CN114943511A - Government affair office automation platform and optimization implementation method thereof - Google Patents
Government affair office automation platform and optimization implementation method thereof Download PDFInfo
- Publication number
- CN114943511A CN114943511A CN202210539471.4A CN202210539471A CN114943511A CN 114943511 A CN114943511 A CN 114943511A CN 202210539471 A CN202210539471 A CN 202210539471A CN 114943511 A CN114943511 A CN 114943511A
- Authority
- CN
- China
- Prior art keywords
- module
- message
- face
- document
- information
- 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 111
- 238000005457 optimization Methods 0.000 title abstract description 8
- 230000006870 function Effects 0.000 claims abstract description 31
- 238000003058 natural language processing Methods 0.000 claims abstract description 16
- 238000010276 construction Methods 0.000 claims abstract description 14
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 12
- 230000003993 interaction Effects 0.000 claims abstract description 10
- 238000004458 analytical method Methods 0.000 claims abstract description 7
- 238000013500 data storage Methods 0.000 claims abstract description 6
- 230000008569 process Effects 0.000 claims description 47
- 238000012549 training Methods 0.000 claims description 37
- 238000012545 processing Methods 0.000 claims description 27
- 230000001815 facial effect Effects 0.000 claims description 20
- 238000007726 management method Methods 0.000 claims description 16
- 238000012360 testing method Methods 0.000 claims description 13
- 238000010586 diagram Methods 0.000 claims description 10
- 238000013527 convolutional neural network Methods 0.000 claims description 8
- 238000002716 delivery method Methods 0.000 claims description 6
- 230000002452 interceptive effect Effects 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 6
- 230000007935 neutral effect Effects 0.000 claims description 6
- 230000002457 bidirectional effect Effects 0.000 claims description 5
- 238000004891 communication Methods 0.000 claims description 5
- 238000011068 loading method Methods 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 claims description 5
- 238000013145 classification model Methods 0.000 claims description 4
- 230000007246 mechanism Effects 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000012938 design process Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 230000000977 initiatory effect Effects 0.000 claims description 3
- 238000012423 maintenance Methods 0.000 claims description 3
- 239000011159 matrix material Substances 0.000 claims description 3
- 238000011176 pooling Methods 0.000 claims description 3
- 238000013515 script Methods 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000013523 data management Methods 0.000 abstract description 3
- 230000008447 perception Effects 0.000 abstract description 3
- 239000008186 active pharmaceutical agent Substances 0.000 description 15
- 238000005516 engineering process Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 230000002441 reversible effect Effects 0.000 description 2
- 239000005441 aurora Substances 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 210000000554 iris Anatomy 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 239000000243 solution Substances 0.000 description 1
- 239000012086 standard solution Substances 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- 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
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/242—Dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C1/00—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people
- G07C1/10—Registering, indicating or recording the time of events or elapsed time, e.g. time-recorders for work people together with the recording, indicating or registering of other data, e.g. of signs of identity
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Tourism & Hospitality (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Human Computer Interaction (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Software Systems (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Biomedical Technology (AREA)
- Marketing (AREA)
- Multimedia (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Molecular Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Economics (AREA)
- Educational Administration (AREA)
- Bioinformatics & Computational Biology (AREA)
- Primary Health Care (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Development Economics (AREA)
- Evolutionary Biology (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A government affair office automation platform and an optimization realization method thereof belong to the field of electronic government affair office. The office efficiency of the government affair office automation platform is improved. The system comprises an environment construction module based on a JBPM workflow engine, a face recognition attendance checking module, a natural language processing document classification module, a micro-service message pushing module and a docx document online editing module built by an embedded NPAPI plug-in, and a government office automation platform is built by the environment construction module based on the JBPM workflow engine. The invention achieves the purpose of improving the data management capability in an information-based manner on a data storage, interaction and analysis platform; the face recognition attendance checking module is used for accurately positioning a target face based on a deep face recognition algorithm by supporting an autonomous developed non-perception attendance checking function; and automatically, the documents are intelligently generated into classification labels, the messages in government offices are pushed in a micro-service mode, and the docx document online editing module enables the cooperative work to be more efficient.
Description
Technical Field
The invention belongs to the field of electronic government affair office, and particularly relates to a government affair office automation platform and an optimization implementation method thereof.
Background
The government affair office system is a relatively mature system developed in government informatization, and the promotion of the government affair office system enables the government departments to realize paperless office work, greatly improves the working efficiency and also enables the office flow in the government to realize standardization and normalization. Various nodes such as receipt, handling and archiving are arranged in the workflow of the government affair system, each node is a processing link and is provided with a fixed processor, and the processors need to give instructions according to the contents of incoming documents and send tasks to the processors in the next working link.
The work of making approval and next division of labor requires a processor to be very familiar with the door division of labor, to have wide knowledge on business work, to speak rigorously and properly and to handle quickly. The node also becomes a pain point in government office processes, on one hand, all the processes pass through the node, the requirement on time effectiveness is high, and once the delay occurs, many subsequent links are influenced.
Electronic government affairs is one of the most important fields of modern informatization as an organic combination of electronic information technology and management. The electronic government affairs is to apply modern information and communication technology, integrate management and service through network technology, realize the optimization and recombination of organizational structure and workflow on the internet, surpass the separation limit between time, space and departments, and provide high-quality, all-round, standard and transparent management and service according with international standard for society.
Along with the deep popularization of the government affair information system, the data volume of the electronic government affair system is larger and larger, the application functions are more and more, and the application range is wider and wider, so that the optimization design of the government affair office automation platform is urgently needed.
Disclosure of Invention
The invention aims to provide an optimized government affair office automation platform and an optimized implementation method thereof.
In order to achieve the purpose, the invention is realized by the following technical scheme:
a government office automation platform comprises an environment construction module based on a JBPM workflow engine, a face recognition attendance module, a natural language processing document classification module, a micro-service message pushing module and an embedded NPAPI plug-in building docx document online editing module, wherein the government office automation platform is built based on the environment construction module of the JBPM workflow engine;
the environment construction module based on the JBPM workflow engine is used for managing data storage, interaction and analysis and determining personnel information;
the facial recognition attendance module is used for accurately positioning the facial key feature points;
the natural language processing official document classification module is used for training, evaluating and classifying the user-defined text classification corpus;
the micro-service message pushing module is used for pushing a pushing request initiated by a platform message, carrying out weight judgment by the micro-service module, and pushing data of the message to a third-party application according to a weight judgment result of the message request;
the embedded NPAPI plug-in builds a docx document online editing module, is used for accessing a third-party document editor by calling an API mode, returning an IWebBrowser object into the terminal equipment, performing event processing and DOM operation on the IWebBrowser object interface, communicating with a plugin object of NPAPI, and finally connecting with the inside of the terminal equipment through a connecting interface to call the online editor.
The invention relates to an optimized implementation method of a government office automation platform, which comprises the following steps:
step S1, building an environment platform based on a JBPM workflow engine;
step S2, face recognition attendance is realized;
step S3, processing the classification of natural language documents;
step S4, pushing micro service information;
and step S5, editing the embedded NPAPI plug-in online to build a docx document.
Further, the method for building the environment platform based on the JBPM workflow engine in step S1 includes the following steps:
s1.1, a workflow API is responsible for explaining the definition of the process, a process model is defined by the jPdl through a process interactive interface, the jPdl is used for self-defining a logic interface, the data of all design processes are maintained, and the circulation process of the process is promoted;
s1.2, regarding a flow as a UML state diagram by a jPdl self-defining flow, defining the starting and ending of the state diagram and the conversion of the intermediate state by the jPdl, and describing a service flow through graphical flow definition;
s1.3, performing interactive data access management on a Hibernate API management database and an environment platform based on a JBPM workflow engine, receiving an access request of the platform, calling a workflow interface application according to the access request to acquire and process flow data, and sending different data to the database according to different flow data so that the database analyzes different information data and then forwards the information data to different database nodes for execution;
s1.4, monitoring the execution process of the process instance and suspending or recovering the operation of the process instance through a JVMP monitoring process interface;
s1.5, form reuse, database table reuse, field reuse and complex circulation modes are managed through the Tasks plug-in.
Further, the complex circulation mode in step S1.5 includes flow task distribution and summarization, M selects N (M is the number of flow nodes, and N is the number of branches) branches and multiple branches, self-circulation, xor branch selection, xor convergence, and multiple convergence.
Further, the method for realizing attendance checking through facial recognition in step S2 includes the following steps:
s2.1, establishing a face database, using a convolutional neural network for a face photo, wherein the convolutional neural network comprises a plurality of layers of convolutional networks connected with a plurality of layers of depth networks, using double 3 x 3 convolutional kernels for image processing to form a face image information file, and generating face image information into face feature codes and storing the face feature codes in the face database;
s2.2, acquiring a face image of a human face to generate an image, acquiring the human face of a person by adopting a camera, intercepting the image from a video stream, and generating a face image characteristic information file by utilizing a convolution kernel algorithm;
the convolution kernel expression is:
the summation part in the formula is equivalent to solving one-time cross correlation, b is deviation amount, Z l And Z l+1 Represents the convolutional input and output of layer L +1, L l+1 Is z l+1 Z (i, j) corresponds to the pixel of the feature map, where (i, j) is the pixel coordinate, ω is the derivative of the pooling layer, K is the number of channels of the feature map, K is the number of convolution kernels, f is the size of the convolution kernel, x, y are the coordinates of the deviation, s 0 Performing matrix element multiplication summation and superposition deviation quantity on input characteristics in a receptive field, wherein the convolution step length is p and the number of filling layers is p;
and S2.3, searching and matching the extracted facial feature information of the human face and features stored in a human face database by utilizing a multithreading and 1-to-N millisecond operation mode, comparing and acquiring two images of the facial features and the human face database information, setting the height and the height of a threshold value, judging that the human face is the same person if the similarity is greater than the threshold value through standardization processing, and outputting a result.
Further, the step of the normalization process in step S2.3 is:
s2.3.1, correcting the information image in the face database, mainly correcting the face contained in the original image to a uniform size according to the coordinates of human eyes;
s2.3.2, correcting the image size by using bilinear interpolation of gray interpolation, convolving the sampled image by using a trigonometric function, and carrying out face detection on the interpolated image;
s2.3.3, obtaining gray difference value according to corresponding points after graying according to the collected human face characteristics and human face database information and the human face areas of the two images, and carrying out face recognition according to the gray difference value.
Further, the method for processing natural language official document classification in step S3 includes the following steps:
s3.1, establishing a dictionary table: acquiring previous related text information and storing the previous related text information into a database through a bidirectional maximum matching algorithm; marking the text information through manual retrieval, and matching and removing illegal characters by using a regular expression when reading the database;
s3.2, a THUCTC Chinese text classification tool is introduced, a feature dictionary is extracted and constructed, the feature dictionary is constructed by combining a Chi-square dimensionality reduction and tf-idf method, and a Chi-square formalization function is as follows:
where A is the actual value of the tag data, T is the speculative value of the tag data from the program, x 2 A difference routine representing the actual value and the speculative value;
dividing a Chinese text into different vocabularies which are used as basic units for matching feature words in a dictionary table in a training stage, namely when the text is divided into words, firstly dividing the text into a plurality of parts, enabling each part to correspond to the dictionary table one by one, if the words are in the dictionary table, successfully dividing the words, and if not, continuously dividing and matching until the words are successful, and marking the part of speech on the divided vocabularies;
s3.3, model training, namely selecting bigrams as feature units, and performing model training by using LibSVM feature selection; the LibSVM linear kernel function classification method comprises
Wherein P, N, M represents positive, negative and neutral, respectively, C represents category, and represents 0, 1 and 2, respectively;
when the positive probability is the maximum, C is 0, the negative probability is the maximum, C is 1, the neutral probability is the maximum, and C is 2;
verifying by using an n-fold cross method, wherein n-5 sequentially uses 4 parts of 5 equal parts as a training set, and 1 part of data is used as a test set to train the accuracy and the recall rate of the model;
the method also comprises the processing of stop words, wherein the stop words comprise some adverbs, adjectives and some conjunctions;
s3.4, learning of a classification model: dividing the participles into a training set A and a testing set B according to a proportion, using training data A to carry out n-fold verification, dividing the training set into a 1-an, n parts in total, checking each ai in the n parts in turn, using the rest n-1 parts as training, training for n times, taking the average value of errors obtained by the n times as the final error evaluation standard of a model, selecting the optimal model parameter, training a model under the selected optimal model parameter by using the whole training set A, and finally testing the testing set B to give a classification result, thereby achieving classification of natural language processing official documents.
Further, the step S4 of pushing the microservice message includes the following steps:
s4.1, initiating a message sending request, storing messages generated by short-time high concurrency in a queue through asynchronous processing, judging the request weight, and confirming whether the messages are added into a sending queue according to a judgment result;
s4.2, dividing the message push into each small service in a transverse dividing mode, and providing WebSocket message communication for the services by the sending queue;
s4.3, the APIGateway realizes the only entrance of all message queues through an internal maintenance routing table, requests service from a remote computer through a network by using an RPC API remote procedure call protocol interface, and matches the message with each micro-service module;
and S4.4, the third party application micro-service module encapsulates the message and the user information to be sent to the equipment terminal into a character string in a json format, and the character string of the message is connected to the equipment terminal in an opposite mode, so that the micro-service message is delivered.
Further, the message delivery method of the step S4 microservice message pushing module is based on the MVC pattern architecture of the apigataway and the message queue.
Further, the step S5 method for editing the embedded NPAPI plug-in online to build the docx document includes the following steps:
s5.1, a docx document editing function is loaded through an npapi dynamic link library, a document is opened or created through a streaming document software interface,
s5.2, interacting with the browser by using an NPAPI mechanism, returning an IWebBrowser object to the browser, performing event processing on an IWebBrowser object interface, operating a DOM document to communicate with an object model and a plugin object, finally connecting the IWebBrowser object interface with the inside of the browser through a connecting interface, integrating scripts, documents, images and other controls into a Web page, and finishing online document editing;
and S5.3, selecting a loading method of the plug-in corresponding to the NPAPI plug-in according to the type of the loaded document editor by the NPAPI plug-in, loading the bottom API plug-in, converting the plug-in object into a uniform plug-in object, and providing the API for the browser to call by the connecting interface according to the upper browser so as to realize the interaction between the browser and the NPAPI.
The invention has the beneficial effects that:
the government affair office automation platform is a standard solution designed aiming at daily office informatization of public institutions, government departments and the like, is based on a cooperative office concept, electronizes, networks, standardizes and unifies daily office, realizes cross-department and cross-region cooperative office, and achieves the purposes of saving office time, saving office cost and improving work efficiency. With the daily use of cooperative office, a complete information base in a unit is gradually built.
According to the government affair office automation platform, a complicated and low-efficiency office mode is changed, so that personnel in a unit can conveniently and quickly share information and can cooperatively finish work at high speed. By building an environment based on a JBPM workflow engine, a work flow can be designed by a user of an enterprise, and a document node executor can be modified at any time, so that the combination of rigid control and flexible management is achieved. In the document circulation process, the formality, seriousness and normalization of document circulation are supported, and rigid control of document management is realized.
According to the government office automation platform, most of the existing official documents are semi-structural or unstructured data, and in order to quickly and effectively obtain required information, the information of the official documents is classified by using a natural language classification function, so that the purposes of not requiring human resources and reducing classification time are achieved; the non-sensing attendance checking function is a technology for carrying out identity recognition by capturing face characteristic information of a person, can finish the non-perceived recognition only under the condition of a specific light source, and has the characteristics of non-contact property, difficulty in counterfeiting and the like.
The optimization implementation method of the government office automation platform is based on environment construction of a JBPM workflow engine, achieves data storage, interaction and analysis platform, and improves data management capacity in an informationization mode; the face recognition attendance checking module is used for accurately positioning a target face based on a deep face recognition algorithm by supporting an autonomous developed non-perception attendance checking function; the natural language processing document classification module automatically generates classification labels for documents intelligently; the micro-service message pushing module is used for pushing messages in government offices in a micro-service mode; the embedded NPAPI plug-in builds a docx document online editing module, and the cooperative work is more efficient.
Drawings
Fig. 1 is a functional framework structure diagram of a government office automation platform according to the present invention;
FIG. 2 is a flow chart of a construction method of an environment platform based on a JBPM workflow engine according to the invention;
fig. 3 is a flow chart of the identification method of the facial recognition attendance module according to the invention;
FIG. 4 is a flow chart of a method of the natural language processing document classification module according to the present invention;
FIG. 5 is a flowchart of a message delivery method of a microservice message push module according to the present invention;
fig. 6 is a flowchart of an editing method for building a docx document online editing module by using an embedded NPAPI plug-in according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and the detailed description. It is to be understood that the embodiments described herein are illustrative only and are not limiting, i.e., that the embodiments described are only a few embodiments, rather than all, of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations, and the present invention may have other embodiments.
Thus, the following detailed description of specific embodiments of the present invention, presented in the accompanying drawings, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be derived by a person skilled in the art from the detailed description of the invention without inventive step, are within the scope of protection of the invention.
For a further understanding of the contents, features and effects of the invention, reference will now be made to the following detailed description taken in conjunction with the accompanying drawings in which:
the first embodiment is as follows:
a government office automation platform comprises an environment construction module based on a JBPM workflow engine, a face recognition attendance module, a natural language processing document classification module, a micro-service message pushing module and an embedded NPAPI plug-in building docx document online editing module, wherein the government office automation platform is built based on the environment construction module of the JBPM workflow engine;
the environment construction module based on the JBPM workflow engine is used for managing data storage, interaction and analysis;
the facial recognition attendance module is used for accurately positioning the facial key feature points and determining personnel information;
the natural language processing official document classification module is used for training, evaluating and classifying the user-defined text classification corpus;
the micro-service message pushing module is used for pushing a pushing request initiated by the platform message, performing weight judgment by the micro-service module, and pushing data which is the message according to a message request weight judgment result to a third-party application;
the embedded NPAPI plug-in builds a docx document online editing module which is used for accessing a third-party document editor in an API calling mode, returning an IWebbrowser object into the terminal equipment, enabling an IWebbrowser object interface to perform event processing, operate DOM, communicate with a plugin object of NPAPI, and finally connect with the inside of the terminal equipment through a connection interface to call the online editor.
Furthermore, the environment construction based on the JBPM (Java Business Process management) workflow engine uses a flexible and extensible workflow management engine as the business process input by the server during JBPM operation, and uses simple and powerful language to express and package the business process in a process file. JBPM combines the convenience of workflow application development with outstanding Enterprise Application Integration (EAI) capabilities.
Furthermore, the face recognition attendance module is an independently developed face recognition algorithm, adopts a progressive self-coding network mode, effectively prevents overfitting, improves the stability of various occasions, realizes the accurate positioning of key feature points of rapid multi-face parts, and can achieve millisecond-level positioning without accuracy loss.
Furthermore, the natural language processing document classification module is a THUCTC Chinese text classification tool, and can automatically and efficiently realize the training, evaluating and classifying functions of the user-defined text classification corpus. The text classification generally comprises three steps of feature selection, feature dimension reduction and classification model learning.
Furthermore, the micro-service message pushing module is used for carrying out weight judgment on a pushing request initiated by the platform message, and pushing data of the message to a third-party application (such as applications of 'WeChat', 'nail', 'aurora pushing' and the like) service module according to a judgment result of the weight of the message request, so that the functions of PC (personal computer) end release and terminal equipment (such as a computer, a mobile phone, a tablet and the like) browsing are realized.
Furthermore, the embedded npapi (netscape plug Application Programming interface) plug-in builds a docx document online editing module, accesses a third-party document editor (such as WPS, perpetual, OpenOffice, NTKO Office and the like) by calling an API mode, returns an IWebBrowser object into the terminal device, performs event processing and DOM operation on the IWebBrowser object interface, communicates with the plug-in object of npapi, and is finally connected with the terminal device through a connection interface, so as to realize calling the online editor function.
Fig. 1 is a functional framework structure diagram of the government office automation platform.
The second embodiment is as follows:
according to a specific embodiment, the method for optimizing the government office automation platform includes the following steps:
step S1, constructing an environment platform based on the JBPM workflow engine:
s1.1, a workflow API is responsible for explaining the definition of the process, a process model is defined by the jPdl through a process interactive interface, the jPdl is used for self-defining a logic interface, the data of all design processes are maintained, and the circulation process of the process is promoted;
s1.2, a jPdl self-defining flow regards a flow as a UML state diagram, the jPdl defines the starting and ending of the state diagram and the conversion of the intermediate state, and a service flow is described through graphical flow definition;
s1.3, performing interactive data access management on a Hibernate API management database and an environment platform based on a JBPM workflow engine, receiving an access request of the platform, calling a workflow interface application according to the access request to acquire and process flow data, and sending different data to the database according to different flow data so that the database analyzes different information data and then forwards the information data to different database nodes for execution;
s1.4, monitoring the execution process of the process instance and suspending or resuming the operation of the process instance through a JVMP monitoring process interface;
s1.5, form reuse, database table reuse, field reuse and complex circulation modes are managed through the Tasks plug-in.
Further, the complex circulation mode in step S1.5 includes flow task distribution and summarization, M selects N (M is the number of flow nodes, and N is the number of branches) branches and multiple branches, self-circulation, xor branch selection, xor convergence, and multiple convergence.
Furthermore, the JBPM workflow engine-based environment is built, and the JBPM takes the current popular Hibernate as a persistent layer thereof, so that the storage, interaction and analysis of data are managed more conveniently. A flow is regarded as a UML state diagram by adopting a self-defined jBPM Process definition language (jPdl). jPdl defines each part of the state diagram in detail (eg: the transition of the initial, end and intermediate states), visually describes the business process through the graphic process definition, and supports the execution of the tasks related to multiple persons or multiple departments; the workflow management can generate and process an electronic form of a task, and personnel of each department can interact through the form, so that the personnel can participate in the whole business process.
Step S2, realizing facial recognition attendance:
s2.1, establishing a face database, using a convolutional neural network for a face picture, wherein the convolutional neural network comprises a plurality of layers of convolutional networks and a plurality of layers of depth networks, using double 3 x 3 convolutional kernels for image processing to form a face image information file, and generating face image information into face feature codes and storing the face image information into the face database;
s2.2, acquiring a face image of a human face to generate an image, acquiring the human face of a person by adopting a camera, intercepting the image from a video stream, and generating a face image characteristic information file by utilizing a convolution kernel algorithm;
the convolution kernel expression is:
the summation part in the formula is equivalent to solving one-time cross correlation, b is deviation amount, Z l And Z l+1 Represents the convolutional input and output of layer L +1, L l+1 Is z l+1 Z (i, j) corresponds to the pixel of the feature map, wherein (i, j) is the pixel coordinate, omega is the derivative of the pooling layer, K is the number of channels of the feature map, K is the number of convolution kernels, f is the size of the convolution kernels, x, y are the coordinates of the deviation, s 0 Performing matrix element multiplication summation and superposition deviation quantity on the input characteristics in the receptive field for convolution step length and p for the number of filling layers;
and S2.3, searching and matching the extracted facial feature information of the human face and features stored in a human face database by utilizing multithreading and a 1-N millisecond operation mode, comparing and acquiring two images of the facial features and the information of the human face database, setting the threshold value high and low, judging that the human face is the same person if the similarity is greater than the threshold value high through standardization processing, and outputting a result.
Further, in the method for implementing optimization of the government office automation platform, the step of standardized processing in step S2.3 is as follows:
s2.3.1, correcting the information image in the face database, mainly correcting the face contained in the original image to a uniform size according to the coordinates of human eyes;
s2.3.2, correcting the image size by using bilinear interpolation of gray interpolation, convolving the sampled image by using a trigonometric function, and carrying out face detection on the interpolated image;
s2.3.3, obtaining gray difference value according to corresponding points after graying according to the collected human face characteristics and human face database information and the human face areas of the two images, and carrying out face recognition according to the gray difference value.
Furthermore, the convolutional neural network can process multidimensional data and perform 360 key point positioning on the facial picture data, wherein the 360 key point positioning comprises characteristic extraction of attributes such as the size, the position, the distance and the like of facial image facial contours such as eye irises, nasal wings, mouth corners and the like.
Further, the Receptive Field (received Field) is the size of the area where the pixels on the feature map output by each layer of the convolutional neural network are mapped on the input map, that is, one point on the feature map corresponds to the area on the input map
Further, before graying is carried out in the standardization processing process, the human face areas of the two images are adjusted to be uniform in size according to a certain ratio, meanwhile, hair and other influences are ignored, and the human face areas are reduced as far as possible to reduce errors with a threshold value, so that the face recognition function is achieved. The gray interpolation is to use the gray values of the known neighboring pixels of the image (or the tristimulus values in the rgb image) to generate the values of the unknown pixels. The bilinear interpolation is to convolute the sampled image by a trigonometric function.
Furthermore, the facial recognition attendance module relies on the self-developed non-sensing attendance function, compares after the high-precision snapshot of the intelligent AI camera, avoids any unnecessary data delay, realizes the extremely fast second signature, deals with the complex conditions of various attendance schemes, and the attendance management is more efficient. By matching with an advanced organization architecture and an attendance group management mechanism, the non-sensing attendance system acquires the change of the organization personnel in real time and updates the attendance personnel characteristics in an incremental manner.
Step S3, the processing of natural language document classification includes the following steps:
s3.1, establishing a dictionary table: acquiring previous related text information and storing the previous related text information into a database through a bidirectional maximum matching algorithm; marking the text information through manual retrieval, and matching and removing illegal characters by using a regular expression when reading the database;
s3.2, referring to a THUCTC Chinese text classification tool, extracting and constructing a feature dictionary, wherein the feature dictionary is constructed by combining a Chi-square reduction and tf-idf method, and a Chi-square formalization function is as follows:
where A is the actual value of the marked data, T is the speculative value of the program to the marked data, x 2 A difference program representing the actual value and the speculative value;
dividing a Chinese text into different vocabularies which are used as basic units for matching feature words in a dictionary table in a training stage, namely when the text is divided into words, firstly dividing the text into a plurality of parts, enabling each part to correspond to the dictionary table one by one, if the words are in the dictionary table, successfully dividing the words, and if not, continuously dividing and matching until the words are successful, and marking the part of speech on the divided vocabularies;
s3.3, model training, namely selecting bigrams as feature units, and performing model training by using LibSVM feature selection; the LibSVM linear kernel function classification method comprises
Wherein P, N, M represents positive, negative and neutral, respectively, C represents category, and represents 0, 1 and 2, respectively;
when the positive probability is the maximum, C is 0, the negative probability is the maximum, C is 1, the neutral probability is the maximum, and C is 2;
verifying by using an n-fold cross method, wherein n-5 sequentially uses 4 parts of 5 equal parts as a training set, and 1 part of data is used as a test set to train the accuracy and the recall rate of the model;
the method also comprises the processing of stop words, wherein the stop words comprise some adverbs, adjectives and some conjunctions;
s3.4, learning a classification model: dividing the participles into a training set A and a testing set B according to a proportion, using training data A to carry out n-fold verification, dividing the training set into a 1-an, n parts in total, checking each ai in the n parts in turn, using the rest n-1 parts as training, training for n times, taking the average value of errors obtained by the n times as the final error evaluation standard of a model, selecting the optimal model parameter, training a model under the selected optimal model parameter by using the whole training set A, and finally testing the testing set B to give a classification result, thereby achieving classification of natural language processing official documents.
Furthermore, the natural language processing official document classification module finds a relation model between text features and text categories according to a labeled training document set (dictionary table), and then performs category judgment on the text by using the relation model obtained by learning; in English segmentation, a blank space is used as a natural delimiter among words, different from English, Chinese natural language does not have a formal delimiter, and a dictionary table is established by a bidirectional maximum matching algorithm on the basis of Chinese text classification. The bidirectional maximum matching algorithm is used for carrying out left-to-right and right-to-left scanning word segmentation on the text through forward maximum matching and reverse maximum matching. The positive maximum match is from the first Chinese character on the left to the right, each time an attempt is made to match a word present in the dictionary table, and then the next word is continued to be matched. This process does not require looking up a single word in the dictionary table at a time, and can use a hash table (hash table) or a trie for efficient matching. The reverse maximum match is then a single word in the dictionary table looking up from the end of the right text to the left.
Step S4, pushing the micro service message, the message delivery method of the micro service message pushing module includes the following steps:
s4.1, initiating a message sending request, storing messages generated by short-time high concurrency in a queue through asynchronous processing, judging the request weight, and confirming whether the messages are added into the sending queue according to the judgment result;
s4.2, dividing the message push into each small service in a transverse dividing mode, and providing WebSocket message communication for the services by the sending queue;
s4.3, the APIGateway realizes the only entrance of all message queues through an internal maintenance routing table, requests service from a remote computer through a network by using an RPC API remote procedure call protocol interface, and matches the message with each micro-service module;
and S4.4, the third party application micro-service module encapsulates the message and the user information to be sent to the equipment terminal into a character string in a json format, and the character string of the message is connected to the equipment terminal in an opposite mode, so that the micro-service message is delivered.
Further, the message delivery method of the step S4 microservice message pushing module is based on the MVC pattern architecture of the APIGateway and the message queue.
Further, the micro-service message pushing module pushes the message to be sent to each functional service by combining the micro-service mode. The method aims at dividing the function of assisting office application into each small service, the services are mutually related, so that one point generates information, multiple points browse, and the method has a diversified mode for supporting the common use of equipment such as a computer, a mobile phone, a tablet and the like.
Step S5, editing the docx document on line by the embedded NPAPI plug-in, wherein the method for editing the docx document on-line editing module built by the embedded NPAPI plug-in comprises the following steps:
s5.1, a docx document editing function is loaded through an npapi dynamic link library, a document is opened or created through a streaming document software interface,
s5.2, interacting with the browser by using an NPAPI mechanism, returning an IWebBrowser object to the browser, carrying out event processing on an IWebBrowser object interface, operating a DOM document to carry out communication on an object model and a plugin object, finally connecting the IWebBrowser object interface with the inside of the browser through a connecting interface, and integrating a script, a document, an image and other controls into a Web page to finish online document editing;
and S5.3, selecting a loading method of the plug-in corresponding to the NPAPI plug-in according to the type of the loaded document editor by the NPAPI plug-in, loading the bottom API plug-in, converting the plug-in object into a uniform plug-in object, and providing the API for the browser to call by the connecting interface according to the upper browser so as to realize the interaction between the browser and the NPAPI.
Furthermore, the embedded NPAPI plug-in builds a docx document online editing module, refers to the current mainstream NPAPI plug-in, embeds the third-party text editor function into the government office automation platform, and is managed by the NPAPI dynamic link library to realize online editing of the docx document. The embedded docx text editor function keeps the habit of a user for editing the document and increases the efficiency of cooperative work.
Furthermore, the embedded NPAPI plug-in builds a docx document online editing module, a secondary development technical scheme can be realized, a user-defined method judges whether images, characters and the like are automatically loaded, the NPAPI plug-in comprises various APIs, and various document storage functions (synchronous, asynchronous, landing, non-landing and remote storage modes) can be realized.
Furthermore, the method for building the environment platform based on the JBPM workflow engine achieves a big data storage, interaction and analysis platform customized for the E-government affairs, and improves the data management capability in an informationized manner.
Furthermore, the recognition method of the face recognition attendance module depends on the self-researched non-perception attendance function, and based on the depth face recognition algorithm, the target face is accurately positioned, and more than 360 key point positions are brushed every second.
Furthermore, the processing method of the natural language processing official document classification module automatically and intelligently generates a classification label for uploading official documents, and establishes a dictionary table by using the previous related text information through embedding a THUCTC Chinese text classification tool. When a text is segmented, firstly, the text is segmented into a plurality of parts, each part corresponds to a dictionary table one by one, if the word is in the dictionary table, the segmentation is successful, otherwise, the segmentation and matching are continued until the word is successful. The method comprises the steps of evaluating the importance degree of a text to one of documents in a document set or a corpus, matching according to the characteristics of the text, selecting an optimal matching result, and judging the category of the text according to the fact that whether the text has words with the same category or not and then adding synonyms for comparison. The natural language processing official document classification module has good universality on long texts and has the advantages of high accuracy and high test speed.
Further, a message delivery method of a micro-service message pushing module is characterized in that messages in government offices are pushed in a micro-service mode, a message pushing function is divided into a group of small services in a transverse division mode from a public and independent function dimension, namely, in a mode that the public is called by a plurality of other services and dependent resources are independent and not coupled with other modules, request information is firstly acquired among the services to be added into a sending queue, then, sending requests are sequentially generated by an API gateway according to the route of the sending queue, and then, the RPC is used for carrying out pushing tasks according to the sending requests to carry out mutual coordination and cooperation so as to provide diversified functions for users.
Furthermore, the embedded NPAPI plug-in builds an editing method of the docx document online editing module, so that the cooperative work is more efficient, the use is more convenient and faster, and the modification records can be immediately stored and saved.
It is noted that relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
While the application has been described above with reference to specific embodiments, various modifications may be made and equivalents may be substituted for elements thereof without departing from the scope of the application. In particular, the various features of the embodiments disclosed herein may be used in any combination that is not inconsistent with the structure, and the failure to exhaustively describe such combinations in this specification is merely for brevity and resource conservation. Therefore, it is intended that the application not be limited to the particular embodiments disclosed, but that the application will include all embodiments falling within the scope of the appended claims.
Claims (10)
1. The utility model provides a government affairs official working automation platform which characterized in that: the system comprises an environment construction module based on a JBPM workflow engine, a face recognition attendance module, a natural language processing document classification module, a micro-service message pushing module and an embedded NPAPI plug-in building docx document online editing module, wherein the government office automation platform is built based on the environment construction module of the JBPM workflow engine;
the environment construction module based on the JBPM workflow engine is used for managing data storage, interaction and analysis;
the facial recognition attendance module is used for accurately positioning the facial key feature points and determining personnel information;
the natural language processing official document classification module is used for training, evaluating and classifying the user-defined text classification corpus;
the micro-service message pushing module is used for pushing a pushing request initiated by a platform message, carrying out weight judgment by the micro-service module, and pushing data of the message to a third-party application according to a weight judgment result of the message request;
the embedded NPAPI plug-in builds a docx document online editing module, is used for accessing a third-party document editor by calling an API mode, returning an IWebBrowser object into the terminal equipment, performing event processing and DOM operation on the IWebBrowser object interface, communicating with a plugin object of NPAPI, and finally connecting with the inside of the terminal equipment through a connecting interface to call the online editor.
2. An optimized implementation method of a government office automation platform according to claim 1, characterized by comprising the following steps: the method comprises the following steps:
step S1, building an environment platform based on the JBPM workflow engine;
step S2, face recognition attendance is realized;
step S3, processing the classification of natural language documents;
step S4, pushing micro service information;
and step S5, editing the embedded NPAPI plug-in online to build a docx document.
3. The method for optimizing the government office automation platform according to claim 2, wherein the method comprises the following steps: the method for building the environment platform based on the JBPM workflow engine in the step S1 comprises the following steps:
s1.1, a workflow API is responsible for explaining the definition of the process, a process model is defined by the jPdl through a process interactive interface, the jPdl is used for self-defining a logic interface, the data of all design processes are maintained, and the circulation process of the process is promoted;
s1.2, a jPdl self-defining flow regards a flow as a UML state diagram, the jPdl defines the starting and ending of the state diagram and the conversion of the intermediate state, and a service flow is described through graphical flow definition;
s1.3, performing interactive data access management on a Hibernate API management database and an environment platform based on a JBPM workflow engine, receiving an access request of the platform, calling a workflow interface application according to the access request to acquire and process flow data, and sending different data to the database according to different flow data so that the database analyzes different information data and then forwards the information data to different database nodes for execution;
s1.4, monitoring the execution process of the process instance and suspending or resuming the operation of the process instance through a JVMP monitoring process interface;
s1.5, managing form reuse, database table reuse, field reuse and complex circulation modes through Tasks plug-ins.
4. The optimized implementation method of the government office automation platform according to claim 3, characterized in that: the complex circulation mode in step S1.5 includes flow task distribution and summarization, M selects N branches and multiple branches, self-circulation, exclusive or branch selection, exclusive or convergence, and multiple convergence, where M is the number of flow nodes and N is the number of branches.
5. The optimized implementation method of the government office automation platform according to claim 2, characterized in that: the method for realizing the facial recognition attendance in the step S2 comprises the following steps:
s2.1, establishing a face database, using a convolutional neural network for a face photo, wherein the convolutional neural network comprises a plurality of layers of convolutional networks connected with a plurality of layers of depth networks, using double 3 x 3 convolutional kernels for image processing to form a face image information file, and generating face image information into face feature codes and storing the face feature codes in the face database;
s2.2, acquiring a face image of a human face to generate an image, acquiring the human face of a person by adopting a camera, intercepting the image from a video stream, and generating a face image characteristic information file by utilizing a convolution kernel algorithm;
wherein the convolution kernel expression is:
(i，j)∈(0，1，...L l+1 }
the summation part in the formula is equivalent to solving one-time cross correlation, b is deviation amount, Z l And Z l+1 Represents the convolutional input and output of layer L +1, L l+1 Is z l+1 Z (i, j) corresponds to the pixel of the feature map, where (i, j) is the pixel coordinate, ω is the derivative of the pooling layer, K is the number of channels of the feature map, K is the number of convolution kernels, f is the size of the convolution kernel, x, y are the coordinates of the deviation, s 0 Performing matrix element multiplication summation and superposition deviation quantity on input characteristics in a receptive field, wherein the convolution step length is p and the number of filling layers is p;
and S2.3, searching and matching the extracted facial feature information of the human face and features stored in a human face database by utilizing multithreading and a 1-N millisecond operation mode, comparing and acquiring two images of the facial features and the information of the human face database, setting the threshold value high and low, judging that the human face is the same person if the similarity is greater than the threshold value high through standardization processing, and outputting a result.
6. The optimized implementation method of the government office automation platform according to claim 5, wherein the method comprises the following steps: the step of the standardization process in step S2.3 is:
s2.3.1, correcting the information image in the face database, mainly correcting the face contained in the original image to a uniform size according to the coordinates of human eyes;
s2.3.2, correcting the image size by using bilinear interpolation of gray interpolation, convolving the sampled image by using a trigonometric function, and carrying out face detection on the interpolated image;
s2.3.3, obtaining gray difference value according to corresponding points after graying according to the collected human face characteristics and human face database information and the human face areas of the two images, and carrying out face recognition according to the gray difference value.
7. The method for optimizing the government office automation platform according to claim 2, wherein the method comprises the following steps: the method for processing natural language official document classification in the step S3 comprises the following steps:
s3.1, establishing a dictionary table: acquiring previous related text information and storing the previous related text information into a database through a bidirectional maximum matching algorithm; marking the text information through manual retrieval, and matching and removing illegal characters by using a regular expression when reading the database;
s3.2, a THUCTC Chinese text classification tool is introduced, a feature dictionary is extracted and constructed, the feature dictionary is constructed by combining a Chi-square dimensionality reduction and tf-idf method, and a Chi-square formalization function is as follows:
where A is the actual value of the tag data, T is the speculative value of the tag data from the program, x 2 A difference program representing the actual value and the speculative value;
dividing a Chinese text into different vocabularies which are used as basic units for matching feature words in a dictionary table in a training stage, namely when the text is divided into words, firstly dividing the text into a plurality of parts, enabling each part to correspond to the dictionary table one by one, if the words are in the dictionary table, successfully dividing the words, and if not, continuously dividing and matching until the words are successful, and marking the part of speech on the divided vocabularies;
s3.3, model training, namely selecting bigrams as feature units, and performing model training by using LibSVM feature selection; the LibSVM linear kernel function classification method comprises
Wherein P, N, M represents positive, negative and neutral, respectively, C represents category, and represents 0, 1 and 2, respectively;
when the positive probability is the maximum, C is 0, the negative probability is the maximum, C is 1, the neutral probability is the maximum, and C is 2;
verifying by using an n-fold cross method, wherein n-5 sequentially uses 4 parts of 5 equal parts as a training set, and 1 part of data is used as a test set to train the accuracy and the recall rate of the model;
the method also comprises the processing of stop words, wherein the stop words comprise some adverbs, adjectives and some conjunctions thereof;
s3.4, learning of a classification model: the method comprises the steps of dividing word segmentation into a training set A and a testing set B according to a proportion, using training data A to carry out n-fold verification, dividing the training set into a 1-an, n parts in total, checking each ai in the n parts in turn, using the rest n-1 parts as training, carrying out n times of training, taking the error average value obtained in n times as the final error evaluation standard of a model, selecting the optimal model parameter, using the whole training set A to train a model under the selected optimal model parameter, and finally testing the testing set B to give a classification result, thereby achieving natural language processing official document classification.
8. The method for optimizing the government office automation platform according to claim 2, wherein the method comprises the following steps: the method for pushing the micro-service message in the step S4 comprises the following steps:
s4.1, initiating a message sending request, storing messages generated by short-time high concurrency in a queue through asynchronous processing, judging the request weight, and confirming whether the messages are added into the sending queue according to the judgment result;
s4.2, dividing the message push into each small service in a transverse dividing mode, and providing WebSocket message communication for the services by the sending queue;
s4.3, the APIGateway realizes the only entrance of all message queues through an internal maintenance routing table, requests service from a remote computer through a network by using an RPC API remote procedure call protocol interface, and matches the message with each micro-service module;
and S4.4, the third party application micro-service module encapsulates the message and the user information to be sent to the equipment terminal into a character string in a json format, and the character string of the message is connected to the equipment terminal in an opposite mode, so that the micro-service message is delivered.
9. The method for optimizing the government office automation platform according to claim 8, wherein the method comprises the following steps: the message delivery method of the step S4 microservice message push module is based on the MVC pattern architecture of APIGateway and message queue.
10. The method for optimizing the government office automation platform according to claim 2, wherein the method comprises the following steps: the method for editing the embedded NPAPI plug-in on line to build the docx document in the step S5 comprises the following steps:
s5.1, a docx document editing function is loaded through an npapi dynamic link library, a document is opened or created through a streaming document software interface,
s5.2, interacting with the browser by using an NPAPI mechanism, returning an IWebBrowser object to the browser, performing event processing on an IWebBrowser object interface, operating a DOM document to communicate with an object model and a plugin object, finally connecting the IWebBrowser object interface with the inside of the browser through a connecting interface, integrating scripts, documents, images and other controls into a Web page, and finishing online document editing;
s5.3, the NPAPI plug-in selects a loading method of the plug-in corresponding to the loaded document editor type according to the loaded document editor type, loads the bottom API plug-in, converts the plug-in object into a uniform plugin object, provides the API for the browser to call according to the upper browser through the connecting interface, and achieves interaction of the browser and the NPAPI.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210539471.4A CN114943511A (en) | 2022-05-17 | 2022-05-17 | Government affair office automation platform and optimization implementation method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210539471.4A CN114943511A (en) | 2022-05-17 | 2022-05-17 | Government affair office automation platform and optimization implementation method thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114943511A true CN114943511A (en) | 2022-08-26 |
Family
ID=82907017
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210539471.4A Pending CN114943511A (en) | 2022-05-17 | 2022-05-17 | Government affair office automation platform and optimization implementation method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114943511A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116050802A (en) * | 2023-02-21 | 2023-05-02 | 深圳市神州通在线科技有限公司 | Enterprise collaborative office application platform based on image analysis |
| CN116308220A (en) * | 2023-05-25 | 2023-06-23 | 北京联讯星烨科技有限公司 | Online debugging optimization method and system for workflow data |
| CN116882945A (en) * | 2023-09-05 | 2023-10-13 | 圣奥科技股份有限公司 | Collaborative office method, equipment and medium based on office area station information |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109981716A (en) * | 2017-12-28 | 2019-07-05 | 北京奇虎科技有限公司 | A kind of micro services call method and device |
| CN110309303A (en) * | 2019-05-22 | 2019-10-08 | 浙江工业大学 | A kind of judicial dispute data visualization analysis method based on Weighted T F-IDF |
| CN110414305A (en) * | 2019-04-23 | 2019-11-05 | 苏州闪驰数控系统集成有限公司 | Artificial intelligence convolutional neural networks face identification system |
| CN110489749A (en) * | 2019-08-07 | 2019-11-22 | 北京航空航天大学 | Intelligent Office-Automation System Work Flow Optimizing |
| CN112214262A (en) * | 2020-12-09 | 2021-01-12 | 南京中孚信息技术有限公司 | Method and system for realizing online content marking of document based on browser plug-in |
| WO2021008567A1 (en) * | 2019-07-18 | 2021-01-21 | 深圳前海微众银行股份有限公司 | Request transmission method and apparatus based on full duplex communication protocol |
-
2022
- 2022-05-17 CN CN202210539471.4A patent/CN114943511A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109981716A (en) * | 2017-12-28 | 2019-07-05 | 北京奇虎科技有限公司 | A kind of micro services call method and device |
| CN110414305A (en) * | 2019-04-23 | 2019-11-05 | 苏州闪驰数控系统集成有限公司 | Artificial intelligence convolutional neural networks face identification system |
| CN110309303A (en) * | 2019-05-22 | 2019-10-08 | 浙江工业大学 | A kind of judicial dispute data visualization analysis method based on Weighted T F-IDF |
| WO2021008567A1 (en) * | 2019-07-18 | 2021-01-21 | 深圳前海微众银行股份有限公司 | Request transmission method and apparatus based on full duplex communication protocol |
| CN110489749A (en) * | 2019-08-07 | 2019-11-22 | 北京航空航天大学 | Intelligent Office-Automation System Work Flow Optimizing |
| CN112214262A (en) * | 2020-12-09 | 2021-01-12 | 南京中孚信息技术有限公司 | Method and system for realizing online content marking of document based on browser plug-in |
Non-Patent Citations (1)
| Title |
|---|
| 袁雄伟: "基于JBPM的高可靠性电子政务系统的研究和实现", 中国优秀硕士学位论文全文数据库-信息科技辑, no. 1, 15 January 2015 (2015-01-15), pages 138 - 416 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116050802A (en) * | 2023-02-21 | 2023-05-02 | 深圳市神州通在线科技有限公司 | Enterprise collaborative office application platform based on image analysis |
| CN116050802B (en) * | 2023-02-21 | 2024-04-02 | 深圳市神州通在线科技有限公司 | Enterprise collaborative office application platform based on image analysis |
| CN116308220A (en) * | 2023-05-25 | 2023-06-23 | 北京联讯星烨科技有限公司 | Online debugging optimization method and system for workflow data |
| CN116308220B (en) * | 2023-05-25 | 2023-08-15 | 北京联讯星烨科技有限公司 | Online debugging optimization method and system for workflow data |
| CN116882945A (en) * | 2023-09-05 | 2023-10-13 | 圣奥科技股份有限公司 | Collaborative office method, equipment and medium based on office area station information |
| CN116882945B (en) * | 2023-09-05 | 2023-12-26 | 圣奥科技股份有限公司 | Collaborative office method, equipment and medium based on office area station information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11055527B2 (en) | System and method for information extraction with character level features | |
| CN110020424B (en) | Contract information extraction method and device and text information extraction method | |
| CN114943511A (en) | Government affair office automation platform and optimization implementation method thereof | |
| US20200004815A1 (en) | Text entity detection and recognition from images | |
| US11687570B2 (en) | System and method for efficient multi-relational entity understanding and retrieval | |
| CN111177532A (en) | Vertical search method, device, computer system and readable storage medium | |
| WO2024098533A1 (en) | Image-text bidirectional search method, apparatus and device, and non-volatile readable storage medium | |
| CN111098312A (en) | Window government affairs service robot | |
| CN112036906A (en) | Data processing method, device and equipment | |
| CN111783471B (en) | Semantic recognition method, device, equipment and storage medium for natural language | |
| CN113051914A (en) | Enterprise hidden label extraction method and device based on multi-feature dynamic portrait | |
| CN113986864A (en) | Log data processing method and device, electronic equipment and storage medium | |
| WO2023142451A1 (en) | Workflow generation methods and apparatuses, and electronic device | |
| CN115730597A (en) | Multi-level semantic intention recognition method and related equipment thereof | |
| CN112883980A (en) | Data processing method and system | |
| CN115114419A (en) | Question and answer processing method and device, electronic equipment and computer readable medium | |
| CN111078855A (en) | Information processing method, information processing device, electronic equipment and storage medium | |
| CN112199954A (en) | Disease entity matching method and device based on voice semantics and computer equipment | |
| US20230004715A1 (en) | Method and apparatus for constructing object relationship network, and electronic device | |
| CN116030375A (en) | Video feature extraction and model training method, device, equipment and storage medium | |
| CN113722584A (en) | Task pushing method and device and storage medium | |
| CN113139751A (en) | Method for determining micro-service user service type based on big data | |
| US11157538B2 (en) | System and method for generating summary of research document | |
| CN113656579A (en) | Text classification method, device, equipment and medium | |
| CN114118072A (en) | Document structuring method and device, electronic equipment and computer readable storage medium |
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 |