CN117112776A - Enterprise knowledge base management and retrieval platform and method based on large language model - Google Patents
Enterprise knowledge base management and retrieval platform and method based on large language model Download PDFInfo
- Publication number
- CN117112776A CN117112776A CN202311233830.4A CN202311233830A CN117112776A CN 117112776 A CN117112776 A CN 117112776A CN 202311233830 A CN202311233830 A CN 202311233830A CN 117112776 A CN117112776 A CN 117112776A
- Authority
- CN
- China
- Prior art keywords
- data
- module
- knowledge
- neural network
- convolutional neural
- 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 26
- 238000007726 management method Methods 0.000 claims abstract description 68
- 238000013523 data management Methods 0.000 claims abstract description 26
- 230000003993 interaction Effects 0.000 claims abstract description 25
- 238000004364 calculation method Methods 0.000 claims abstract description 16
- 238000013500 data storage Methods 0.000 claims abstract description 14
- 238000011156 evaluation Methods 0.000 claims abstract description 9
- 238000012544 monitoring process Methods 0.000 claims abstract description 9
- 238000007405 data analysis Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims abstract description 8
- 238000013527 convolutional neural network Methods 0.000 claims description 99
- 238000010801 machine learning Methods 0.000 claims description 28
- 238000012545 processing Methods 0.000 claims description 27
- 238000012549 training Methods 0.000 claims description 15
- 238000005516 engineering process Methods 0.000 claims description 12
- 238000003058 natural language processing Methods 0.000 claims description 9
- 230000014759 maintenance of location Effects 0.000 claims description 6
- 238000013138 pruning Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 claims description 5
- 238000011161 development Methods 0.000 claims description 5
- 238000013135 deep learning Methods 0.000 claims description 4
- 238000005065 mining Methods 0.000 claims description 3
- 238000001556 precipitation Methods 0.000 claims description 3
- 238000012795 verification Methods 0.000 claims description 3
- 238000003860 storage Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/338—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Computational Linguistics (AREA)
- Databases & Information Systems (AREA)
- Entrepreneurship & Innovation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Animal Behavior & Ethology (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Economics (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses an enterprise knowledge base management and retrieval platform and method based on a large language model, and relates to the technical field of data management. The system comprises a data management center and a knowledge management center, wherein the data management center comprises a data architecture management module, a data quality management module, a data asset management module, a metadata management module, a data life cycle management module, a data storage module, an AI calculation module, a data analysis module, a data monitoring module and a data management module, and the knowledge management center comprises an enterprise knowledge base, a knowledge interaction module, a data classification management module, an asset evaluation module, a knowledge map module, a permission management module and a knowledge query module. The method is applicable to the system described above. The application systematically integrates the data management and knowledge management functions, realizes the clear and reasonable data management of enterprise knowledge, and the convenient and accurate retrieval of the enterprise knowledge, and improves the enterprise data management efficiency.
Description
Technical Field
The application relates to the technical field of data management, in particular to an enterprise knowledge base management and retrieval platform and method based on a large language model.
Background
Reasonable management of enterprise data is a precondition for benign development of enterprises. In the environment of rapid development of computer and network technology, the management problem of various information of enterprises is more prominent. On the one hand, the storage and management of various complicated information of enterprises and on the other hand, the accurate consulting and calling of various required information are realized, so that systematic management of enterprise information is the basis for realizing reasonable information management and accurate information retrieval, and only then, the enterprise benefit and efficiency can be improved based on the utilization of the enterprise information.
Traditional enterprise information management realizes the storage and matching of complicated data by using a large data or cloud management mode, but the data management mode has slow speed and low retrieval efficiency for data calling and accurate matching, and is not beneficial to quickly and accurately acquiring the required business demand data on the premise of a large number of data bases.
Disclosure of Invention
The application aims to provide an enterprise knowledge base management and retrieval platform and method based on a large language model, so as to solve the technical problems in the background technology.
In order to achieve the above purpose, the present application discloses the following technical solutions:
in a first aspect, the application discloses an enterprise knowledge base management and retrieval platform based on a large language model, which comprises a data management center and a knowledge management center;
the data governance center includes: a data architecture management module configured to manage data distribution, a data quality management module configured to intelligently rectify data and track execution, a data asset management module configured to identify and classify database assets, a metadata management module configured to identify, classify and archive metadata, a data lifecycle management module configured to lifecycle manage and optimize data, the system comprises a data storage module, an AI calculation module, a data analysis module, a data monitoring module, a data management module and a data management module, wherein the data storage module is configured to reasonably store data resources based on data characteristics, the AI calculation module is configured to generate a knowledge graph and a machine learning model according to service requirements and the data characteristics, the data analysis module is configured to intelligently analyze the data and generate a statistical report and an image-text report, the data monitoring module is configured to monitor the data and generate safety precautions, and the data management module is configured to manage the data according to the data management requirements and strategies;
the knowledge management center includes: the system comprises an enterprise knowledge base, a knowledge interaction module, a data classification management module, an asset assessment module, a knowledge map module, a right management module and a knowledge query module, wherein the enterprise knowledge base is configured to provide knowledge retrieval and knowledge recommendation for users based on natural language processing technology, the knowledge interaction module is configured to assist staff in solving problems based on data in the enterprise knowledge base, the data classification management module is configured to generate classification labels for knowledge documents, the asset assessment module is configured to analyze contribution and influence of knowledge assets and evaluate the value of the assets, the knowledge map module is configured to generate and update a knowledge map based on data update in the enterprise knowledge base, the right management module is configured to allocate access rights for users, and the knowledge query module is configured to acquire query results based on user feedback and query history; and the enterprise knowledge base performs data interaction with the data storage module.
In one embodiment, the AI computation module is specifically configured to:
generating a knowledge graph based on natural language understanding according to service requirements and data characteristics;
and generating a machine learning model based on the convolutional neural network according to the service requirements and the data characteristics, wherein the machine learning model is used for deep learning of data assets and metadata and providing a data matching function for data retrieval and calling of the knowledge interaction module and the enterprise knowledge base.
In one embodiment, the generating a machine learning model based on the convolutional neural network specifically includes:
constructing an original convolutional neural network model, and training the original convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a front convolutional neural network model;
performing light weight processing on the characteristic channels and the data characteristics of the front-stage convolutional neural network model to obtain a front-stage light weight convolutional neural network model;
training the lightweight convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a post convolutional neural network model;
performing light weight processing on the characteristic channels of the back-stage convolutional neural network model to obtain a back-stage light weight convolutional neural network;
and adopting model parameters of the front-stage convolutional neural network model to carry out parameter adjustment on the rear-stage lightweight convolutional neural network so as to obtain a required machine learning model.
In one embodiment, the light-weight processing of the characteristic channel of the preceding convolutional neural network specifically includes:
acquiring an optimal retention state of a characteristic channel of the front-stage convolutional neural network model, and deleting a redundant characteristic channel in the front-stage convolutional neural network model;
and deleting redundant features in the business requirements and the data features to perform network pruning processing based on the corresponding data asset value in the asset evaluation module when the original convolutional neural network model is trained.
In one embodiment, the knowledge management center further comprises: the system comprises one or more of a driven growth module configured to analyze employee needs and context and generate personalized training and development plans, a knowledge mining module configured to generate knowledge reports based on analysis of data within the enterprise knowledge base, a sharing collaboration module configured to co-contribute with needs and formulate knowledge sharing incentives and rewards schemes by analysts, and a knowledge precipitation module configured to analyze results based on data in the enterprise knowledge base and identify duplicate and redundant information.
In a second aspect, the present application provides a method for managing and retrieving a platform of an enterprise knowledge base based on a large language model, the method comprising the following steps:
creating a user identity and defining the query authority of the user;
verifying the logged-in user identity and identifying the inquiry authority of the user;
after the user identity verification is passed, matching corresponding data in an enterprise knowledge base based on user feedback;
based on the query rights of the user, data within the range of the query rights among the data matched in the enterprise knowledge base is provided to the user.
In one embodiment, the user feedback includes business requirements and data characteristics, the enterprise knowledge base performs data interaction with a data storage module configured to rationally store data resources based on the data characteristics; the matching of corresponding data in the enterprise knowledge base is realized through an AI calculation module, and the AI calculation module is configured to generate a knowledge graph and a machine learning model according to service requirements and data characteristics;
the machine learning model is generated based on a convolutional neural network, and the specific steps of generating the machine learning model comprise:
constructing an original convolutional neural network model, and training the original convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a front convolutional neural network model;
performing light weight processing on the characteristic channels and the data characteristics of the front-stage convolutional neural network model to obtain a front-stage light weight convolutional neural network model;
training the lightweight convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a post convolutional neural network model;
performing light weight processing on the characteristic channels of the back-stage convolutional neural network model to obtain a back-stage light weight convolutional neural network;
and adopting model parameters of the front-stage convolutional neural network model to carry out parameter adjustment on the rear-stage lightweight convolutional neural network so as to obtain a required machine learning model.
In one embodiment, the light-weight processing of the characteristic channel of the preceding convolutional neural network specifically includes:
acquiring an optimal retention state of a characteristic channel of the front-stage convolutional neural network model, and deleting a redundant characteristic channel in the front-stage convolutional neural network model;
and deleting redundant features in the business requirements and the data features to perform network pruning processing based on the corresponding data asset value in the asset evaluation module when the original convolutional neural network model is trained.
In one embodiment, the user feedback includes questions posed by the user through the knowledge interaction module and query keywords input by the user through the knowledge query module;
when a user puts forward a problem through a knowledge interaction module, extracting service requirements and data characteristics corresponding to the problem, matching corresponding data in an enterprise knowledge base based on the service requirements and the data characteristics, and displaying the matched data in the authority range to the user through the knowledge interaction module in a question-and-answer mode based on the inquiry authority of the user;
when the user is the query keyword input through the knowledge query module, matching corresponding data in the enterprise knowledge base based on the input query keyword, and displaying the matched data in the authority range to the user through the knowledge query module in a result display mode based on the query authority of the user.
In one embodiment, the matching the corresponding data in the enterprise knowledge base based on the input query keyword specifically includes:
and identifying and analyzing the short meaning words and the synonyms of the input query keywords through a natural language processing technology, acquiring corresponding business requirements and data characteristics, and matching corresponding data in an enterprise knowledge base based on the business requirements and the data characteristics.
The beneficial effects are that: the enterprise knowledge base management and retrieval platform and method based on the large language model integrate data management and knowledge management functions systematically, comprehensively process data identification, classification, storage, monitoring, management and the like in data management, reasonably arrange data calling, matching, evaluation, retrieval and the like in knowledge management, realize clear and reasonable data management of enterprise knowledge, and convenient and accurate retrieval of enterprise knowledge, and improve enterprise data management efficiency.
Drawings
In order to more clearly illustrate the embodiments of the application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic diagram illustrating a functional profile of a data management center in accordance with an embodiment of the present application;
FIG. 2 is a schematic diagram of a functional profile of a knowledge management center in accordance with an embodiment of the application;
FIG. 3 is a block flow diagram of enterprise knowledge base management and retrieval based on a large language model in an embodiment of the application.
Detailed Description
The following description of the technical solutions in the embodiments of the present application will be clear and complete, and it is obvious that the described embodiments are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
In this document, 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 … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.
The embodiment discloses an enterprise knowledge base management and retrieval platform based on a large language model in a first aspect, which comprises a data management center and a knowledge management center.
Specifically, as shown in the schematic functional profile diagram of fig. 1, the data management center includes: the system comprises a data architecture management module configured for managing data distribution, a data quality management module configured for intelligently rectifying data and tracking execution conditions, a data asset management module configured for identifying and classifying database assets, a metadata management module configured for identifying, classifying and archiving metadata, a data life cycle management module configured for carrying out life cycle management and optimization on the data, a data storage module configured for reasonably storing data resources based on data characteristics, an AI calculation module configured for generating a knowledge graph and a machine learning model according to service requirements and data characteristics, a data analysis module configured for intelligently analyzing the data and generating a statistical report and an image-text report, a data monitoring module configured for monitoring the data and generating a safety warning, and a data management module configured for carrying out data management on the data according to data management requirements and strategies.
Specifically, as shown in the schematic diagram of the functional profile in fig. 2, the knowledge management center includes: the system comprises an enterprise knowledge base, a knowledge interaction module, a data classification management module, an asset assessment module, a knowledge map module, a right management module and a knowledge query module, wherein the enterprise knowledge base is configured to provide knowledge retrieval and knowledge recommendation for users based on natural language processing technology, the knowledge interaction module is configured to assist staff in solving problems based on data in the enterprise knowledge base, the data classification management module is configured to generate classification labels for knowledge documents, the asset assessment module is configured to analyze contribution and influence of knowledge assets and evaluate the value of the assets, the knowledge map module is configured to generate and update a knowledge map based on data update in the enterprise knowledge base, the right management module is configured to allocate access rights for users, and the knowledge query module is configured to acquire query results based on user feedback and query history; and the enterprise knowledge base performs data interaction with the data storage module. And, as a possible implementation manner of this embodiment, the knowledge management center further includes: the system comprises a driven growth module configured to analyze employee requirements and backgrounds and generate personalized training and development plans, a knowledge mining module configured to generate knowledge reports based on analysis of data in an enterprise knowledge base, a sharing collaboration module configured to co-contribute with requirements and formulate knowledge sharing incentive and rewards schemes by analysts, and a knowledge precipitation module configured to analyze results and identify duplicate and redundant information based on the data in the enterprise knowledge base.
As a preferred implementation manner of this embodiment, in order to improve efficiency of data analysis, classification, and matching, and achieve effects of load reduction and energy reduction, in this embodiment, the AI computation module is specifically configured to:
generating a knowledge graph based on natural language understanding according to service requirements and data characteristics;
and generating a machine learning model based on the convolutional neural network according to the service requirements and the data characteristics, wherein the machine learning model is used for deep learning of data assets and metadata and providing a data matching function for data retrieval and calling of the knowledge interaction module and the enterprise knowledge base.
Further preferably, the generating a machine learning model based on the convolutional neural network specifically includes:
constructing an original convolutional neural network model, and training the original convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a front convolutional neural network model;
performing light weight processing on the characteristic channels and the data characteristics of the front-stage convolutional neural network model to obtain a front-stage light weight convolutional neural network model;
training the lightweight convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a post convolutional neural network model;
performing light weight processing on the characteristic channels of the back-stage convolutional neural network model to obtain a back-stage light weight convolutional neural network;
and adopting model parameters of the front-stage convolutional neural network model to carry out parameter adjustment on the rear-stage lightweight convolutional neural network so as to obtain a required machine learning model.
The light-weight processing of the characteristic channel of the front-stage convolutional neural network specifically comprises the following steps:
acquiring an optimal retention state of a characteristic channel of the front-stage convolutional neural network model, and deleting a redundant characteristic channel in the front-stage convolutional neural network model;
and deleting redundant features in the business requirements and the data features to perform network pruning processing based on the corresponding data asset value in the asset evaluation module when the original convolutional neural network model is trained.
Based on the above, the calculation pressure of the machine learning model is reduced by the model light-weight processing mode, the data matching efficiency is improved, and convenience and efficiency are improved for subsequent data retrieval and knowledge calling. Further, by means of secondary light weight processing, on the premise of improving the weight of high-value data in the model, a matching channel of low-value data in the model is reserved, neglect of the light weight processed model to data features corresponding to some low-value data is avoided, comprehensiveness of data matching is guaranteed, and a stable and reliable data matching basis is provided for enterprise knowledge management and retrieval.
In a second aspect, the present embodiment discloses a large language model-based enterprise knowledge base management and retrieval method as shown in fig. 3, which is used in the large language model-based enterprise knowledge base management and retrieval system. Specifically, the method comprises the following steps:
s101, creating a user identity and defining the query authority of a user;
s102-verifying the logged-in user identity and identifying the query authority of the user, wherein it can be understood that the data in the enterprise knowledge base sets a corresponding query authority range based on the user identity, and for people of different levels, only the data in the corresponding query authority range of the user identity can be queried;
s103, after the user identity verification is passed, matching corresponding data in an enterprise knowledge base based on user feedback;
s104, based on the query authority of the user, providing the data which is matched in the enterprise knowledge base and is within the range of the query authority to the user.
Correspondingly, the user feedback comprises service requirements and data characteristics, the enterprise knowledge base performs data interaction with a data storage module, and the data storage module is configured to reasonably store data resources based on the data characteristics; the matching of corresponding data in the enterprise knowledge base is realized through an AI calculation module, and similarly, in order to improve the efficiency of data analysis, classification and matching, the AI calculation module is configured to generate a knowledge graph and a machine learning model according to service requirements and data characteristics.
The machine learning model is generated based on a convolutional neural network, and the specific steps of generating the machine learning model comprise:
constructing an original convolutional neural network model, and training the original convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a front convolutional neural network model;
performing light weight processing on the characteristic channels and the data characteristics of the front-stage convolutional neural network model to obtain a front-stage light weight convolutional neural network model;
training the lightweight convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a post convolutional neural network model;
performing light weight processing on the characteristic channels of the back-stage convolutional neural network model to obtain a back-stage light weight convolutional neural network;
and adopting model parameters of the front-stage convolutional neural network model to carry out parameter adjustment on the rear-stage lightweight convolutional neural network so as to obtain a required machine learning model.
Further, the light-weight processing of the characteristic channel of the front-stage convolutional neural network specifically includes:
acquiring an optimal retention state of a characteristic channel of the front-stage convolutional neural network model, and deleting a redundant characteristic channel in the front-stage convolutional neural network model;
and deleting redundant features in the business requirements and the data features to perform network pruning processing based on the corresponding data asset value in the asset evaluation module when the original convolutional neural network model is trained.
As a preferred implementation manner of this embodiment, the user feedback includes questions posed by the user through the knowledge interaction module and query keywords input by the user through the knowledge query module, which correspond to the function profiles corresponding to the respective modules shown in fig. 1 and 2.
When a user puts forward a problem through a knowledge interaction module, extracting service requirements and data characteristics corresponding to the problem, matching corresponding data in an enterprise knowledge base based on the service requirements and the data characteristics, and displaying the matched data in the authority range to the user through the knowledge interaction module in a question-and-answer mode based on the inquiry authority of the user.
When the user is the query keyword input through the knowledge query module, matching corresponding data in the enterprise knowledge base based on the input query keyword, and displaying the matched data in the authority range to the user through the knowledge query module in a result display mode based on the query authority of the user. Further, the matching of the corresponding data in the enterprise knowledge base based on the input query keyword specifically includes:
and identifying and analyzing the short meaning words and the synonyms of the input query keywords through a natural language processing technology, acquiring corresponding business requirements and data characteristics, and matching corresponding data in an enterprise knowledge base based on the business requirements and the data characteristics.
Further, in the recognition and analysis of the keywords, the service requirements and the data characteristics are obtained specifically through the following steps:
step 1: analyzing the keywords, and acquiring first semantics and second semantics corresponding to the keywords based on a natural language processing technology;
step 2: matching a reference semantic similar to the first semantic and the second semantic in a semantic database established by adopting business common semantics in a natural language processing technology;
step 3: calculating the similarity rho between the first semantic meaning and the second semantic meaning respectively, wherein the calculation formula of the similarity is rho= Σk i *M i Wherein M is i K being similarity values corresponding to subject, predicate, object i For the weight coefficients of subjects, predicates and objects in keyword matching respectively, in the calculation process, ρ=k Main unit *M Main unit +K So-called "two-way" system *M So-called "two-way" system +K Bin (guest) *M Bin (guest) When the subject, predicate or object in the first or second semantics is the same as the word in the corresponding reference semantics, corresponding M i 1, otherwise, obtaining a corresponding similarity value through a word sense matching calculation model obtained based on a natural language processing technology and a deep learning technology, wherein the similarity value obtained through the model is smaller than 1;
step 4: comparing the acquired similarity, and selecting a reference semantic corresponding to the first semantic or the second semantic with large similarity as the semantic of the keyword, thereby completing the recognition and analysis of the keyword.
Based on the above, the large language model-based enterprise knowledge base management and retrieval platform and method disclosed in the embodiment systematically integrate data management and knowledge management functions, and in data management, the data is comprehensively processed in recognition, classification, storage, monitoring, management and the like, and in knowledge management, the data is reasonably arranged in data calling, matching, evaluation, retrieval and the like, so that clear and reasonable data management of enterprise knowledge and convenient and accurate retrieval of enterprise knowledge are realized, and the enterprise data management efficiency is improved.
In the embodiments provided by the present application, it is to be understood that the embodiments described herein may be implemented in hardware, software, firmware, middleware, code, or any suitable combination thereof. For a hardware implementation, the processor may be implemented in one or more of the following units: an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to perform the functions described herein, or a combination thereof. For a software implementation, some or all of the flow of an embodiment may be accomplished by a computer program to instruct the associated hardware. When implemented, the above-described programs may be stored in or transmitted as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a computer. The computer readable media can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.
Finally, it should be noted that: the foregoing description is only illustrative of the preferred embodiments of the present application, and although the present application has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that modifications may be made to the embodiments described, or equivalents may be substituted for elements thereof, and any modifications, equivalents, improvements or changes may be made without departing from the spirit and principles of the present application.
Claims (10)
1. An enterprise knowledge base management and retrieval platform based on a large language model is characterized by comprising a data management center and a knowledge management center;
the data governance center includes: a data architecture management module configured to manage data distribution, a data quality management module configured to intelligently rectify data and track execution, a data asset management module configured to identify and classify database assets, a metadata management module configured to identify, classify and archive metadata, a data lifecycle management module configured to lifecycle manage and optimize data, the system comprises a data storage module, an AI calculation module, a data analysis module, a data monitoring module, a data management module and a data management module, wherein the data storage module is configured to reasonably store data resources based on data characteristics, the AI calculation module is configured to generate a knowledge graph and a machine learning model according to service requirements and the data characteristics, the data analysis module is configured to intelligently analyze the data and generate a statistical report and an image-text report, the data monitoring module is configured to monitor the data and generate safety precautions, and the data management module is configured to manage the data according to the data management requirements and strategies;
the knowledge management center includes: the system comprises an enterprise knowledge base, a knowledge interaction module, a data classification management module, an asset assessment module, a knowledge map module, a right management module and a knowledge query module, wherein the enterprise knowledge base is configured to provide knowledge retrieval and knowledge recommendation for users based on natural language processing technology, the knowledge interaction module is configured to assist staff in solving problems based on data in the enterprise knowledge base, the data classification management module is configured to generate classification labels for knowledge documents, the asset assessment module is configured to analyze contribution and influence of knowledge assets and evaluate the value of the assets, the knowledge map module is configured to generate and update a knowledge map based on data update in the enterprise knowledge base, the right management module is configured to allocate access rights for users, and the knowledge query module is configured to acquire query results based on user feedback and query history; and the enterprise knowledge base performs data interaction with the data storage module.
2. The large language model based enterprise knowledge base management and retrieval platform of claim 1, wherein the AI computation module is specifically configured to:
generating a knowledge graph based on natural language understanding according to service requirements and data characteristics;
and generating a machine learning model based on the convolutional neural network according to the service requirements and the data characteristics, wherein the machine learning model is used for deep learning of data assets and metadata and providing a data matching function for data retrieval and calling of the knowledge interaction module and the enterprise knowledge base.
3. The large language model based enterprise knowledge base management and retrieval platform as claimed in claim 2, wherein the convolutional neural network based machine learning model generation specifically comprises:
constructing an original convolutional neural network model, and training the original convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a front convolutional neural network model;
performing light weight processing on the characteristic channels and the data characteristics of the front-stage convolutional neural network model to obtain a front-stage light weight convolutional neural network model;
training the lightweight convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a post convolutional neural network model;
performing light weight processing on the characteristic channels of the back-stage convolutional neural network model to obtain a back-stage light weight convolutional neural network;
and adopting model parameters of the front-stage convolutional neural network model to carry out parameter adjustment on the rear-stage lightweight convolutional neural network so as to obtain a required machine learning model.
4. The large language model based enterprise knowledge base management and retrieval platform as claimed in claim 3, wherein said lightweight processing of the characteristic channels of the front-level convolutional neural network specifically comprises:
acquiring an optimal retention state of a characteristic channel of the front-stage convolutional neural network model, and deleting a redundant characteristic channel in the front-stage convolutional neural network model;
and deleting redundant features in the business requirements and the data features to perform network pruning processing based on the corresponding data asset value in the asset evaluation module when the original convolutional neural network model is trained.
5. The large language model based enterprise knowledge base management and retrieval platform of claim 1, wherein the knowledge management center further comprises: the system comprises one or more of a driven growth module configured to analyze employee needs and context and generate personalized training and development plans, a knowledge mining module configured to generate knowledge reports based on analysis of data within the enterprise knowledge base, a sharing collaboration module configured to co-contribute with needs and formulate knowledge sharing incentives and rewards schemes by analysts, and a knowledge precipitation module configured to analyze results based on data in the enterprise knowledge base and identify duplicate and redundant information.
6. A method for managing and retrieving a large language model-based enterprise knowledge base, comprising the large language model-based enterprise knowledge base management and retrieval platform as claimed in any one of claims 1-5, the method comprising the steps of:
creating a user identity and defining the query authority of the user;
verifying the logged-in user identity and identifying the inquiry authority of the user;
after the user identity verification is passed, matching corresponding data in an enterprise knowledge base based on user feedback;
based on the query rights of the user, data within the range of the query rights among the data matched in the enterprise knowledge base is provided to the user.
7. The large language model based enterprise knowledge base management and retrieval method of claim 6, wherein the user feedback includes business requirements and data characteristics, the enterprise knowledge base is in data interaction with a data storage module, the data storage module is configured to reasonably store data resources based on the data characteristics; the matching of corresponding data in the enterprise knowledge base is realized through an AI calculation module, and the AI calculation module is configured to generate a knowledge graph and a machine learning model according to service requirements and data characteristics;
the machine learning model is generated based on a convolutional neural network, and the specific steps of generating the machine learning model comprise:
constructing an original convolutional neural network model, and training the original convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a front convolutional neural network model;
performing light weight processing on the characteristic channels and the data characteristics of the front-stage convolutional neural network model to obtain a front-stage light weight convolutional neural network model;
training the lightweight convolutional neural network model by adopting business requirements and data characteristics of an enterprise to obtain a post convolutional neural network model;
performing light weight processing on the characteristic channels of the back-stage convolutional neural network model to obtain a back-stage light weight convolutional neural network;
and adopting model parameters of the front-stage convolutional neural network model to carry out parameter adjustment on the rear-stage lightweight convolutional neural network so as to obtain a required machine learning model.
8. The method for managing and retrieving a knowledge base of an enterprise based on a large language model according to claim 7, wherein the light-weight processing is performed on the characteristic channels of the front-level convolutional neural network, specifically comprising:
acquiring an optimal retention state of a characteristic channel of the front-stage convolutional neural network model, and deleting a redundant characteristic channel in the front-stage convolutional neural network model;
and deleting redundant features in the business requirements and the data features to perform network pruning processing based on the corresponding data asset value in the asset evaluation module when the original convolutional neural network model is trained.
9. The method for managing and retrieving a knowledge base of an enterprise based on a large language model according to claim 6, wherein the user feedback includes questions posed by a user through a knowledge interaction module and query keywords input by the user through a knowledge query module;
when a user puts forward a problem through a knowledge interaction module, extracting service requirements and data characteristics corresponding to the problem, matching corresponding data in an enterprise knowledge base based on the service requirements and the data characteristics, and displaying the matched data in the authority range to the user through the knowledge interaction module in a question-and-answer mode based on the inquiry authority of the user;
when the user is the query keyword input through the knowledge query module, matching corresponding data in the enterprise knowledge base based on the input query keyword, and displaying the matched data in the authority range to the user through the knowledge query module in a result display mode based on the query authority of the user.
10. The method for managing and retrieving a large language model based enterprise knowledge base according to claim 9, wherein said matching corresponding data in the enterprise knowledge base based on the input query keyword comprises:
and identifying and analyzing the short meaning words and the synonyms of the input query keywords through a natural language processing technology, acquiring corresponding business requirements and data characteristics, and matching corresponding data in an enterprise knowledge base based on the business requirements and the data characteristics.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311233830.4A CN117112776A (en) | 2023-09-23 | 2023-09-23 | Enterprise knowledge base management and retrieval platform and method based on large language model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311233830.4A CN117112776A (en) | 2023-09-23 | 2023-09-23 | Enterprise knowledge base management and retrieval platform and method based on large language model |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117112776A true CN117112776A (en) | 2023-11-24 |
Family
ID=88794836
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311233830.4A Pending CN117112776A (en) | 2023-09-23 | 2023-09-23 | Enterprise knowledge base management and retrieval platform and method based on large language model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117112776A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117688135A (en) * | 2024-01-30 | 2024-03-12 | 中国标准化研究院 | Digital resource emergency management method and system |
CN117709909A (en) * | 2024-02-06 | 2024-03-15 | 绵阳师范学院 | Business data processing method based on large language model |
CN117725423A (en) * | 2024-02-18 | 2024-03-19 | 青岛海尔科技有限公司 | Method and device for generating feedback information based on large model |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150095331A1 (en) * | 2012-12-21 | 2015-04-02 | Cloud Computing Center Chinese Academy Of Sciences | Establishing and querying methods of knowledge library engine based on emergency management |
CN110674206A (en) * | 2019-09-30 | 2020-01-10 | 广东鼎义互联科技股份有限公司 | Knowledge sharing platform applied to enterprises |
CN113657576A (en) * | 2021-07-21 | 2021-11-16 | 浙江大华技术股份有限公司 | Convolutional neural network model lightweight method and device, and image identification method |
CN114722169A (en) * | 2022-03-18 | 2022-07-08 | 西安电子科技大学 | Enterprise intelligent question-answering system, control method, medium, equipment and terminal |
CN115062587A (en) * | 2022-06-02 | 2022-09-16 | 北京航空航天大学 | Knowledge graph embedding and reply generation method based on surrounding information |
CN116762080A (en) * | 2021-02-01 | 2023-09-15 | 利普麦德株式会社 | Neural network generation device, neural network operation device, edge device, neural network control method, and software generation program |
CN116756085A (en) * | 2023-05-25 | 2023-09-15 | 广州棉花糖科技有限公司 | Intelligent cooperative control management system for enterprise knowledge base |
-
2023
- 2023-09-23 CN CN202311233830.4A patent/CN117112776A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20150095331A1 (en) * | 2012-12-21 | 2015-04-02 | Cloud Computing Center Chinese Academy Of Sciences | Establishing and querying methods of knowledge library engine based on emergency management |
CN110674206A (en) * | 2019-09-30 | 2020-01-10 | 广东鼎义互联科技股份有限公司 | Knowledge sharing platform applied to enterprises |
CN116762080A (en) * | 2021-02-01 | 2023-09-15 | 利普麦德株式会社 | Neural network generation device, neural network operation device, edge device, neural network control method, and software generation program |
CN113657576A (en) * | 2021-07-21 | 2021-11-16 | 浙江大华技术股份有限公司 | Convolutional neural network model lightweight method and device, and image identification method |
CN114722169A (en) * | 2022-03-18 | 2022-07-08 | 西安电子科技大学 | Enterprise intelligent question-answering system, control method, medium, equipment and terminal |
CN115062587A (en) * | 2022-06-02 | 2022-09-16 | 北京航空航天大学 | Knowledge graph embedding and reply generation method based on surrounding information |
CN116756085A (en) * | 2023-05-25 | 2023-09-15 | 广州棉花糖科技有限公司 | Intelligent cooperative control management system for enterprise knowledge base |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117688135A (en) * | 2024-01-30 | 2024-03-12 | 中国标准化研究院 | Digital resource emergency management method and system |
CN117688135B (en) * | 2024-01-30 | 2024-04-30 | 中国标准化研究院 | Digital resource emergency management method and system |
CN117709909A (en) * | 2024-02-06 | 2024-03-15 | 绵阳师范学院 | Business data processing method based on large language model |
CN117709909B (en) * | 2024-02-06 | 2024-04-09 | 绵阳师范学院 | Business data processing method based on large language model |
CN117725423A (en) * | 2024-02-18 | 2024-03-19 | 青岛海尔科技有限公司 | Method and device for generating feedback information based on large model |
CN117725423B (en) * | 2024-02-18 | 2024-05-24 | 青岛海尔科技有限公司 | Method and device for generating feedback information based on large model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11500818B2 (en) | Method and system for large scale data curation | |
US11663254B2 (en) | System and engine for seeded clustering of news events | |
CN117112776A (en) | Enterprise knowledge base management and retrieval platform and method based on large language model | |
CN106407278B (en) | Architecture design system of big data platform | |
US20200242134A1 (en) | Method and apparatus for grouping data records | |
Weng et al. | Using text classification and multiple concepts to answer e-mails | |
US11250513B2 (en) | Computer implemented system for generating assurance related planning process and documents for an entity and method thereof | |
Liu | Using big data database to construct new GFuzzy text mining and decision algorithm for targeting and classifying customers | |
CN115547466B (en) | Medical institution registration and review system and method based on big data | |
US11620453B2 (en) | System and method for artificial intelligence driven document analysis, including searching, indexing, comparing or associating datasets based on learned representations | |
US11694020B2 (en) | Systems and methods for XBRL tag suggestion and validation | |
KR20200145299A (en) | Intelligent recruitment support platform based on online interview video analysis and social media information analysis | |
CN110310012B (en) | Data analysis method, device, equipment and computer readable storage medium | |
CN115358481A (en) | Early warning and identification method, system and device for enterprise ex-situ migration | |
CN115982429B (en) | Knowledge management method and system based on flow control | |
US20220374401A1 (en) | Determining domain and matching algorithms for data systems | |
CN114611841A (en) | Scenic spot tourist flow prediction method and device | |
CN113222471A (en) | Asset wind control method and device based on new media data | |
WO2019246252A1 (en) | Systems and methods for identifying and linking events in structured proceedings | |
Zirui | An evaluation approach of financial performance of university based on big data | |
CN117668205B (en) | Smart logistics customer service processing method, system, equipment and storage medium | |
Wenjun et al. | Research on brand crisis identify index model based on cluster analysis | |
Wang | Human resource network information recommendation method based on machine learning | |
Valavala et al. | Methodology: Automatic Database Index Tuning Using Machine Learning | |
CN118037250A (en) | Data mining method and system applying text informatization system |
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 |