the semantic recognition result of all sentences;
And S8-8, summarizing semantic recognition results of all sentences to form a complete structured data packet.
11. An automatic circulation system of community worksheets, comprising:
the data acquisition module is used for capturing service work order information to form a first data packet;
the service processing module is used for analyzing the first data packet, selecting the service category to which the data packet belongs, and automatically filling and popping up a service order;
the data storage module is used for storing historical steps of the business orders and data information;
The encryption processing module is used for creating an encrypted blockchain channel for the service order and encrypting channel data writing and reading of the blockchain channel;
The RPA configuration module is used for configuring data through selecting a mode type through the image visualization operation interface to form an RPA task description file;
The RPA scheduling module is used for performing operation scheduling on the RPA task description file, providing an RPA task automation execution context environment, recording an RPA task operation log, and realizing automatic triggering, abnormal interruption and state rollback of the task;
The RPA automatic login module is used for automatically logging in a third-party service system through preconfigured authorization information, acquiring RPA automatic operation authorization and automatically acquiring service order data;
The RPA data acquisition module automatically acquires the authorization page data of the third party service system, performs text recognition on the third party authorization image and PDF file data through an OCR algorithm, performs semantic extraction on text, and finally constructs a complete structured data packet;
the RPA data filling module is used for automatically classifying pictures and characters in the collected structured data packet, wherein the character classification comprises title details and contents corresponding to the title details, and after the corresponding names of the report are checked according to the title details and the contents corresponding to the title details, text data filling and file data uploading are automatically realized on an authorized page, so that data reporting is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410629691.5A CN118211941A (en) | 2024-05-21 | 2024-05-21 | Automatic community work order circulation method and system based on RPA |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410629691.5A CN118211941A (en) | 2024-05-21 | 2024-05-21 | Automatic community work order circulation method and system based on RPA |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118211941A true CN118211941A (en) | 2024-06-18 |
Family
ID=91454698
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410629691.5A Pending CN118211941A (en) | 2024-05-21 | 2024-05-21 | Automatic community work order circulation method and system based on RPA |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118211941A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110610094A (en) * | 2019-07-25 | 2019-12-24 | 温州医科大学 | College data increment treatment system based on block chain |
| WO2020224253A1 (en) * | 2019-05-07 | 2020-11-12 | 深圳壹账通智能科技有限公司 | Blockchain information pushing method and apparatus, computer device and storage medium |
| CN114117499A (en) * | 2021-12-06 | 2022-03-01 | 中电万维信息技术有限责任公司 | Authority management based trusted data exchange method |
| CN116823422A (en) * | 2023-06-12 | 2023-09-29 | 中国建设银行股份有限公司 | Form data processing method and device |
| CN116894005A (en) * | 2023-07-24 | 2023-10-17 | 山东核电有限公司 | File processing method, device, electronic equipment and storage medium |
-
2024
- 2024-05-21 CN CN202410629691.5A patent/CN118211941A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020224253A1 (en) * | 2019-05-07 | 2020-11-12 | 深圳壹账通智能科技有限公司 | Blockchain information pushing method and apparatus, computer device and storage medium |
| CN110610094A (en) * | 2019-07-25 | 2019-12-24 | 温州医科大学 | College data increment treatment system based on block chain |
| CN114117499A (en) * | 2021-12-06 | 2022-03-01 | 中电万维信息技术有限责任公司 | Authority management based trusted data exchange method |
| CN116823422A (en) * | 2023-06-12 | 2023-09-29 | 中国建设银行股份有限公司 | Form data processing method and device |
| CN116894005A (en) * | 2023-07-24 | 2023-10-17 | 山东核电有限公司 | File processing method, device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102452123B1 (en) | Apparatus for Building Big-data on unstructured Cyber Threat Information, Method for Building and Analyzing Cyber Threat Information | |
| US9245243B2 (en) | Concept-based analysis of structured and unstructured data using concept inheritance | |
| CN107992764B (en) | Sensitive webpage identification and detection method and device | |
| CN109391706A (en) | Domain name detection method, device, equipment and storage medium based on deep learning | |
| CN112749284B (en) | Knowledge graph construction method, device, equipment and storage medium | |
| CN111866004B (en) | Security assessment method, apparatus, computer system, and medium | |
| CN114218391B (en) | Sensitive information identification method based on deep learning technology | |
| CN113806548A (en) | Petition factor extraction method and system based on deep learning model | |
| CN117473512B (en) | Vulnerability risk assessment method based on network mapping | |
| CN116910104B (en) | Construction industry construction safety intelligent log recording method based on large language model | |
| CN111881398A (en) | Page type determination method, device and equipment and computer storage medium | |
| CN114860882A (en) | Fair competition review auxiliary method based on text classification model | |
| CN115545671A (en) | Method and system for structured processing of laws and regulations | |
| CN117454426A (en) | Method, device and system for desensitizing and collecting information of claim settlement data | |
| CN117195319A (en) | Verification method and device for electronic part of file, electronic equipment and medium | |
| CN113971283A (en) | Malicious application program detection method and device based on features | |
| CN114842982B (en) | Knowledge expression method, device and system for medical information system | |
| CN118211941A (en) | Automatic community work order circulation method and system based on RPA | |
| CN115618085A (en) | Interface data exposure detection method based on dynamic label | |
| CN113688346A (en) | Illegal website identification method, device, equipment and storage medium | |
| CN114338058A (en) | Information processing method, device and storage medium | |
| CN117591770B (en) | Policy pushing method and device and computer equipment | |
| CN115082174B (en) | Method, device, computer equipment and storage medium for identifying similar quality control of bonds | |
| CN117763607B (en) | File security grading method, system, equipment and storage medium based on large model | |
| CN112187768B (en) | Method, device and equipment for detecting bad information website and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |