CN115238710A - Intelligent document generation and management method and device - Google Patents
Intelligent document generation and management method and device Download PDFInfo
- Publication number
- CN115238710A CN115238710A CN202211154504.XA CN202211154504A CN115238710A CN 115238710 A CN115238710 A CN 115238710A CN 202211154504 A CN202211154504 A CN 202211154504A CN 115238710 A CN115238710 A CN 115238710A
- Authority
- CN
- China
- Prior art keywords
- document
- information
- documents
- data
- user
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/93—Document management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Business, Economics & Management (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Machine Translation (AREA)
Abstract
The invention discloses a method and a device for intelligently generating and managing documents, which are characterized in that related candidate document data information is respectively retrieved from a local database, a professional database and a network database based on user requirements and is simultaneously and respectively input into a RNN (radio network node) and LSTM (local node module) -based document generation module, a first content document and a second content document are intelligently generated, and a final high-quality document is output and stored according to automatic evaluation or manual evaluation. In addition, when the local database can not retrieve accurate document information, the server automatically converts the document information into document generation operation, and the efficiency of intelligently generating the document is further improved.
Description
Technical Field
The present application relates to the field of new generation information processing technologies, and in particular, to a method and an apparatus for intelligently generating and managing documents.
Background
In the existing automatic document generation method and system, most of the existing automatic document generation methods and systems are applied to statistics and calculation, namely, a document template is given, and a certain specific service is processed, counted or calculated to obtain a document; or a document template is designed in advance according to a business rule, relevant information is extracted according to certain rules in specific business and is filled, and a document is generated, for example, a score sheet in office automation or report generation in other fields has the technical problems of insufficient intellectualization, single generation rule, single data source acquisition and the like, and generally cannot meet the actual requirements of users.
In addition, although many professional office software (such as financial software, tax professional software, legal management system, etc.) exist, including resource management systems such as ERP and the like, and large and small ERP derivative products such as enterprise management systems, documents generated by the software are basically highly structured messages, forms, or calculations of some formulas, document templates are made in advance and are not changeable, the accuracy of document generation is low, the efficiency is low, and retrieval and query in the management process are also very inconvenient.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides the method and the device for intelligently generating and managing the document, which can intelligently generate the reference document according to the requirements of the user and improve the accuracy and the efficiency of document generation. In order to realize the purpose, the invention provides the following technical scheme:
a method for intelligently generating and managing documents comprises the following steps:
(1) A user inputs and generates document task requirement information through a client and submits the task requirement information to a server;
further, the document task generating requirement information is input text information or voice information, wherein the voice information needs to be further converted into text information so as to facilitate subsequent semantic analysis;
(2) The server executes semantic analysis on the task demand information to obtain a plurality of keywords and semantic association relations, and retrieves from a local database, a professional database and a network database respectively to obtain related candidate data information based on the keywords and the semantic association relations, wherein the local database comprises data information related to the inside of a user organization, the professional database stores professional data information of each industry, and the network database comprises internet platform related data information and is obtained through a webpage crawler;
further, the professional database stores financial field professional data, tax field professional data, legal field professional data and the like;
(3) The server performs data acquisition and splicing, respectively extracts M1 data files from a local database, M2 data files from a professional database and M3 data files from a network database through a web crawler on the basis of the retrieved candidate data information, wherein M1, M2 and M3 are user-defined parameters; splicing the M1+ M2+ M3 files into a merged file;
(4) The server executes data preprocessing operation on the merged file data information, wherein the preprocessing operation specifically comprises cleaning, duplicate removal and Word2vec feature extraction and conversion into a Word vector v;
(5) The server respectively inputs the sequences of the word vectors v generated through data preprocessing into a first document intelligent generation module and a second document intelligent generation module, wherein the first document intelligent generation module is a sequence-to-sequence model based on a Recurrent Neural Network (RNN), and the second document intelligent generation module is a sequence-to-sequence model based on a long-short term memory artificial neural network (LSTM);
the first document intelligent generation module is realized as follows: at time t, RNN inputs X at the current time t And the hidden state S at the previous moment t-1 As input, the two are simultaneously hidden state S for the current time t Updating, wherein the calculation formula is as follows:
wherein f represents a nonlinear activation function Sigmoid function, U and W are parameters in the model and respectively correspond to X t And hidden state S t-1 Corresponding weight matrix, p represents offset, and the output at time t is O t The formula is as follows:
v is a parameter needing to be learned, each unit of the RNN shares one group of parameters, the training parameters can be greatly reduced, and a first content document is output through the steps;
further, f can also be a Relu function, a tanh function, and the like;
the second document intelligent generation module is realized as follows: firstly, inputting the preprocessed data, namely word vector v into the encoding end of the LSTM network model Encoder,
r i-1 indicating a state of concealment from the previous time, r i For a hidden state output at each time i, the function g represents the transition of the hidden layer:
setting n words in the input model sequence, and converting the hidden state of each moment into an intermediate content vector c by the Encoder through a user-defined function q:
an attention mechanism is introduced into LSTM, the attention mechanism makes the c vector coded by an Encoder encoding end and the output in the decoding process of a Decoder decoding end carry out weighting operation, the weight is adjusted in each decoding process to obtain an independent c vector, and r in the following calculation process j For the hidden state of the encoder at each time, the total sequence length is T, and the vector at the ith time c is calculated as follows:
given the time of day i,
is a probability distribution at time j =1, 2., T,
calculating by using a softmax function; wherein c output from the Encoder end at the previous moment and r hidden state of output decoded at the Decode end at the previous moment i-1 Inputting a word vector for the next moment of a Decoder decoding end and predicting the word vector at the previous moment;
splicing the intermediate content vector and the hidden state of the decoding end together at the current decoding moment by the Decoder decoding end, inputting a full connection layer, simultaneously obtaining the probability distribution of words in a word list by utilizing a softmax function, and outputting the content to form a second content document;
(6) The server adopts a preset automatic evaluation algorithm to determine the quality of the first content document and the second content document, sequences according to the quality, outputs the final document according to a mode selected by a user, selects the document with high automatic evaluation quality to store and output and display in a local database when the user selects the mode I, and selects the final document to store and output and display in the local database according to the manual evaluation of the user when the user selects the mode II.
The preset automatic evaluation algorithm is an evaluation algorithm which comprehensively considers the information coverage rate, the information fluency and the information redundancy of the document content.
And when the user selects the second mode, an editing function is provided on the reading interface aiming at the documents manually evaluated and selected, an editing button is clicked, the editing interface is skipped, the documents are saved in a local database after the manual editing is finished, and the quality of the generated documents is further improved.
Further, the user submits the task requirement information of the retrieval content document in the local database to the server through the client, if the server hits the document containing all the keywords of the task requirement information of the retrieval content document in the local database through precise matching, the precisely matched document is output for the user to preview, edit or download, otherwise, the server automatically converts the task requirement information of the retrieval content document into the task requirement information of the generated document, and the server generates the document according to the steps (2) to (6) and outputs the document to the user.
It is another object of the present invention to provide a document intelligent generation and management apparatus, comprising a processor, a memory, and one or more programs stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps in the method.
It is another object of the present invention to provide a computer-readable storage medium storing a computer program for intelligent generation and management of documents, wherein the computer program causes a computer to execute the steps in the method.
By combining all the technical schemes, compared with the prior art, the invention has the following advantages:
(1) The source database generated by the document is rich, has both local files and professional databases, and can acquire real-time latest data information through a web crawler;
(2) The document generation is respectively based on the RNN model and the LSTM model, the document generation step is more intelligent, the generated document can be selected from the RNN model and the LSTM model according to quality evaluation, the finally formed document is high in quality, manual evaluation and editing are supported, and the quality of the generated document is further improved;
(3) The man-machine interaction is intelligent and convenient, and the document retrieval and management are convenient.
(4) When the local database can not retrieve accurate document information, the operation is automatically converted into a document generation operation, and documents required by the user are intelligently generated for browsing and searching of the user.
Drawings
FIG. 1 is a flowchart illustrating the implementation of the document intelligent generation and management method of the present invention.
FIG. 2 is a schematic structural diagram of the document intelligent generation and management apparatus of the present invention.
Detailed Description
This is further described below by way of example in conjunction with figures 1 and 2. It is to be understood that the embodiments described are only some of the embodiments of the invention, and not all of them. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without inventive step, are within the scope of the present invention.
The execution flow chart of the intelligent document generation and management method is shown in figure 1. As shown in FIG. 1, the present invention provides a method for intelligent generation and management of documents, comprising the following steps:
(1) A user inputs and generates document task demand information through a client and submits the task demand information to a server;
further, the document task generation requirement information is input text information or voice information, wherein the voice information needs to be further converted into text information so as to facilitate subsequent semantic analysis;
(2) The server executes semantic analysis on the task demand information to obtain a plurality of keywords and semantic association relations, and retrieves from a local database, a professional database and a network database respectively to obtain related candidate data information based on the keywords and the semantic association relations, wherein the local database comprises data information related to the inside of a user organization, the professional database stores professional data information of each industry, and the network database comprises internet platform related data information and is obtained through a webpage crawler;
further, the professional database stores financial field professional data, tax field professional data, legal field professional data and the like;
(3) The server performs data acquisition and splicing, respectively extracts M1 data files from a local database, M2 data files from a professional database and M3 data files from a network database through a web crawler on the basis of the retrieved candidate data information, wherein M1, M2 and M3 are user-defined parameters; splicing the M1+ M2+ M3 files into a merged file;
(4) The server executes data preprocessing operation on the merged file data information, wherein the preprocessing operation specifically comprises cleaning, duplicate removal, word2vec feature extraction and conversion into a Word vector v;
(5) The server respectively inputs the sequences of the word vectors v generated through data preprocessing into a first document intelligent generation module and a second document intelligent generation module, wherein the first document intelligent generation module is a sequence-to-sequence model based on a Recurrent Neural Network (RNN), and the second document intelligent generation module is a sequence-to-sequence model based on a long-short term memory artificial neural network (LSTM);
as a preferred embodiment, the first document intelligent generation module is implemented as follows: at time t, RNN inputs X at the current time t And the hidden state S at the previous moment t-1 As input, the two are simultaneously hidden state S for the current time t Updating, wherein the calculation formula is as follows:
wherein f represents a nonlinear activation function Sigmoid function, U and W are parameters in the model and respectively correspond to X t And a hidden state S t-1 Corresponding weight matrix, p represents offset, and output at time t is O t The formula is as follows:
v is a parameter needing to be learned, each unit of the RNN shares a group of parameters, the training parameters can be greatly reduced, and a first content document is output through the steps;
further, f can also be a Relu function, a tanh function, and the like;
as a preferred embodiment, the second document intelligent generation module is implemented as follows: firstly, inputting the preprocessed data, namely the word vector v into the encoding end of the LSTM network model Encoder,
r i-1 indicating a state of concealment from the previous time, r i For a hidden state output at each time i, the function g represents the transition of the hidden layer:
setting n words in the input model sequence, and converting the hidden state of each moment into an intermediate content vector c by the Encoder through a user-defined function q:
an attention mechanism is introduced into LSTM, the attention mechanism makes the c vector coded by an Encoder encoding end and the output in the decoding process of a Decoder decoding end carry out weighting operation, the weight is adjusted in each decoding process to obtain an independent c vector, and r in the following calculation process j For the hidden state of the encoder at each time, the total length of the sequence is T, and the vector at the ith time c is calculated as follows:
given the time of day i,
is a probability distribution at time j =1,2,. Times T,
calculating by using a softmax function; wherein c output from the Encoder end at the previous moment and r hidden state of output decoded at the Decode end at the previous moment i-1 The word vectors are input at the next moment of a Decoder end and predicted at the previous moment;
the Decoder side splices the intermediate content vector and the hidden state of the Decoder side together at the current decoding time, obtains probability distribution of words in a word list by using a softmax function while inputting a full connection layer, and outputs the content to form a second content document;
(6) The server adopts a preset automatic evaluation algorithm to determine the quality of the first content document and the second content document, sequences according to the quality, outputs the final document according to a mode selected by a user, selects the document with high automatic evaluation quality to store and output and display in a local database when the user selects the mode I, and selects the final document to store and output and display in the local database according to the manual evaluation of the user when the user selects the mode II.
The second mode is convenient for a user to manually evaluate the document quality, obtains and stores manually considered high-quality documents as source data documents when other content documents are subsequently generated, and improves the quality of subsequent document generation.
The preset automatic evaluation algorithm is an evaluation algorithm which comprehensively considers the information coverage rate, the information fluency and the information redundancy of the document content.
And when the user selects the second mode, an editing function is provided on the reading interface aiming at the document manually evaluated and selected, the editing button is clicked, the editing interface is skipped to, the document is stored in a local database after the manual editing is finished, and the quality of the generated document is further improved.
As a preferred embodiment, a user submits the task requirement information of the retrieval content document in a local database to a server through a client, if the server hits a document containing all keywords of the task requirement information of the retrieval content document in the local database through precise matching, the precisely matched document is output for the user to preview, edit or download, otherwise, the server automatically converts the task requirement information of the retrieval content document into the task requirement information of the generation document, and the server generates the document according to the steps (2) to (6) and outputs the document to the user.
In addition, the document intelligent generation and management method of the present invention is shown in fig. 2.
According to fig. 2, the present invention provides a document intelligent generation and management apparatus comprising a processor, a memory, and one or more programs stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps in the method. It should be noted that, as will be understood by those skilled in the art, the implementation functions of the modules shown in the embodiment of the intelligent document generation and management apparatus can be understood by referring to the relevant description of the intelligent document generation and management method. The functions of the respective modules shown in the embodiment of the document intelligent generation and management apparatus may be realized by a program (executable instructions) running on a processor, or may be realized by a specific logic circuit.
Accordingly, the present invention also provides a computer-readable storage medium having stored therein computer-executable instructions that, when executed by a processor, implement the method embodiments of the present application. Computer-readable storage media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device.
It should be understood that the above description is only for the preferred embodiment of the present disclosure, and is not intended to limit the scope of the present disclosure. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of the present disclosure should be included in the scope of protection of one or more embodiments of the present disclosure.
Claims (10)
1. A method for intelligently generating and managing documents is characterized by comprising the following steps:
(1) A user inputs and generates document task demand information through a client and submits the task demand information to a server;
(2) The server executes semantic analysis on the task demand information to obtain a plurality of keywords and semantic association relations, and retrieves from a local database, a professional database and a network database respectively to obtain related candidate data information based on the keywords and the semantic association relations, wherein the local database comprises data information related to the inside of a user organization, the professional database stores professional data information of each industry, and the network database comprises internet platform related data information and is obtained through a webpage crawler;
(3) The server performs data acquisition and splicing, respectively extracts M1 data files from a local database, M2 data files from a professional database and M3 data files from a network database through a web crawler on the basis of the retrieved candidate data information, wherein M1, M2 and M3 are user-defined parameters; splicing the M1+ M2+ M3 files into a merged file;
(4) The server executes data preprocessing operation on the merged file data information, wherein the preprocessing operation comprises data cleaning, duplicate removal and Word2vec feature extraction and conversion into a Word vector v;
(5) The server respectively inputs the sequences of the word vectors v generated through data preprocessing into a first document intelligent generation module and a second document intelligent generation module, wherein the first document intelligent generation module is a sequence-to-sequence model based on a Recurrent Neural Network (RNN), and the second document intelligent generation module is a sequence-to-sequence model based on a long-short term memory artificial neural network (LSTM);
the first document intelligent generation module is realized as follows: at time t, RNN inputs X at the current time t And hidden state S at the previous moment t-1 As input, the two are simultaneously hidden state S for the current time t Update and calculate the publicThe formula is as follows:
wherein f is Sigmoid function, U and W are parameters in the model and respectively correspond to X t And hidden state S t-1 Corresponding weight matrix, p represents offset, and output at time t is O t The formula is as follows:
wherein V is a parameter needing to be learned, and outputting a first content document through the steps;
the second document intelligent generation module is realized as follows: firstly, inputting the preprocessed data, namely word vector v into the encoding end of the LSTM network model Encoder,
r i-1 indicating a hidden state from the previous time, r i For each hidden state output at time i, the function g represents the hidden layer transition:
setting n words in the input model sequence, and converting the hidden state of each moment into an intermediate content vector c by the Encoder through a user-defined function q:
an attention mechanism is introduced in LSTM, and the attention mechanism enables the c vector coded by an Encoder encoding end and the output in the decoding process of a Decoder decoding endPerforming weighting operation, adjusting weight in each decoding process to obtain independent c vector, and in the following calculation process, r j For the hidden state of the encoder at each time, the total length of the sequence is T, and the vector at the ith time c is calculated as follows:
given the time of day i,
is a probability distribution at time j =1, 2., T,
calculating by using a softmax function; wherein c output from the Encoder end at the previous moment and r hidden state of output decoded at the Decode end at the previous moment i-1 The word vectors are input at the next moment of a Decoder end and predicted at the previous moment;
splicing the intermediate content vector and the hidden state of the decoding end together at the current decoding moment by the Decoder decoding end, inputting a full connection layer, simultaneously obtaining the probability distribution of words in a word list by utilizing a softmax function, and outputting the content to form a second content document;
(6) The server adopts a preset automatic evaluation algorithm to determine the quality of the first content document and the second content document, orders the first content document and the second content document according to the quality, outputs the final documents according to a mode selected by a user, selects the documents with high automatic evaluation quality to be stored in a local database and output and displayed when the user selects a mode I, and selects the final documents to be stored in the local database and output and displayed according to the manual evaluation of the user when the user selects a mode II.
2. The method for intelligent generation and management of documents as claimed in claim 1, wherein said generating document task requirement information in step (1) is text information or voice information, wherein the voice information is further converted into text information by the client for subsequent semantic analysis.
3. The method for intelligent generation and management of documents according to claim 1, wherein said professional database in step (2) stores at least one of the following professional data: financial field professional data, tax field professional data, legal field professional data.
4. The method for intelligent generation and management of documents according to claim 1, wherein f in step (5) is a Relu function.
5. The method for intelligent generation and management of documents as claimed in claim 1, wherein f in step (5) is a tanh function.
6. The method for intelligent generation and management of documents as claimed in claim 1, wherein said preset automatic evaluation algorithm in step (6) is an evaluation algorithm comprehensively considering information coverage, information fluency and information redundancy of document contents.
7. The method for intelligent generation and management of documents as claimed in claim 1, wherein in step (6), when the user selects mode two, the editing function is provided on the viewing interface for the documents selected by manual evaluation, the editing button is clicked, the editing interface is skipped, and the documents are saved in the local database after the manual editing is completed.
8. The method for intelligent generation and management of documents as claimed in claim 1, further comprising the steps of: and (3) submitting the task requirement information of the retrieval content document in the local database to the server by the user through the client, outputting the precise matching document for the user to preview, edit or download if the server hits the document containing all the keywords of the task requirement information of the retrieval content document in the local database through precise matching, or automatically converting the task requirement information of the retrieval content document into the task requirement information of the generated document by the server, and generating the document by the server according to the steps (2) to (6) and outputting the document to the user.
9. An apparatus for intelligent generation and management of documents, comprising a processor, a memory, and one or more programs stored in the memory and configured to be executed by the processor, the programs comprising instructions for performing the steps of the method of any of claims 1-8.
10. A computer-readable storage medium, in which a computer program for intelligent generation and management of documents is stored, wherein the computer program causes a computer to perform the method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211154504.XA CN115238710B (en) | 2022-09-22 | 2022-09-22 | Intelligent document generation and management method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211154504.XA CN115238710B (en) | 2022-09-22 | 2022-09-22 | Intelligent document generation and management method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115238710A true CN115238710A (en) | 2022-10-25 |
| CN115238710B CN115238710B (en) | 2022-11-22 |
Family
ID=83667012
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211154504.XA Active CN115238710B (en) | 2022-09-22 | 2022-09-22 | Intelligent document generation and management method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115238710B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117350266A (en) * | 2023-12-06 | 2024-01-05 | 本溪钢铁（集团）信息自动化有限责任公司 | Method and system for automatically generating document |
| CN117494672A (en) * | 2023-11-13 | 2024-02-02 | 北京大学长沙计算与数字经济研究院 | Method and device for generating industry document and computer readable storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436964A (en) * | 2008-12-17 | 2009-05-20 | 北京航空航天大学 | Top layer design platform for electronic system |
| US20120233205A1 (en) * | 2008-03-07 | 2012-09-13 | Inware, Llc | System and method for document management |
| CN106886408A (en) * | 2017-01-22 | 2017-06-23 | 广州真诺电子科技有限公司 | It is applied to the intelligent decision system of robot software engineer |
| CN110222160A (en) * | 2019-05-06 | 2019-09-10 | 平安科技（深圳）有限公司 | Intelligent semantic document recommendation method, device and computer readable storage medium |
| CN114020862A (en) * | 2021-11-04 | 2022-02-08 | 中国矿业大学 | Retrieval type intelligent question-answering system and method for coal mine safety regulations |
-
2022
- 2022-09-22 CN CN202211154504.XA patent/CN115238710B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120233205A1 (en) * | 2008-03-07 | 2012-09-13 | Inware, Llc | System and method for document management |
| CN101436964A (en) * | 2008-12-17 | 2009-05-20 | 北京航空航天大学 | Top layer design platform for electronic system |
| CN106886408A (en) * | 2017-01-22 | 2017-06-23 | 广州真诺电子科技有限公司 | It is applied to the intelligent decision system of robot software engineer |
| CN110222160A (en) * | 2019-05-06 | 2019-09-10 | 平安科技（深圳）有限公司 | Intelligent semantic document recommendation method, device and computer readable storage medium |
| CN114020862A (en) * | 2021-11-04 | 2022-02-08 | 中国矿业大学 | Retrieval type intelligent question-answering system and method for coal mine safety regulations |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117494672A (en) * | 2023-11-13 | 2024-02-02 | 北京大学长沙计算与数字经济研究院 | Method and device for generating industry document and computer readable storage medium |
| CN117350266A (en) * | 2023-12-06 | 2024-01-05 | 本溪钢铁（集团）信息自动化有限责任公司 | Method and system for automatically generating document |
| CN117350266B (en) * | 2023-12-06 | 2024-03-08 | 本溪钢铁（集团）信息自动化有限责任公司 | Method and system for automatically generating document |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115238710B (en) | 2022-11-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11599714B2 (en) | Methods and systems for modeling complex taxonomies with natural language understanding | |
| CN115238710B (en) | Intelligent document generation and management method and device | |
| CN111914085B (en) | Text fine granularity emotion classification method, system, device and storage medium | |
| CN110968695A (en) | Intelligent labeling method, device and platform based on active learning of weak supervision technology | |
| US11481387B2 (en) | Facet-based conversational search | |
| JP7208872B2 (en) | Systems and methods for generating proposals based on request for proposals (RFPs) | |
| CN111723295B (en) | Content distribution method, device and storage medium | |
| CN112528010B (en) | Knowledge recommendation method and device, computer equipment and readable storage medium | |
| CN111723256A (en) | Government affair user portrait construction method and system based on information resource library | |
| US20240020538A1 (en) | Systems and methods for real-time search based generative artificial intelligence | |
| CN116522912A (en) | Training method, device, medium and equipment for package design language model | |
| CN116756281A (en) | Knowledge question-answering method, device, equipment and medium | |
| CN116151235A (en) | Article generating method, article generating model training method and related equipment | |
| CN113641817A (en) | Report generation method and device based on subject and timeline abstract | |
| KR20220079029A (en) | Method for providing automatic document-based multimedia content creation service | |
| CN117336539B (en) | Video script production method and system for short video IP (Internet protocol) construction | |
| CN117609444B (en) | Searching question-answering method based on large model | |
| CN111539208B (en) | Sentence processing method and device, electronic device and readable storage medium | |
| CN118133793A (en) | Form generation method, form generation device, computer equipment and storage medium | |
| CN115481616A (en) | Target text acquisition method and device, computer equipment and storage medium | |
| Kurniasari et al. | Integrated Web-Based E-Fulfillment Services Information Systems | |
| CN115617949A (en) | Target object matching method and device and computer equipment | |
| Kopp et al. | Towards the Enterprise Architecture Web Mining Approach and Software Tool. | |
| CN115759041A (en) | Model training and abstract generating method, device and storage medium | |
| CN117787491A (en) | Stock section profit prediction method and equipment based on Transformer model |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |