WO2023004807A1 - 知识管理系统、方法、装置、电子设备和存储介质 - Google Patents

知识管理系统、方法、装置、电子设备和存储介质 Download PDF

Info

Publication number
WO2023004807A1
WO2023004807A1 PCT/CN2021/109880 CN2021109880W WO2023004807A1 WO 2023004807 A1 WO2023004807 A1 WO 2023004807A1 CN 2021109880 W CN2021109880 W CN 2021109880W WO 2023004807 A1 WO2023004807 A1 WO 2023004807A1
Authority
WO
WIPO (PCT)
Prior art keywords
data set
knowledge management
target semantic
node
index file
Prior art date
Application number
PCT/CN2021/109880
Other languages
English (en)
French (fr)
Inventor
高亮
王文科
李强
Original Assignee
西门子股份公司
西门子（中国）有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 西门子股份公司, 西门子（中国）有限公司 filed Critical 西门子股份公司
Priority to PCT/CN2021/109880 priority Critical patent/WO2023004807A1/zh
Priority to CN202180100338.2A priority patent/CN117716349A/zh
Publication of WO2023004807A1 publication Critical patent/WO2023004807A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/10—File systems; File servers
    • G06F16/13—File access structures, e.g. distributed indices

Definitions

  • the present application relates to the technical field of knowledge management, in particular to a knowledge management system, method, device, electronic equipment and storage medium.
  • Knowledge graph is a knowledge system formed by structuring knowledge, which contains basic facts, general rules and other relevant information, and is often used in tasks such as knowledge extraction, storage and reasoning in the field of artificial intelligence.
  • Semantics Datasets can be used to build knowledge graphs, and users can use the knowledge graphs built based on semantic datasets to develop applications, such as developing data retrieval applications.
  • a centralized management system is usually used to manage knowledge.
  • all parties can create new semantic datasets and update existing semantic datasets.
  • the creation/update records of sets may also be tampered with, so the security of semantic datasets cannot be guaranteed.
  • the knowledge management system, method, device, electronic equipment and storage medium provided by the present application can improve the security of semantic data sets.
  • the embodiment of the present application provides a knowledge management system, at least one knowledge management node, and the at least one knowledge management node is deployed on the block chain;
  • the knowledge management node is configured to perform the following processing:
  • the unique identifier of the target semantic data set is stored on the block chain.
  • the embodiment of the present application also provides a knowledge management method, which is executed by a knowledge management system, wherein the knowledge management system includes at least two knowledge management nodes deployed on the block chain, and the knowledge management method includes :
  • the knowledge management node obtains the target semantic data set input by the knowledge creator
  • the knowledge management node After the target semantic data set is stored, the knowledge management node obtains the unique identifier of the target semantic data set;
  • the knowledge management node After the knowledge management node determines that the target semantic data set is credible according to the smart contract stored on the block chain, it stores the unique identifier of the target semantic data set on the block chain.
  • the embodiment of the present application further provides a knowledge management device, including a module for implementing the operations performed by the knowledge management node in the method provided in the second aspect.
  • the embodiment of the present application also provides an electronic device, including: a processor, a communication interface, a memory, and a communication bus, and the processor, the memory, and the communication interface complete mutual communication through the communication bus. Communication;
  • the memory is used to store at least one executable instruction, and the executable instruction causes the processor to execute the operations performed by the knowledge management node in the method provided by the second aspect above.
  • the embodiment of the present application also provides a computer-readable storage medium, where computer instructions are stored on the computer-readable storage medium, and when the computer instructions are executed by a processor, the processor executes the above-mentioned The operation performed by the knowledge management node in the method provided by the second aspect.
  • the unique identifier of the target semantic data set is stored on the blockchain, that is, the corresponding unique identifier is stored in the block
  • the semantic data set on the chain is approved and credible, which guarantees the correctness and security of the target semantic data set from the source.
  • the unique identifier of the semantic data set is obtained through the blockchain, and the semantic data is verified through the obtained unique identifier, so as to Determine whether the acquired semantic data set has been tampered with, and send it to the user after determining that the semantic data set has not been tampered with, so as to ensure that the semantic data set sent to the user has not been tampered with, thereby further ensuring the semantic data set safety.
  • the knowledge management system further includes a data storage node, and the data storage node is deployed on a distributed storage system.
  • the knowledge management node sends the target semantic data set to the data storage node, and the data storage node stores the target semantic data set on the distributed storage system, and generates the first index file for indexing the ontology file in the target semantic data set.
  • the first index file is stored, the unique identifier of the first index file is calculated, and the unique identifier of the first index file is sent to the knowledge management node.
  • the first index file The unique identifier is stored on the block chain, wherein the first index file is linked to the second index file, and the second index file is used to index the ontology file in the semantic dataset from the knowledge management node last time.
  • the unique identifier of the first index file After storing the target semantic dataset on the distributed storage system, generate a first index file for indexing the target semantic dataset, store the first index file on the distributed storage system, and then calculate the first index file After verifying the credibility of the target semantic data set, the unique identifier of the first index file is stored on the blockchain, so that the unique identifier of the first index file will not be tampered with, so that the unique identifier of the first index file can be The unique identifier of the first index file verifies whether the storage path of the target semantic dataset in the distributed storage system has been tampered with, and can also search for the first index file from the distributed storage system through the unique identifier of the first index file, which is more convenient Search the semantic dataset created or updated by the user every time from the distributed storage system. When the user consults the semantic dataset, the user can feedback the required semantic dataset to the user more quickly, thereby improving the user experience.
  • the data storage node generates at least one domain folder according to at least one business domain involved in the target semantic dataset, and then generates the first domain folder for linking to each domain folder and the second index file Index file, in which, each business domain corresponds to a domain folder, and different business domains correspond to different domain folders.
  • Each domain folder records file links, and the file links correspond to the domain folders in the distributed storage system. Links to each ontology file of the business domain and the latest version, the ontology file includes the data of the business domain in the target semantic dataset.
  • the knowledge creator may create or update semantic datasets in different business domains at the same time, when storing the target semantic dataset input by the knowledge creator, create a domain folder according to the business domain involved in the target semantic dataset, so that the target semantic Different business domains involved in the data set correspond to different domain folders, and then according to the business domain to which each ontology file in the target semantic data set belongs, the link of the ontology file in the target semantic data set is recorded in the domain folder corresponding to the corresponding business domain, so that Each domain folder is linked to the latest version of the ontology file of its corresponding business domain, so that semantic datasets can be stored by domain and version, which is convenient for management and version traceability of semantic datasets.
  • the data storage node calculates the unique identifier of each domain folder and the unique identifier of the ontology file recorded in each domain folder, and calculates the unique identifier of each domain folder and each The unique identifier of the ontology file recorded in each domain folder is sent to the knowledge management node. After the knowledge management node determines that the target semantic dataset is credible, the unique identifier of each domain folder and the records recorded in each domain folder The unique identifier of the ontology file is stored on the blockchain.
  • the unique identifier of the first index file can be obtained by triggering the smart contract. According to the unique identifier of the first index file, from The first index file is obtained on the distributed storage system, and then the target semantic data set can be obtained through the first index file, so as to facilitate searching for the semantic data set required by the user from the distributed storage system.
  • the unique identification of each domain folder linked to by the first index file and the unique identification of each ontology file recorded in each domain folder are stored in the block chain, which can avoid that each domain folder linked to the first index file.
  • the unique identification of each domain folder and the unique identification of each ontology file recorded in each domain folder are tampered with, and then the unique identification of each domain folder linked to the first index file stored in the blockchain and the unique identification of each domain folder can be used.
  • the unique identifier of each ontology file recorded verifies whether the corresponding ontology file has been tampered with, and ensures the security of the semantic dataset.
  • the knowledge management node when the knowledge management node stores the unique identifier of the first index file on the blockchain, it encrypts the unique identifier of the first index file with a key set by the knowledge creator, And store the unique identifier of the encrypted first index file on the block chain.
  • the knowledge management system further includes at least one application node, and at least one application node is deployed on the block chain.
  • the application node obtains the knowledge query request of the user querying the target semantic data set, and sends the acquired knowledge query request to the knowledge management node; when the knowledge management node has stored the target semantic data set locally, the knowledge management node according to the knowledge query request After verifying the credibility of the target semantic data set through the smart contract, the target semantic data set is sent to the application node for the application node to visualize the target semantic data set.
  • the knowledge management node when the knowledge management node does not store the target semantic dataset locally, the knowledge management node obtains the unique identifier of the first index file from the blockchain according to the knowledge query request, and stores the first index file
  • the unique identifier of the data storage node is sent to the data storage node; the data storage node obtains the first index file from the distributed storage system according to the unique identifier of the first index file received, and obtains the first index file from the distributed storage system according to the obtained first index file
  • the knowledge management node When the user needs to access the target semantic data set, if the knowledge management node does not store the target semantic data set locally, the target semantic data set is obtained from the distributed storage system, and the target semantic data set is stored after verifying that the target semantic data set is credible local to the knowledge management node, and send the target semantic dataset to the application node for visualization.
  • the knowledge management node stores the semantic datasets accessed by users locally, which can reduce the number of visits to the distributed storage system, reduce the computing pressure of the knowledge management system, and return the required semantic datasets to users faster.
  • the block chain is a private chain. Since the unique identifier of the index file is stored on the block chain, and the smart contract is also stored on the block chain, and for security reasons, the index file and smart contract do not want all users to be able to access, and the private chain can restrict access to Access rights to content stored on the blockchain.
  • the distributed storage system can be an interplanetary file system, which can be better combined with the blockchain to achieve safe storage of data, thereby further ensuring the security of the managed knowledge.
  • FIG. 1 is a schematic diagram of a knowledge management system provided by an embodiment of the present application.
  • Fig. 2 is a schematic diagram of another knowledge management system provided by the embodiment of the present application.
  • FIG. 3 is a schematic diagram of a data storage structure in a distributed storage system provided by an embodiment of the present application
  • Fig. 4 is a schematic diagram of another knowledge management system provided by the embodiment of the present application.
  • Fig. 5 is a schematic diagram of a semantic data set storage method provided by the embodiment of the present application.
  • FIG. 6 is a flow chart of a knowledge management method provided in an embodiment of the present application.
  • FIG. 7 is a schematic diagram of a knowledge management device provided by an embodiment of the present application.
  • Fig. 8 is a schematic diagram of another knowledge management device provided by the embodiment of the present application.
  • FIG. 9 is a schematic diagram of an electronic device shut down according to an embodiment of the present application.
  • Identification acquisition module 703 Storage module 704: Transceiver module
  • the knowledge management node obtains the target semantic dataset input by the knowledge creator
  • the knowledge management node acquires the unique identifier of the target semantic dataset after the target semantic dataset is stored
  • semantic datasets, data models and knowledge graphs are all stored centrally on the server, and multiple parties involved in the preparation of semantic datasets can create new semantic datasets and The original semantic data set is updated, and the creation record and update record of the semantic data set can also be modified. Since all parties are free to create and update semantic datasets, and the creation and update records of semantic datasets can also be modified, there is a possibility of maliciously creating or updating semantic datasets by tampering with the creation and update records of semantic datasets. risk, so the security of the semantic dataset cannot be guaranteed.
  • the semantic data set is stored, and the unique identifier of the semantic data set is obtained, and the semantic data is determined according to the smart contract stored on the blockchain
  • the unique identifier of the semantic dataset is stored on the blockchain.
  • the credibility of the semantic data set is verified through the smart contract, which ensures the legality of the content of the semantic data set uploaded by the knowledge creator.
  • the unique identifier of the semantic data set is stored on the blockchain, and the semantic data can be guaranteed based on the characteristics of the blockchain.
  • the unique identifier of the dataset cannot be tampered with, and then it is possible to verify whether the stored semantic dataset has been tampered with based on the unique identifier of the semantic dataset stored on the blockchain, thereby ensuring the security of the semantic dataset.
  • the embodiment of the present application provides a knowledge management system 10, including: at least one knowledge management node 11, and each knowledge management node 11 is deployed on a block chain 20;
  • the knowledge management node 11 is used to obtain the target semantic data set input by the knowledge creator, after the target semantic data set is stored, obtain the unique identifier of the target semantic data set, and determine the target semantic data according to the smart contract stored on the block chain 20 After the data set is credible, the unique identifier of the target semantic data set is stored on the block chain 20 .
  • the unique identifier of the target semantic data set is stored on the blockchain, that is, the corresponding unique identifier
  • the semantic dataset stored on the blockchain is approved and credible, which guarantees the correctness and security of the target semantic dataset from the source.
  • the unique identifier of the semantic data set is obtained through the blockchain, and the semantic data is verified through the obtained unique identifier, so as to Determine whether the acquired semantic data set has been tampered with, and send it to the user after determining that the semantic data set has not been tampered with, so as to ensure that the semantic data set sent to the user has not been tampered with, thereby further ensuring the semantic data set safety.
  • the knowledge management node 11 is located in the blockchain 20, which may be a peer-to-peer (P2P) network, and the smart contract and the unique identifier of the semantic data set are stored on the blockchain 20 .
  • P2P peer-to-peer
  • the unique identifier of the semantic dataset is used to identify the semantic dataset, and different semantic datasets correspond to different unique identifiers, and it can be determined whether the semantic dataset has been tampered with according to the unique identifier.
  • the unique identifier of a semantic dataset can be the hash value, MD5 value, etc. of the semantic dataset, that is, the hash algorithm, MD5 algorithm and other encryption algorithms are used to calculate the semantic dataset to obtain the hash value that can uniquely identify the semantic dataset. Hash value, MD5 value, etc., as the unique identifier of the semantic dataset.
  • FIG. 2 is a schematic diagram of another knowledge management system 10 provided by the embodiment of the present application. As shown in FIG. 2 , the knowledge management system 10 also includes a data storage node 12, and the data storage node 12 is deployed on a distributed storage system 30;
  • the knowledge management node 11 is used to send the target semantic data set to the data storage node 12;
  • the data storage node 12 is used to store the received target semantic data set on the distributed storage system 30, generate a first index file for indexing ontology files in the target semantic data set, and store the first index file in On the distributed storage system 30, wherein, the first index file is linked to the second index file, and the second index file is used to index the ontology file in the semantic data set from the knowledge management node 11 last time;
  • the data storage node 12 is also used to calculate the unique identifier of the first index file, and send the unique identifier of the first index file to the knowledge management node 11;
  • the knowledge management node 11 is also used to store the unique identifier of the first index file on the block chain 20 after determining that the target semantic data set is credible.
  • the knowledge management node 11 after receiving the target semantic data set input by the knowledge creator, the knowledge management node 11 sends the target semantic data set to the data storage node 12, and the data storage node 12 stores the target semantic data set in the distributed storage system 30.
  • Storing the target semantic data set on the distributed storage system 30, based on the characteristics of the distributed storage system 30, can increase the difficulty of the target semantic data set being tampered with, and can reduce the risk of the target semantic data set being lost due to accidental reasons, To improve the security of the target semantic dataset.
  • the data storage node 12 After the data storage node 12 stores the target semantic data set on the distributed storage system 30, the data storage node 12 generates a first index file for indexing ontology files in the target semantic data set, and the first index file is also linked to The second index file is used to index the ontology file in the semantic data set received by the knowledge management node 11 last time.
  • the ontology file included in the target semantic data set can be obtained from the distributed storage system 30 through the first index file, since the target semantic data set is a semantic data set newly created by the knowledge creator or after the version update of the original semantic data set Therefore, after the target semantic dataset is stored on the distributed storage system 30, the ontology files included in the target semantic dataset are all the latest versions, that is, the latest version can be obtained from the distributed storage system 30 through the first index file. version of the ontology file.
  • the second index file is used to index the ontology file last stored in the semantic data set of the distributed storage system 30, and the second index file is also linked to the The index file generated by the semantic data set previously stored on the distributed storage system 30, that is, the second index file can be obtained through the first index file, and the old version of ontology can be obtained from the distributed storage system 30 through the second index file files, so that the version traceability and version management of the semantic dataset can be conveniently carried out through the index file.
  • the data storage node 12 After the data storage node 12 generates the first index file, it stores the first index file on the distributed storage system 30, then calculates the unique identifier of the first index file, and sends the unique identifier of the first index file to the knowledge management node 11.
  • the knowledge management node 11 stores the unique identifier of the first index file on the block chain 20 after verifying that the target semantic data set is credible.
  • the knowledge management node 11 After the knowledge management node 11 determines that the target semantic data set is credible, it stores the unique identifier of the first index file on the blockchain 20, so that the unique identifier of the first index file will not be tampered with, so that the unique identifier of the first index file can be The unique identifier of the first index file verifies whether the storage path of the target semantic dataset in the distributed storage system 30 has been tampered with, and the first index file can also be searched from the distributed storage system 30 through the unique identifier of the first index file.
  • the unique identifier of the first index file may be a hash value, an MD5 value, etc. of the first index file.
  • storing the target semantic data set on the distributed storage system is only an optional implementation of the embodiment of the present application.
  • the target semantic data set can be stored in the knowledge
  • the local storage of the management node 11 can either store the target semantic data set on the blockchain, or store the target semantic data set in the cloud server.
  • the embodiment of the present application does not limit the storage location of the target semantic data set.
  • the data storage node 12 after receiving the target semantic data set, the data storage node 12 generates at least one domain folder according to at least one business domain involved in the target semantic data set, and then generates a file for linking to each domain folder and the first index file of the second index file, wherein each business domain corresponds to a domain folder, different business domains correspond to different domain folders, and each domain folder records file links, which are distributed storage Links to the latest versions of ontology files belonging to the business domain corresponding to the domain folder in the system 30.
  • the ontology files include the data of the business domain in the target semantic dataset.
  • the knowledge creator may create or update semantic datasets in different business fields at the same time
  • create Domain folders so that different business domains involved in the target semantic dataset correspond to different domain folders, and then according to the business domain to which each ontology file in the target semantic dataset belongs, the link of the ontology file in the target semantic dataset is recorded to the corresponding business domain
  • each domain folder is linked to the latest version of the ontology file of its corresponding business domain, so that the semantic dataset can be stored by domain and version, which is convenient for the management and version traceability of the semantic dataset .
  • the target semantic dataset is a semantic dataset newly created by the knowledge creator or a semantic dataset after the version update of the original semantic dataset
  • the target semantic dataset is stored in the distributed storage
  • the ontology files included in the target semantic dataset are all the latest versions, so the links of the ontology files in the target semantic dataset will be recorded in corresponding domain folders.
  • the knowledge management system 10 can manage knowledge in different business fields, in order to more conveniently manage knowledge belonging to different business fields, domain folders can be generated according to business domains, so as to obtain The corresponding ontology file.
  • the domain folder newly created by the data storage node 12 is called the first domain folder
  • the domain folder previously created by the data storage node 12 is called the second domain folder. Since the knowledge creator may not update all semantic datasets stored on the distributed storage system 30 at the same time, the first domain folder and the second domain folder may record links to the same ontology file.
  • the recently generated second domain folder contains links to ontology file 1-V1, ontology file 2-V2 and ontology file 3-V1, but the knowledge creator only updates ontology file 1-V1 to ontology file 1- V2, after the knowledge creator updates ontology file 1-V1 to ontology file 1-V2, ontology file 1-V2, ontology file 2-V2 and ontology file 3-V1 are recorded in the generated first domain folder the link to.
  • the data storage node 12 after generating each domain folder for the target semantic data set, the data storage node 12 will also generate a first index file for linking to each domain folder. Since the first index file is linked to each domain folder, each domain folder can be obtained after obtaining the first index file, and then the ontology file can be obtained according to the file links recorded in the domain folder.
  • the index files linked to each first domain folder are referred to as first index files
  • the index files linked to each second domain folder are referred to as second index files.
  • an index file will be generated, which is linked to The latest versions of each ontology file in the corresponding business domain after the update of the semantic dataset are recorded in the domain folders obtained, so that the user can obtain the latest version of each ontology file after the update of the semantic dataset according to the index file, so that Semantic datasets can be easily versioned using index files.
  • the newly generated index file will record the index file link pointing to the previously generated index file. Through the index file link, each old version of the ontology file can be obtained, so that the version of the semantic dataset can be easily traced through the index file.
  • the latest domain folder generated for this business domain can be linked to the domain folder previously generated for this business domain, that is, the domain folder that records the new version of the ontology file. Link to the domain folder that records the old version of the ontology file, so that the version of the semantic dataset can be easily traced through the domain folder.
  • the first index file is linked to the second index file, specifically, at least one index file link pointing to the second index file is recorded in the first index file, and when an index file link is recorded in the first index file, the index file link Can point to the second domain folder generated recently (that is, the domain folder formed according to the semantic data set stored in the distributed storage system 30 last time), when multiple index file links are recorded in the first index file , these index file links may point to the most recently generated N (N>2) second domain folders in turn.
  • each domain folder records the latest version of each ontology file of the corresponding business domain after the semantic dataset is created or updated, and each generated domain folder is linked to an index file, so the index file can
  • the data storage node 12 will store each index file on the distributed storage system 30, and the knowledge management node 11 will use the unique identifier of each index file as a transaction
  • the records are stored on the block chain 20, thereby ensuring that the creation records and update records of the semantic dataset cannot be tampered with, further improving the security of the semantic dataset.
  • the data storage node 12 is also used to calculate the unique identifier of each domain folder and the ontology file recorded in each domain folder (that is, the ontology corresponding to the file link recorded in the domain folder file), and send the unique identifier of each domain folder and the unique identifier of the ontology file recorded in each domain folder to the knowledge management node 11.
  • the knowledge management node 11 stores the received unique identifier of each domain folder and the unique identifier of the ontology file recorded in each domain folder on the block chain 20 .
  • the knowledge management node 11 will store the unique identifier of the first index file on the blockchain 20, and will link the first index file to the The unique identification of each domain folder and the unique identification of each ontology file recorded in each domain folder are stored in the block chain 20 . Store the unique identifier of the first index file on the blockchain.
  • the unique identifier of the first index file can be obtained by triggering the smart contract, and the unique identifier of the first index file will be sent to After the data storage node 12, the data storage node 12 can obtain the first index file from the distributed storage system 30 according to the unique identifier of the first index file, and then can obtain the target semantic data set through the first index file.
  • the unique identification of each domain folder linked to by the first index file and the unique identification of each ontology file recorded in each domain folder are stored in the block chain 20, which can avoid that each domain file linked to the first index file.
  • the unique identification of each domain folder and the unique identification of each ontology file recorded in each domain folder are tampered with, and then the unique identification of each domain folder linked to the first index file stored in the block chain 20 and the unique identification of each domain folder can be used.
  • the unique identification of each ontology file recorded in the folder and verify whether the corresponding ontology file has been tampered with.
  • the unique identifier of the domain folder can be the hash value, MD5 value, etc. of the domain folder
  • the unique identifier of the ontology file can be the hash value or MD5 value of the ontology file wait.
  • FIG. 3 schematically shows the structural form of data storage in the distributed storage system 30.
  • the distributed storage system 30 stores ontology files OF1-V1, ontology files OF2-V1, and ontology files OF1.
  • the domain folder DF-V1 is linked to the ontology file OF1-V1, ontology file OF2-V1, and ontology file OF3 (that is, the ontology file OF1-V1, ontology file OF2-V1, and ontology file OF3 are recorded in the domain folder DF-V1 file link), domain folder DF-V2 is linked to ontology file OF1-V2, ontology file OF2-V1 and ontology file OF3 (that is, ontology file OF1-V2, ontology file OF2-V1 and ontology file OF2-V1 are recorded in domain folder DF-V2 File link to ontology file OF3).
  • the index file C1 is linked to the domain folder DF-V1
  • the index file C2 is linked to the domain folder DF-V2
  • the index file Cn is linked to the domain folder DF-Vn (the domain folder DF-Vn is not shown in the figure)
  • An index file link P1 pointing to the index file C1 is recorded in the index file C2
  • an index file link Pn-1 pointing to the index file Cn-1 is recorded in the index file Cn.
  • the index file C1 to the index file Cn are sequentially stored in each block 1 to block n in the block chain 20 , wherein block n is located at the end of the block chain 20 .
  • FIG. 3 is only an example of the storage structure in the distributed storage system 30, and the actual distributed storage system 30 includes more ontology files, domain folders and index files, and there may be links to the same index file to multiple domain folders.
  • the knowledge management node 11 when the knowledge management node 11 stores the unique identifier of the first index file on the block chain 20, it first uses the key set by the knowledge creator to perform the unique identifier of the first index file. Encrypt, and then store the encrypted unique identifier on the block chain 20.
  • the knowledge creator will set a key before uploading the target semantic dataset, and the knowledge management node 11 will use the key set by the knowledge creator to encrypt the unique identifier of the first index file, and then the encrypted
  • the final unique identifier is stored on the block chain 20. Since the unique identifier of the first index file is used to search for the first index file from the distributed storage system 30, by encrypting the unique identifier of the first index file, only verified users can obtain the first index file The unique identifier of the first index file, and then access the data on the distributed storage system 30 according to the unique identifier of the first index file, so that the security of the data on the distributed storage system 30 can be guaranteed.
  • the knowledge management system 10 may further include at least one application node 13, wherein each application node 13 is deployed in on blockchain 20;
  • the application node 13 is used to obtain the knowledge query request of the user querying the target semantic data set, and send the knowledge query request to the knowledge management node 11;
  • the knowledge management node 11 After the knowledge management node 11 receives the knowledge query request, if the target semantic data set is stored locally in the knowledge management node 11, it will send the target semantic data set to To the application node 13, so that the application node 13 can visualize the target semantic data set.
  • the knowledge management node 11 when the application node 13 sends the knowledge query request to the knowledge management node 11, the knowledge management node 11 first checks whether the target semantic data set that the user wants to access has been stored locally according to the knowledge query request, if the knowledge management node 11 The target semantic data set has been stored locally. After the knowledge management node 11 verifies the credibility of the locally stored target semantic data set through the smart contract, it sends the locally stored target semantic data set to the application node 13 for the application node 13 to Target semantic datasets for visualization.
  • the semantic data set required by the user is stored locally in the knowledge management node 11, after verifying that the locally stored semantic data set is credible, the locally stored semantic data set is sent to the user, thereby eliminating the need for frequent access to the distributed storage system 30 Obtaining the semantic data set can reduce the network bandwidth required by the knowledge management system 10 on the one hand, and can also increase the rate of feedback of the semantic data set to the user, thereby improving the user experience.
  • the smart contract when the knowledge management node 11 verifies the target semantic data set through the smart contract, the smart contract can obtain the unique identifier of the target semantic data set from the blockchain 20, and the smart contract follows the preset
  • the encryption algorithm calculates the target semantic dataset to obtain the unique identifier of the target semantic dataset, and then compares the calculated unique identifier with the unique identifier obtained from the blockchain 20, and if the two are the same, determine the target semantic The dataset is credible, and if the two are different, it is determined that the target semantic dataset is not credible.
  • the knowledge management node 11 after the knowledge management node 11 receives the knowledge query request, if the target semantic data set that the user wants to access is not stored locally in the knowledge management node 11, then according to the knowledge query request Obtain the unique identifier of the first index file from the block chain 20, and send the unique identifier of the first index file to the data storage node 12.
  • the data storage node 12 obtains the first index file from the distributed storage system 30 according to the received unique identifier of the first index file, and then obtains the first index file from the distributed storage system 30 according to the obtained first index file Link to each domain folder, then according to each file link recorded in each domain folder obtained, obtain the target semantic data set from the distributed storage system 30, and send the acquired target semantic data set to the knowledge management Node 11.
  • the knowledge management node 11 verifies whether the received target semantic data set is credible through the smart contract, and after determining that the target semantic data set is credible, stores the target semantic data set locally in the knowledge management node 11, and sends the target semantic data set to To the application node 13, so that the application node 13 can visualize the target semantic data set.
  • the knowledge management node 11 when the application node 13 sends the knowledge query request to the knowledge management node 11, the knowledge management node 11 first checks whether the target semantic data set that the user wants to access has been stored locally according to the knowledge query request, if the knowledge management node 11 If the target semantic data set is not stored locally, then the target semantic data set is obtained from the distributed storage system 30 through the application node 13, and after verifying that the target semantic data set is credible, the target semantic data set is stored locally in the knowledge management node 11, And send the target semantic data set to the application node 13.
  • the knowledge management node 11 locally stores the semantic data sets accessed by users, which can reduce the number of visits to the distributed storage system 30, reduce the computing pressure of the knowledge management system 10, and return the required semantic data sets to users faster .
  • the knowledge management system 10 further includes: a coordinating node and at least one knowledge approval node. On block chain 20;
  • the coordinating node is used to assign accounts to users. Different accounts have different permissions. After the user logs in to the device through the account assigned by the coordinating node, according to the different permissions of the logged-in account, the device acts as the knowledge management node 11 and the application in the knowledge management system 10. Node 13 or knowledge approval node;
  • the coordinating node also distributes different user identifications for each knowledge management node 11, each application node 13, and each knowledge approval node, so that knowledge management nodes 11, application nodes 13, and knowledge approval nodes can be based on each other's user IDs. identification for information exchange.
  • the knowledge management node 11 and the coordinating node are at the core of the entire knowledge management system 10, in order to ensure the safety of the knowledge management node 11 and the coordinating node, the knowledge management node 11 and the coordinating node have the highest voting authority Smart contract selection, and the results of voting can be updated at configurable intervals.
  • the coordinating node can assign different user IDs to each knowledge management node 11, each application node 13, and each knowledge approval node, so as to determine the group to which each node belongs through the user ID, so as to facilitate knowledge
  • the management node 11 can perform information interaction with each application node 13 and each knowledge approval node according to the user identification, so as to ensure the orderly and safe operation of the entire knowledge management system 10 .
  • the knowledge management node 11 after acquiring the target semantic data set uploaded by the knowledge creator 40, the knowledge management node 11 sends the target semantic data set to the data storage node 12 on the one hand, and the data storage node 12 Store the target semantic data set on the distributed storage system 30.
  • the knowledge management node 11 will generate a knowledge approval request for requesting the knowledge approver to approve the target semantic data set, and according to the smart contract 21, the generated The knowledge approval request is sent to at least one knowledge approval node 14, and the knowledge approval node 14 that receives the knowledge approval request sends the approval result given by the knowledge approver to the smart contract 21, and the smart contract 21 determines according to the received approval results Whether the target semantic data set is credible, the knowledge management node 11 stores the unique identifier of the target semantic data set on the block chain 20 after obtaining the result that the smart contract 21 determines that the target semantic data set is credible.
  • the knowledge management node 11 will not store the unique identifier of the target semantic data set on the block chain 20, so as to illustrate that the target semantic data set is unreliable. Faithful. Furthermore, after determining that the target semantic data set is untrustworthy, the knowledge management node 11 can delete the target semantic data set from the distributed storage system 30 according to the smart contract, or retain the target semantic data set in the distributed storage system 30 , but set an untrustworthy flag for the target semantic dataset to give an untrustworthy prompt when the user queries the target semantic dataset.
  • the knowledge approver is an expert in the field.
  • the knowledge approval request is sent to one or more knowledge approval nodes, and the corresponding knowledge approver approves the target semantic data set and gives the corresponding
  • the smart contract determines whether the target semantic data set is credible according to the approval results returned by each knowledge approval node.
  • the smart contract can determine whether the target semantic data set is credible according to the proportion of approval in each approval result. For example, when more than 80% of the approval results are approved, it is determined that the target semantic data set is credible.
  • the target semantic data set acquired by the knowledge management node 11 may be a new semantic data set created by the knowledge creator, or an updated version of the stored semantic data set by the knowledge creator, so that the knowledge
  • the management system 10 can securely manage newly created semantic datasets, and can also securely manage different versions of semantic datasets updated by knowledge creators.
  • the knowledge management node 11 can also generate a target knowledge graph according to the target semantic data set, and store the target semantic data set and the generated target knowledge graph in the distributed storage system 30 through the data storage node 12 .
  • the knowledge management node 11 will also generate a knowledge approval request for requesting approval of the target knowledge map, and send the knowledge approval request to at least one knowledge approval node according to the smart contract, and the knowledge approval node can send the approver to the The approval result of the target knowledge graph is sent to the smart contract.
  • the knowledge management node 11 stores the unique identifier of the target knowledge graph on the blockchain 20 . It can be seen that the knowledge management system 10 provided by the embodiment of the present application can not only manage the semantic data set to ensure the security of the semantic data set, but also manage the knowledge map generated based on the semantic data set to ensure The security of knowledge graph.
  • each part of the knowledge management system 10 provided in the embodiment of the present application can process the semantic data set alone, or can process both the semantic data set and the knowledge map generated based on the semantic data set.
  • the knowledge map and semantic data set generated according to the semantic data set can be used as parallel knowledge management objects, and the various parts included in the knowledge management system 10 can adopt the same method for managing the semantic data set Manage knowledge graphs.
  • the block chain 20 may be a private chain. Since the unique identifier of the index file is stored on the block chain 20, and the smart contract is also stored on the block chain 20, and for security reasons, the index file and smart contract do not want all users to be able to access, through the private chain. Restricting access rights to content stored on the blockchain 20 is more suitable for the application scenario of the knowledge management system 10 provided in this application.
  • the distributed storage system 30 can be an InterPlanetary File System (InterPlanetary File System, IPFS), and the InterPlanetary File System can be better Combining with the block chain 20 realizes safe storage of data, thereby further ensuring the security of the knowledge managed by the knowledge management system 10 provided in this application.
  • IPFS InterPlanetary File System
  • the distributed storage system involved in the following knowledge management method can be the aforementioned distributed storage system 30, the knowledge management node involved in the following knowledge management method can be the aforementioned knowledge management node 11, and the following knowledge
  • the data storage node involved in the management method can be the aforementioned data storage node 12, the application node involved in the following knowledge management method can be the aforementioned application node 13, and the blockchain involved in the following knowledge management method can be the aforementioned area Blockchain 20.
  • an embodiment of the present application provides a knowledge management method 60, which may include the following steps:
  • the knowledge management node acquires the target semantic data set input by the knowledge creator;
  • the knowledge management node After the target semantic dataset is stored, the knowledge management node acquires the unique identifier of the target semantic dataset;
  • the knowledge management node After determining that the target semantic dataset is credible according to the smart contract stored on the blockchain, the knowledge management node stores the unique identifier of the target semantic dataset on the blockchain.
  • the unique identifier of the target semantic data set is stored on the blockchain, that is, the corresponding unique identifier
  • the semantic dataset stored on the blockchain is approved and credible, which guarantees the correctness and security of the target semantic dataset from the source.
  • the unique identifier of the semantic data set is obtained through the blockchain, and the semantic data is verified through the obtained unique identifier, so as to Determine whether the acquired semantic data set has been tampered with, and send it to the user after determining that the semantic data set has not been tampered with, so as to ensure that the semantic data set sent to the user has not been tampered with, thereby further ensuring the semantic data set safety.
  • the target semantic dataset may be stored on a distributed storage system through a data storage node.
  • the storage of the target semantic dataset can be achieved through the following steps:
  • the knowledge management node sends the target semantic dataset to the data storage node
  • the data storage node stores the target semantic dataset on the distributed storage system
  • the data storage node generates the first index file for indexing the ontology file in the target semantic dataset, wherein the first index file is linked to the second index file, and the second index file is used to index the previous index file from the knowledge management node. index the ontology files in the semantic dataset;
  • the data storage node stores the first index file
  • the data storage node calculates the unique identifier of the first index file, and sends the unique identifier of the first index file to the knowledge management node;
  • the knowledge management node After determining that the target semantic data set is credible, the knowledge management node stores the unique identifier of the first index file on the blockchain.
  • the target semantic data set after receiving the target semantic data set input by the knowledge creator, the target semantic data set is stored on the distributed storage system. Based on the characteristics of the distributed storage system, the target semantic data set can be increased The difficulty of tampering, and at the same time, it can reduce the risk of loss of the target semantic dataset due to accidental reasons, so as to improve the security of the target semantic dataset.
  • a first index file for indexing the target semantic dataset is generated, and the first index file is stored on the distributed storage system , and then calculate the unique identifier of the first index file, after verifying the credibility of the target semantic dataset, store the unique identifier of the first index file on the blockchain, so that the unique identifier of the first index file will not be tampered with, thus It can be verified whether the storage path of the target semantic dataset in the distributed storage system has been tampered with according to the unique identifier of the first index file on the blockchain, and it is also possible to find the first index file from the distributed storage system through the unique identifier of the first index file.
  • An index file so that it is more convenient to find the semantic dataset created or updated by the user each time from the distributed storage system. When the user consults the semantic dataset, the required semantic dataset can be fed back to the user more quickly, thereby improving User experience.
  • At least one domain folder may be generated according to at least one business domain involved in the target semantic dataset, and then generated for linking to each domain folder and the second
  • the first index file of the index file wherein each business domain corresponds to a domain folder, and different business domains correspond to different domain folders, and each domain folder records a file link, and the file link belongs to the distributed storage system.
  • the ontology file includes the data of the business domain in the target semantic dataset.
  • the knowledge creator may create or update semantic datasets in different business fields at the same time
  • create Domain folders so that different business domains involved in the target semantic dataset correspond to different domain folders, and then according to the business domain to which each ontology file in the target semantic dataset belongs, the link of the ontology file in the target semantic dataset is recorded to the corresponding business domain
  • each domain folder is linked to the latest version of the ontology file of its corresponding business domain, so that the semantic dataset can be stored by domain and version, which is convenient for the management and version traceability of the semantic dataset .
  • the unique identifier of each domain folder and the unique identifier of each domain folder can be calculated.
  • the unique identifier of the first index file is stored on the block chain.
  • the unique identifier of the first index file can be obtained by triggering the smart contract.
  • the first index file is obtained from the distributed storage system, and then the target semantic data set can be obtained through the first index file, so as to facilitate the search for the semantic data set required by the user from the distributed storage system.
  • the unique identification of each domain folder linked to by the first index file and the unique identification of each ontology file recorded in each domain folder are stored in the block chain, which can avoid that each domain folder linked to the first index file.
  • the unique identification of each domain folder and the unique identification of each ontology file recorded in each domain folder are tampered with, and then the unique identification of each domain folder linked to the first index file stored in the blockchain and the unique identification of each domain folder can be used.
  • the unique identifier of each ontology file recorded verifies whether the corresponding ontology file has been tampered with, and ensures the security of the semantic dataset.
  • the unique identifier of the first index file when storing the unique identifier of the first index file on the block chain, first encrypt the unique identifier of the first index file with the key set by the knowledge creator, and then The encrypted unique identifier is stored on the blockchain.
  • the unique identifier of the first index file is used to search for the first index file from the distributed storage system, by encrypting the unique identifier of the first index file, only authenticated users can obtain The unique identifier of the first index file is obtained, and then the data on the distributed storage system is accessed according to the unique identifier of the first index file, thereby ensuring the security of the data on the distributed storage system.
  • the knowledge management method 60 further includes:
  • the application node obtains the knowledge query request of the user querying the target semantic data set
  • the application node sends the acquired knowledge query request to the knowledge management node;
  • the knowledge management node When the knowledge management node has stored the target semantic data set locally, the knowledge management node will send the target semantic data set to the application node after verifying the credibility of the target semantic data set through the smart contract, so that the application node can visualize the target semantic data set .
  • the semantic data set accessed by the user is cached locally in the knowledge management node.
  • the user needs to access a semantic data set, first check whether the local knowledge management node caches the semantic data set that the user wants to access , if the local cache of the knowledge management node has the semantic data set that the user wants to access, there is no need to obtain data from the distributed storage system, and the speed of feeding back the required semantic data set to the user can be improved, thereby improving the user experience.
  • the knowledge management method 60 further includes:
  • the knowledge management node When the knowledge management node does not store the target semantic data set locally, the knowledge management node obtains the unique identifier of the first index file from the smart contract according to the knowledge query request;
  • the knowledge management node sends the unique identifier of the first index file to the data storage node
  • the data storage node acquires the first index file from the distributed storage system according to the received unique identifier of the first index file;
  • the data storage node obtains the domain folders linked to by the first index file from the distributed storage system according to the obtained first index file;
  • the data storage node obtains the target semantic data set from the distributed storage system according to the obtained file links recorded in each domain folder, and sends the obtained target semantic data set to the knowledge management node;
  • the knowledge management node After the knowledge management node verifies the credibility of the target semantic data set obtained through the smart contract, it stores and caches the target semantic data set locally in the knowledge management node;
  • the knowledge management node sends the target semantic data set to the application node, so that the application node can visualize the target semantic data set.
  • the target semantic data set when the user needs to access the target semantic data set, if the knowledge management node does not store the target semantic data set locally, the target semantic data set is obtained from the distributed storage system, and after verifying that the target semantic data set is credible , store the target semantic dataset locally in the knowledge management node, and send the target semantic dataset to the application node for visualization.
  • Knowledge management nodes store semantic data sets (ontology files) accessed by users locally, which can reduce the number of visits to distributed storage systems, reduce the computing pressure of knowledge management systems, and return the required semantic data to users faster set.
  • the block chain may be a private chain. Since the unique identifier of the index file is stored on the block chain, and the smart contract is also stored on the block chain, and for security reasons, the index file and smart contract do not want all users to be able to access, and the private chain can restrict access to Access rights to content stored on the blockchain.
  • the distributed storage system can be an InterPlanetary File System (IPFS), and the InterPlanetary File System can be better integrated with the blockchain The combination realizes the secure storage of data, thereby further ensuring the security of the managed knowledge.
  • IPFS InterPlanetary File System
  • FIG. 7 is a schematic diagram of a knowledge management device 700 provided by an embodiment of the present application.
  • the knowledge management device 700 includes modules for implementing the operations performed by the knowledge management node 11 in the above method embodiment. Specifically, as shown in FIG. 7 , the knowledge management device 700 includes:
  • Data acquisition module 701 for acquiring the target semantic data set input by the knowledge creator
  • An identification acquisition module 702 configured to obtain the unique identification of the target semantic data set after the target semantic data set acquired by the data acquisition module 701 is stored;
  • the storage module 703 is configured to store the unique identifier of the target semantic dataset on the blockchain after determining that the target semantic dataset is credible according to the smart contract stored on the blockchain.
  • the identification acquisition module 702 acquires the unique identification of the target semantic data set, and after the target semantic data set is verified and trusted by the smart contract, the storage module 703 stores The unique identifier of the semantic dataset is stored in the blockchain. Due to the non-tamperability of the blockchain, the semantic dataset corresponding to the unique identifier stored on the blockchain is credible, which guarantees the correctness and security of the target semantic dataset from the source. In addition, the storage module 703 stores the unique identifier of the target semantic dataset on the block chain.
  • the target semantic dataset can be verified through the unique identifier stored on the blockchain to determine the target Whether the semantic data set has been tampered with, after determining that the target semantic data set has not been tampered with, send the target semantic data set to the user, so as to ensure that the target semantic data set sent to the user has not been tampered with, thereby further ensuring semantics Dataset security.
  • FIG. 8 is a schematic diagram of another knowledge management device 700 provided in the embodiment of the present application. As shown in FIG. 8, the knowledge management device 700 further includes:
  • the transceiver module 704 is configured to send the target semantic data set to the data storage node, and the data storage node stores the target semantic data set on the distributed storage system, and receives the unique identifier of the first index file from the data storage node, and sends the The unique identification of the first index file is sent to the storage module 703, wherein the first index file is used to index the ontology file included in the target semantic dataset in the distributed storage system, the first index file is linked to the second index file, and the first index file is linked to the second index file.
  • the second index file is used to index the ontology file in the semantic data set acquired last time by the data acquisition module 701 in the distributed storage system;
  • the storage module 703 is further configured to store the unique identifier of the first index file on the block chain after determining that the target semantic data set is credible.
  • the transceiver module 704 is also configured to receive the unique identifier of each domain folder from the data storage node, and the unique identifier of the ontology file recorded in each domain folder, and receive the The unique identifier of each domain folder and the unique identifier of the ontology file recorded in each domain folder are sent to the storage module 703, and the storage module 703 is also used to store each The unique identifier of the domain folder and the unique identifier of the ontology file recorded in each domain folder are stored on the blockchain.
  • the storage module 703 is also used to use the key set by the knowledge creator to uniquely identify the first index file when storing the unique identifier of the first index file on the block chain. Encrypt and store the unique identifier of the encrypted first index file on the block chain.
  • the transceiver module 704 is also used to receive the knowledge query request from the application node. After the target semantic data set locally stored in the knowledge management node is credible, the target semantic data set is sent to the application node for the application node to visualize the target semantic data set.
  • the transceiver module 704 is also configured to receive a knowledge query request from an application node. If the target semantic data set to be accessed by the knowledge query request is not stored locally in the knowledge management node, then according to the knowledge query request from Obtain the unique identifier of the first index file on the block chain, and send the obtained unique identifier of the first index file to the data storage node.
  • the transceiver module 704 is also used to receive the target semantic data set from the data storage node, after verifying the credibility of the target semantic data set through the smart contract, store the target semantic data set locally in the knowledge management node, and store the target semantic data set Sent to the application node for the application node to visualize the target semantic dataset.
  • FIG. 9 is a schematic diagram of an electronic device provided by an embodiment of the present application, and the specific embodiment of the present application does not limit the specific implementation of the electronic device.
  • an electronic device 900 provided by an embodiment of the present application includes: a processor (processor) 902 , a communication interface (Communications Interface) 904 , a memory (memory) 906 , and a communication bus 908 . in:
  • the processor 902 , the communication interface 904 , and the memory 906 communicate with each other through the communication bus 908 .
  • the communication interface 904 is used for communicating with other electronic devices or servers.
  • the processor 902 is configured to execute the program 910, and may specifically execute the operations performed by the knowledge management node 11 in any one of the foregoing knowledge management method embodiments.
  • the program 910 may include program codes including computer operation instructions.
  • the processor 902 may be a central processing unit CPU, or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.
  • the one or more processors included in the smart device may be of the same type, such as one or more CPUs, or may be different types of processors, such as one or more CPUs and one or more ASICs.
  • the memory 906 is used to store the program 910 .
  • the memory 906 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.
  • the program 910 may be specifically configured to enable the processor 902 to perform operations performed by the knowledge management node 11 in any of the aforementioned knowledge management method embodiments.
  • the unique identifier of the target semantic data set is stored on the blockchain, that is, the corresponding
  • the semantic dataset stored on the blockchain with unique identification is approved and credible, which guarantees the correctness and security of the target semantic dataset from the source.
  • the unique identifier of the semantic data set is obtained through the blockchain, and the semantic data is verified through the obtained unique identifier, so as to Determine whether the acquired semantic data set has been tampered with, and send it to the user after determining that the semantic data set has not been tampered with, so as to ensure that the semantic data set sent to the user has not been tampered with, thereby further ensuring the semantic data set safety.
  • the present application also provides a computer-readable storage medium, which stores instructions for enabling a machine to perform operations performed by the knowledge management node 11 in any of the foregoing knowledge management method embodiments.
  • a system or device equipped with a storage medium may be provided, on which a software program code for realizing the functions of any of the above embodiments is stored, and the computer (or CPU or MPU of the system or device) ) to read and execute the program code stored in the storage medium.
  • the program code itself read from the storage medium can realize the function of any one of the above-mentioned embodiments, so the program code and the storage medium storing the program code constitute a part of the present application.
  • Examples of storage media for providing program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), Tape, non-volatile memory card, and ROM.
  • the program code can be downloaded from a server computer via a communication network.
  • the program code read from the storage medium is written into the memory provided in the expansion board inserted into the computer or written into the memory provided in the expansion module connected to the computer, and then based on the program code
  • the instruction causes the CPU installed on the expansion board or the expansion module to perform some or all of the actual operations, thereby realizing the functions of any one of the above-mentioned embodiments.
  • the hardware modules may be implemented mechanically or electrically.
  • a hardware module may include permanently dedicated circuitry or logic (such as a dedicated processor, FPGA or ASIC) to perform the corresponding operations.
  • the hardware modules may also include programmable logic or circuits (such as general-purpose processors or other programmable processors), which can be temporarily set by software to complete corresponding operations.
  • the specific implementation mechanical way, or a dedicated permanent circuit, or a temporary circuit

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

本申请提供了知识管理系统、方法、装置、电子设备和存储介质，该知识管理系统包括：至少一个知识管理节点，所述至少一个知识管理节点部署在区块链上；所述知识管理节点，用于执行如下处理：获取知识创建者输入的目标语义数据集；在所述目标语义数据集被存储后，获取所述目标语义数据集的唯一标识；在根据所述区块链上存储的智能合约，确定所述目标语义数据集可信后，将所述目标语义数据集的唯一标识存储在所述区块链上。本方案能够提高语义数据集的安全性。

Description

知识管理系统、方法、装置、电子设备和存储介质 技术领域
本申请涉及知识管理技术领域，尤其涉及知识管理系统、方法、装置、电子设备和存储介质。
背景技术
知识图谱(Knowledge Graph)是将知识结构化形成的知识系统，其中包含了基本事实、通用规则和其他有关信息，常用于人工智能领域的知识抽取、存储与推理等任务。语义数据集(Semantics Datasets)可以用于构建知识图谱，用户可以利用基于语义数据集所构建的知识图谱开发应用程序，比如开发数据检索应用。
目前，通常采用集中式管理系统对知识进行管理，当存在多方参与语义数据集的准备工作时，各方均可以创建新的语义数据集并可以对已创建的语义数据集进行更新，而且语义数据集的创建/更新记录也可能被篡改，因此无法保证语义数据集的安全性。
发明内容
有鉴于此，本申请提供的知识管理系统、方法、装置、电子设备和存储介质，能够提高语义数据集的安全性。
第一方面，本申请实施例提供了一种知识管理系统，至少一个知识管理节点，所述至少一个知识管理节点部署在区块链上；
所述知识管理节点，用于执行如下处理：
获取知识创建者输入的目标语义数据集；
在所述目标语义数据集被存储后，获取所述目标语义数据集的唯一标识；
在根据所述区块链上存储的智能合约，确定所述目标语义数据集可信后，将所述目标语义数据集的唯一标识存储在所述区块链上。
第二方面，本申请实施例还提供了一种知识管理方法，由知识管理系统执行，其中所述知识管理系统包括部署在区块链上的至少两个知识管理节点，所述知识管理方法包括：
所述知识管理节点获取知识创建者输入的目标语义数据集；
所述知识管理节点在所述目标语义数据集被存储后，获取所述目标语义数据集的唯一标 识；
所述知识管理节点在根据所述区块链上存储的智能合约，确定所述目标语义数据集可信后，将所述目标语义数据集的唯一标识存储在所述区块链上。
第三方面，本申请实施例还提供了一种知识管理装置，包括用于实现上述第二方面提供的方法中所述知识管理节点所执行操作的模块。
第四方面，本申请实施例还提供了一种电子设备，包括：处理器、通信接口、存储器和通信总线，所述处理器、所述存储器和所述通信接口通过所述通信总线完成相互间的通信；
所述存储器用于存放至少一可执行指令，所述可执行指令使所述处理器执行上述第二方面提供的方法中所述知识管理节点执行的操作。
第五方面，本申请实施例还提供了一种计算机可读存储介质，所述计算机可读存储介质上存储有计算机指令，所述计算机指令在被处理器执行时，使所述处理器执行上述第二方面提供的方法中所述知识管理节点执行的操作。
其中，对于知识创建者上传的目标语义数据集，在通过智能合约验证目标语义数据集可信后，将目标语义数据集的唯一标识存储到区块链上，即对应唯一标识被存储在区块链上的语义数据集是经审批可信的，从源头上保证了目标语义数据集的正确性和安全性。另外，当用户查询语义数据集时，在获取到用户所需的语义数据集后，通过区块链上获取该语义数据集的唯一标识，通过获取到的唯一标识对该语义数据进行验证，以确定获取到的语义数据集是否被篡改，在确定该语义数据集未被篡改后发送给用户，从而可以确保发送给用户的语义数据集是未被篡改过的，从而可以进一步保证语义数据集的安全性。
对于上述任一方面，可选地，知识管理系统还包括数据存储节点，所述数据存储节点部署在分布式存储系统上。知识管理节点将目标语义数据集发送给数据存储节点，数据存储节点将目标语义数据集存储在分布式存储系统上，生成用于对目标语义数据集中的本体文件进行索引的第一索引文件，对第一索引文件进行存储，计算第一索引文件的唯一标识，并将第一索引文件的唯一标识发送给知识管理节点，知识管理节点在确定目标语义数据集可信后，将第一索引文件的唯一标识存储到区块链上，其中，第一索引文件链接到第二索引文件，第二索引文件用于对上一次来自知识管理节点的语义数据集中的本体文件进行索引。
将目标语义数据集存储到分布式存储系统上之后，生成用于对目标语义数据集进行索引 的第一索引文件，并将第一索引文件存储在分布式存储系统上，然后计算第一索引文件的唯一标识，在验证目标语义数据集可信后，将第一索引文件的唯一标识存储在区块链上，使得第一索引文件的唯一标识不会被篡改，从而可以根据区块链上的第一索引文件的唯一标识验证分布式存储系统中的目标语义数据集的存储路径是否被篡改，还能够通过第一索引文件的唯一标识从分布式存储系统上查找第一索引文件，从而更加方便从分布式存储系统上查找用户每次创建或更新后的语义数据集，用户在查阅语义数据集时，可以更加快速的向用户反馈所需的语义数据集，从而提高用户的使用体验。
对于上述任一方面，可选地，数据存储节点根据目标语义数据集涉及的至少一个业务领域，生成至少一个域文件夹，进而生成用于链接到各域文件夹和第二索引文件的第一索引文件，其中，每个业务领域对应一个域文件夹，不同业务领域对应不同域文件夹，每个域文件夹中记录有文件链接，文件链接为分布式存储系统中属于该域文件夹所对应业务领域且为最新版本的各本体文件的链接，本体文件包括目标语义数据集中该业务领域的数据。
由于知识创建者可能同时创建或更新不同业务领域的语义数据集，在对知识创建者输入的目标语义数据集进行存储时，根据目标语义数据集所涉及的业务领域创建域文件夹，使得目标语义数据集涉及的不同业务领域对应不同的域文件夹，之后根据目标语义数据集中各本体文件所属的业务领域，将目标语义数据集中本体文件的链接记录到相应业务领域对应的域文件夹中，使得每一个域文件夹链接到其所对应业务领域的本体文件的最新版本，从而可以分领域、按版本对语义数据集进行存储，便于对语义数据集进行管理和版本追溯。
对于上述任一方面，可选地，数据存储节点计算每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件的唯一标识，并将每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件的唯一标识发送给知识管理节点，知识管理节点在确定目标语义数据集可信后，将每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件的唯一标识存储到区块链上。
将第一索引文件的唯一标识存储到区块链上，当用户需要查询目标语义数据集时，通过触发智能合约可以获取到第一索引文件的唯一标识，根据第一索引文件的唯一标识，从分布式存储系统上获取第一索引文件，进而可以通过第一索引文件获取到目标语义数据集，以方便从分布式存储系统上查找用户所需的语义数据集。将第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识存储到区块链中，可以避免第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识被篡改，进而可以利用区块链中存储的第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识，验证相应的本体文件是否被篡改，保证语义数据集的安全性。
对于上述任一方面，可选地，知识管理节点在将第一索引文件的唯一标识存储到区块链上时，利用知识创建者设定的密钥对第一索引文件的唯一标识进行加密，并将经过加密的第一索引文件的唯一标识存储到区块链上。
通过对第一索引文件的唯一标识进行加密，仅有通过验证的用户才能够获取到第一索引文件的唯一标识，进而根据第一索引文件的唯一标识来访问分布式存储系统上的数据，从而可以保证分布式存储系统上数据的安全性。
对于上述任一方面，可选地，知识管理系统还包括至少一个应用节点，至少一个应用节点部署在区块链上。应用节点获取用户查询目标语义数据集的知识查询请求，并将获取到的知识查询请求发送给知识管理节点；在知识管理节点本地已存储目标语义数据集时，知识管理节点根据所述知识查询请求，通过智能合约校验目标语义数据集可信后，将目标语义数据集发送给应用节点，以供应用节点对目标语义数据集进行可视化。
将用户访问过的语义数据集缓存在知识管理节点的本地，当用户需要访问一个语义数据集时，首先检查知识管理节点的本地是否缓存有用户所要访问的语义数据集，如果知识管理节点的本地缓存有用户所要访问的语义数据集，则无需从分布式存储系统中获取数据，提高向用户反馈所需语义数据集的速度，从而可以提高用户的使用体验。
对于上述任一方面，可选地，在知识管理节点本地未存储目标语义数据集时，知识管理节点根据知识查询请求从区块链上获取第一索引文件的唯一标识，并将第一索引文件的唯一标识发送给数据存储节点；数据存储节点根据接收到的第一索引文件的唯一标识，从分布式存储系统中获取第一索引文件，根据获取到的第一索引文件，从分布式存储系统中获取第一索引文件链接到的各域文件夹，根据获取到的各域文件夹中记录的各文件链接，从分布式存储系统中获取目标语义数据集，并将获取到的目标语义数据集发送给知识管理节点；知识管理节点在通过智能合约校验获取到的目标语义数据集可信后，将目标语义数据集存储到知识管理节点本地，并将目标语义数据集发送给应用节点，以供应用节点对目标语义数据集进行可视化。
在用户需要访问目标语义数据集时，如果知识管理节点本地未存储目标语义数据集，则从分布式存储系统获取目标语义数据集，在验证目标语义数据集可信后，将目标语义数据集存储在知识管理节点的本地，并将目标语义数据集发送给应用节点进行可视化。知识管理节点在本地存储用户访问过的语义数据集，可以减少访问分布式存储系统的次数，可以降低知识管理系统的计算压力，还能够更快地向用户返回所需的语义数据集。
对于上述任一方面，可选地，所述区块链为私有链。由于索引文件的唯一标识存储到区块链上，而且智能合约也存储在区块链上，而为了安全考虑，索引文件和智能合约并不希望 所有用户都能够访问，通过私有链便可以限制对区块链上所存储内容的访问权限。
对于上述任一方面，可选地，分布式存储系统可以是星际文件系统，星际文件系统可以更好的与区块链进行结合实现数据的安全存储，从而进一步保证对所管理知识的安全性。
附图说明
图1是本申请实施例提供的一种知识管理系统的示意图；
图2是本申请实施例提供的另一种知识管理系统的示意图；
图3是本申请实施例提供的一种分布式存储系统中数据存储结构形式的示意图；
图4是本申请实施例提供的又一种知识管理系统的示意图；
图5是本申请实施例提供的一种语义数据集存储方式的示意图；
图6是本申请实施例提供的一种知识管理方法的流程图；
图7是本申请实施例提供的一种知识管理装置的示意图；
图8是本申请实施例提供的另一种知识管理装置的示意图；
图9是本申请实施例停工的一种电子设备的示意图。
附图标记列表：
10：知识管理系统         20：区块链               30：分布式存储系统
60：知识管理方法         11：知识管理节点         12：数据存储节点
13：应用节点             14：知识审批节点         21：智能合约
40：知识创建者           700：知识管理装置        701：数据获取模块
702：标识获取模块        703：存储模块            704：收发模块
900：电子设备            902：处理器              904：通信接口
906：存储器              908：通信总线            910：程序
601：知识管理节点获取知识创建者输入的目标语义数据集
602：知识管理节点在目标语义数据集被存储后，获取目标语义数据集的唯一标识
603：在确定目标语义数据集可信后，将目标语义数据集的唯一标识存储在区块链上
具体实施方式
如前所述，目前通常采用集中式管理系统对知识进行管理，语义数据集、数据模型和知识图谱均集中存储在服务器上，参与语义数据集准备工作的多方均可以创建新的语义数据集以及对原有的语义数据集进行更新，而且还可以对语义数据集的创建记录和更新记录进行修 改。由于各参与方均可以自由创建和更新语义数据集，而且语义数据集的创建记录和更新记录也可以被修改，因此存在通过篡改语义数据集的创建记录和更新记录而恶意创建或更新语义数据集的风险，因此无法保证语义数据集的安全性。
本申请实施例中，知识创建者以创建或更新的方式上传语义数据集后，对语义数据集进行存储，并获取语义数据集的唯一标识，在根据区块链上存储的智能合约确定语义数据集可信后，将语义数据集的唯一标识存储到区块链上。通过智能合约验证语义数据集可信，保证了知识创建者所上传语义数据集的内容合法性，另外语义数据集的唯一标识被存储到区块链上，基于区块链的特性可以保证语义数据集的唯一标识无法被篡改，进而可以根据区块链上存储的语义数据集的唯一标识验证所存储的语义数据集是否被篡改，从而可以保证语义数据集的安全性。
下面结合附图对本申请实施例提供的知识管理系统、方法、装置和电子设备进行详细说明。
如图1所示，本申请实施例提供了一种知识管理系统10，包括：至少一个知识管理节点11，各知识管理节点11部署在区块链20上；
知识管理节点11，用于获取知识创建者输入的目标语义数据集，在目标语义数据集被存储后，获取目标语义数据集的唯一标识，在根据区块链20上存储的智能合约确定目标语义数据集可信后，将目标语义数据集的唯一标识存储在区块链20上。
在本申请实施例中，对于知识创建者上传的目标语义数据集，在通过智能合约验证目标语义数据集可信后，将目标语义数据集的唯一标识存储到区块链上，即对应唯一标识被存储在区块链上的语义数据集是经审批可信的，从源头上保证了目标语义数据集的正确性和安全性。另外，当用户查询语义数据集时，在获取到用户所需的语义数据集后，通过区块链上获取该语义数据集的唯一标识，通过获取到的唯一标识对该语义数据进行验证，以确定获取到的语义数据集是否被篡改，在确定该语义数据集未被篡改后发送给用户，从而可以确保发送给用户的语义数据集是未被篡改过的，从而可以进一步保证语义数据集的安全性。
在本申请实施例中，知识管理节点11位于区块链20内，区块链20可以是点对点(P2P)网络，区块链20上存储有智能合约和语义数据集的唯一标识。
需要说明的是，语义数据集的唯一标识用于标识语义数据集，不同的语义数据集对应不同的唯一标识，而且根据唯一标识可以确定语义数据集是否被篡改。比如，语义数据集的唯一标识可以是语义数据集的哈希值、MD5值等，即通过哈希算法、MD5算法等加密算法对语义数据集进行计算，获得可以唯一的标识语义数据集的哈希值、MD5值等，作为语义数据集的唯一标识。
在一种可能的实现方式中，在图1所示知识管理系统10的基础上，在接收到知识创建者输入的目标语义数据集后，可以将目标语义数据集存储在分布式存储系统上，并生成用于对目标语义数据集中的本体文件进行索引的索引文件，同时将针对本次接收到的语义数据集生成的索引文件与针对上一次接收到的语义数据集生成的索引文件相链接，以便于对语义数据集进行版本管理。图2是本申请实施例提供的另一种知识管理系统10的示意图，如图2所示，该知识管理系统10还包括数据存储节点12，数据存储节点12部署在分布式存储系统30上；
知识管理节点11用于将目标语义数据集发送给数据存储节点12；
数据存储节点12用于将接收到的目标语义数据集存储早分布式存储系统30上，生成用于对目标语义数据集中的本体文件进行索引的第一索引文件，并将第一索引文件存储在分布式存储系统30上，其中，第一索引文件链接到第二索引文件，第二索引文件用于对上一次来自知识管理节点11的语义数据集中的本体文件进行索引；
数据存储节点12还用于计算第一索引文件的唯一标识，并将第一索引文件的唯一标识发送给知识管理节点11；
知识管理节点11还用于在确定目标语义数据集可信后，将第一索引文件的唯一标识存储到区块链20上。
在本申请实施例中，知识管理节点11在接收到知识创建者输入的目标语义数据集后，将目标语义数据集发送给数据存储节点12，数据存储节点12将目标语义数据集存储在分布式存储系统30上。将目标语义数据集存储在分布式存储系统30上，基于分布式存储系统30的特性，能够增大目标语义数据集被篡改的难度，同时能够降低目标语义数据集由于意外原因而丢失的风险，以提高目标语义数据集的安全性。
数据存储节点12将目标语义数据集存储到分布式存储系统30上之后，数据存储节点12生成用于对目标语义数据集中的本体文件进行索引的第一索引文件，同时第一索引文件还链接到第二索引文件，第二索引文件用于对上一次知识管理节点11接收到的语义数据集中的本体文件进行索引。通过第一索引文件可以从分布式存储系统30上获取目标语义数据集包括的本体文件，由于目标语义数据集是知识创建者新创建的语义数据集或者是对原有语义数据集进行版本更新后的语义数据集，因此将目标语义数据集存储到分布式存储系统30上之后，目标语义数据集所包括的本体文件均是最新版本，即通过第一索引文件可以从分布式存储系统30获取最新版本的本体文件。由于第一索引文件还链接到第二索引文件，第二索引文件用于对上一次存储到分布式存储系统30的语义数据集中的本体文件进行索引，而第二索引文件还链接到针对更早之前存储到分布式存储系统30上的语义数据集生成的索引文件，即可以通过第一索引文件获得第二索引文件，进而通过第二索引文件可以从分布式存储系统30上获取旧 版本的本体文件，从而通过索引文件可以方便地对语义数据集进行版本追溯和版本管理。
数据存储节点12在生成第一索引文件后，将第一索引文件存储在分布式存储系统30上，然后计算第一索引文件的唯一标识，并将第一索引文件的唯一标识发送给知识管理节点11，知识管理节点11在验证目标语义数据集可信后，将第一索引文件的唯一标识存储在区块链20上。知识管理节点11在确定目标语义数据集可信后，将第一索引文件的唯一标识存储到区块链20上，使得第一索引文件的唯一标识不会被篡改，从而可以根据区块链上的第一索引文件的唯一标识验证分布式存储系统30中的目标语义数据集的存储路径是否被篡改，还能够通过第一索引文件的唯一标识从分布式存储系统30上查找第一索引文件。
需要说明的是，与语义数据集的唯一标识相对应，第一索引文件的唯一标识可以是第一索引文件的哈希值、MD5值等。
还需要说明的是，将目标语义数据集存储到分布式存储系统上，仅是本申请实施例的一种可选的实现方式，在另一些实施例中，可以将目标语义数据集存储在知识管理节点11的本地存储，或者可以将目标语义数据集存储在区块链上，抑或是将目标语义数据集存储到云端服务器，对于目标语义数据集的存储位置，本申请实施例不进行限定。
在一种可能的实现方式中，数据存储节点12在接收到目标语义数据集后，根据目标语义数据集涉及的至少一个业务领域，生成至少一个域文件夹，进而生成用于链接到各域文件夹和第二索引文件的第一索引文件，其中，每个业务领域对应一个域文件夹，不同业务领域对应不同域文件夹，每个域文件夹中记录有文件链接，文件链接为分布式存储系统30中属于该域文件夹所对应业务领域且为最新版本的各本体文件的链接，本体文件包括目标语义数据集中该业务领域的数据。
在本申请实施例中，由于知识创建者可能同时创建或更新不同业务领域的语义数据集，在对知识创建者输入的目标语义数据集进行存储时，根据目标语义数据集所涉及的业务领域创建域文件夹，使得目标语义数据集涉及的不同业务领域对应不同的域文件夹，之后根据目标语义数据集中各本体文件所属的业务领域，将目标语义数据集中本体文件的链接记录到相应业务领域对应的域文件夹中，使得每一个域文件夹链接到其所对应业务领域的本体文件的最新版本，从而可以分领域、按版本对语义数据集进行存储，便于对语义数据集进行管理和版本追溯。
在本申请实施例中，由于目标语义数据集是知识创建者新创建的语义数据集或者是对原有语义数据集进行版本更新后的语义数据集，因此将目标语义数据集存储到分布式存储系统30上之后，目标语义数据集所包括的本体文件均是最新版本，从而目标语义数据集中本体文件的链接均会被记录在相应的域文件夹中。另外，由于知识管理系统10可以对不同业务领域 的知识进行管理，为了更加方便的对属于不同业务领域的知识进行管理，可以按照业务领域生成域文件夹，以便于通过所生成的域文件夹获取相应本体文件。
在本申请实施例中，每一次知识创建者创建新的语义数据集或更新原有的语义数据集时，数据存储节点12在将语义数据集存储到分布式存储系统30上之后，会根据语义数据集所涉及的业务领域生成一个或多个域文件夹，用于记录新创建语义数据集或更新语义数据集版本后最新版本的各本体文件的链接。为了进行区别，将数据存储节点12最新创建的域文件夹称为第一域文件夹，将数据存储节点12此前创建的域文件夹称为第二域文件夹。由于知识创建者可能不会同时对分布式存储系统30上存储的所有语义数据集进行更新，因此第一域文件夹和第二域文件夹可能记录有相同本体文件的链接。比如，最近生成的第二域文件夹中记录有本体文件1-V1、本体文件2-V2和本体文件3-V1的链接，而知识创建者仅将本体文件1-V1更新为本体文件1-V2，则在知识创建者将本体文件1-V1更新为本体文件1-V2后，所生成的第一域文件夹中记录有本体文件1-V2、本体文件2-V2和本体文件3-V1的链接。
在本申请实施例中，在针对目标语义数据集生成各域文件夹后，数据存储节点12还会生成用于链接到各域文件夹的第一索引文件。由于第一索引文件链接到了各域文件夹，获取到第一索引文件便可以获取到各域文件夹，进而可以根据域文件夹中记录的文件链接获取本体文件。为了便于区别，将链接到各第一域文件夹的索引文件称为第一索引文件，而将链接到各第二域文件夹的索引文件称为第二索引文件。
在本申请实施例中，知识创建者每一次对语义数据集进行更新(包括创建新的语义数据集或对原有语义数据集进行更新)后，均会生成一个索引文件，该索引文件所链接到的各域文件夹中记录有本次语义数据集更新后相应业务领域各本体文件的最新版本，从而用户根据该索引文件可以获取到本次语义数据集更新后各本体文件的最新版本，从而利用索引文件可以方便地对语义数据集进行版本管理。另外，最新生成的索引文件中会记录指向此前所生成索引文件的索引文件链接，通过索引文件链接可以获取到各个旧版本的本体文件，从而可以通过索引文件方便地对语义数据集进行版本追溯。除此之外，对于每一个业务领域，针对该业务领域最新生成的域文件夹，可以链接到此前针对该业务领域生成的域文件夹，即记录有新版本的本体文件的域文件夹，可以链接到记录有旧版本的本体文件的域文件夹，从而可以通过域文件夹方便地对语义数据集进行版本追溯。
第一索引文件链接到第二索引文件，具体为第一索引文件中记录有至少一个指向第二索引文件的索引文件链接，当第一索引文件中记录有一个索引文件链接时，该索引文件链接可以指向最近生成的第二域文件夹(即为根据上一次存储到分布式存储系统30中的语义数据集而形成的域文件夹)，当第一索引文件中记录有多个索引文件链接时，这些索引文件链接可以 依次指向最近生成的N(N＞2)个第二域文件夹。由于每个域文件夹均记录了创建语义数据集后或更新语义数据集后相应业务领域最新版本的各本体文件，而所生成的每一个域文件夹均链接到一个索引文件，因此索引文件可以作为分布式存储系统30上语义数据集的创建记录和更新记录，数据存储节点12会将各索引文件存储在分布式存储系统30上，而知识管理节点11会将各索引文件的唯一标识作为交易记录存储到区块链20上，从而保证了语义数据集的创建记录和更新记录无法被篡改，进一步提高了语义数据集的安全性。
在一种可能的实现方式中，数据存储节点12还用于计算每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件(即域文件夹中所记录文件链接对应的本体文件)的唯一标识，并将每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件的唯一标识发送给知识管理节点11。在智能合约确定目标语义数据集可信后，知识管理节点11将接收到的每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件的唯一标识存储到区块链20上。
在本申请实施例中，在智能合约确定目标语义数据集可信后，知识管理节点11会将第一索引文件的唯一标识存储到区块链20上，并会将第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识存储到区块链20中。将第一索引文件的唯一标识存储到区块链上，当用户需要查询目标语义数据集时，通过触发智能合约可以获取到第一索引文件的唯一标识，将第一索引文件的唯一标识发送给数据存储节点12后，数据存储节点12可以根据第一索引文件的唯一标识，从分布式存储系统30上获取第一索引文件，进而可以通过第一索引文件获取到目标语义数据集。将第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识存储到区块链20中，可以避免第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识被篡改，进而可以利用区块链20中存储的第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识，验证相应的本体文件是否被篡改。
需要说明的是，与语义数据集的唯一标识相对应，域文件夹的唯一标识可以是域文件夹的哈希值、MD5值等，本体文件的唯一标识可以是本体文件的希值、MD5值等。
在本申请实施例中，图3示意性的给出了分布式存储系统30中数据存储的结构形式，分布式存储系统30中存储有本体文件OF1-V1、本体文件OF2-V1、本体文件OF1-V2和本体文件OF3、域文件夹DF-V1、域文件夹DF-V2以及索引文件C1至索引文件Cn，其中，本体文件OF1-V2是本体文件OF1-V1的新版本。
域文件夹DF-V1链接到本体文件OF1-V1、本体文件OF2-V1和本体文件OF3(即域文件夹DF-V1中记录有本体文件OF1-V1、本体文件OF2-V1和本体文件OF3的文件链接)，域文件夹DF-V2链接到本体文件OF1-V2、本体文件OF2-V1和本体文件OF3(即域文件夹DF-V2 中记录有本体文件OF1-V2、本体文件OF2-V1和本体文件OF3的文件链接)。
索引文件C1链接到域文件夹DF-V1，索引文件C2链接到域文件夹DF-V2，索引文件Cn链接到域文件夹DF-Vn(图中未示出域文件夹DF-Vn)，而且索引文件C2中记录有指向索引文件C1的索引文件链接P1，索引文件Cn中记录有指向索引文件Cn-1的索引文件链接Pn-1。
索引文件C1至索引文件Cn依次被存储到区块链20中的各区块1至区块n中，其中区块n位于区块链20的尾部。
当知识创建者对本体文件OF1进行版本更新获得本体文件OF1-V2后，由于本体文件OF2和本体文件OF3未被更新版本，所以本体文件OF2-V1和本体文件OF3仍为各自的最新版本，所以数据存储节点12所生成的域文件夹DF-V2链接到本体文件OF1-V2、本体文件OF2-V1和本体文件OF3。
需要说明的是，图3仅作为分布式存储系统30中存储结构的一种示例，实际分布式存储系统30中包括更多的本体文件、域文件夹和索引文件，而且可能存在同一索引文件链接到多个域文件夹的情况。
在一种可能的实现方式中，知识管理节点11在将第一索引文件的唯一标识存储到区块链20上时，首先利用知识创建者设定的密钥对第一索引文件的唯一标识进行加密，然后将经过加密的唯一标识存储到区块链20上。
在本申请实施例中，知识创建者在上传目标语义数据集之前会设定密钥，知识管理节点11利用知识创建者设定的密钥对第一索引文件的唯一标识进行加密，进而将加密后的唯一标识存储到区块链20上。由于第一索引文件的唯一标识用于从分布式存储系统30上查找第一索引文件，通过对第一索引文件的唯一标识进行加密，使得仅有通过验证的用户才能够获取到第一索引文件的唯一标识，进而根据第一索引文件的唯一标识来访问分布式存储系统30上的数据，从而可以保证分布式存储系统30上数据的安全性。
在一种可能的实现方式中，在图2所示知识管理系统10的基础上，如图4所示，该知识管理系统10还可以包括至少一个应用节点13，其中，各应用节点13部署在区块链20上；
应用节点13用于获取用户查询目标语义数据集的知识查询请求，并将知识查询请求发送给知识管理节点11；
知识管理节点11在接收到知识查询请求后，如果知识管理节点11的本地存储有目标语义数据集，则在通过智能合约校验本地存储的目标语义数据集可信后，将目标语义数据集发送给应用节点13，以供应用节点13对目标语义数据集进行可视化。
在本申请实施例中，当应用节点13将知识查询请求发送给知识管理节点11后，知识管 理节点11首先根据知识查询请求检查本地是否已经存储用户所要访问的目标语义数据集，如果知识管理节点11本地已经存储了目标语义数据集，知识管理节点11通过智能合约验证本地存储的目标语义数据集可信后，将本地存储的目标语义数据集发送给应用节点13，以供该应用节点13对目标语义数据集进行可视化。在知识管理节点11的本地存储有用户所需的语义数据集时，校验本地存储的语义数据集可信后，将本地存储的语义数据集发送给用户，从而无需频繁访问分布式存储系统30获取语义数据集，一方面可以降低知识管理系统10所需的网络带宽，另一方面还能够提高向用户反馈语义数据集的速率，进而提高用户的使用体验。
在本申请实施例中，知识管理节点11在通过智能合约对目标语义数据集进行校验时，智能合约可以从区块链20上获取目标语义数据集的唯一标识，智能合约按照预先设定的加密算法对目标语义数据集进行计算，获得目标语义数据集的唯一标识，然后将计算获得的唯一标识与从区块链20上获取到的唯一标识进行比对，如果两者相同则确定目标语义数据集可信，如果两者不同则确定目标语义数据集不可信。
在一种可能的实现方式中，如图4所示，知识管理节点11在接收到知识查询请求后，如果知识管理节点11的本地未存储用户所要访问的目标语义数据集，则根据知识查询请求从区块链20上获取第一索引文件的唯一标识，并将第一索引文件的唯一标识发送给数据存储节点12。数据存储节点12根据接收到的第一索引文件的唯一标识，从分布式存储系统30中获取第一索引文件，然后根据获取到的第一索引文件从分布式存储系统30中获取第一索引文件链接到的各域文件夹，然后根据获取到的各域文件夹中记录的各文件链接，从分布式存储系统30中获取目标语义数据集，并将获取到的目标语义数据集发送给知识管理节点11。知识管理节点11通过智能合约验证接收到的目标语义数据集是否可信，在确定目标语义数据集可信后，将目标语义数据集存储到知识管理节点11的本地，并将目标语义数据集发送给应用节点13，以供应用节点13对目标语义数据集进行可视化。
在本申请实施例中，当应用节点13将知识查询请求发送给知识管理节点11后，知识管理节点11首先根据知识查询请求检查本地是否已经存储用户所要访问的目标语义数据集，如果知识管理节点11本地未存储目标语义数据集，则通过应用节点13从分布式存储系统30获取目标语义数据集，在验证目标语义数据集可信后，将目标语义数据集存储在知识管理节点11的本地，并将目标语义数据集发送给应用节点13。知识管理节点11在本地存储用户访问过的语义数据集，可以减少访问分布式存储系统30的次数，可以降低知识管理系统10的计算压力，还能够更快地向用户返回所需的语义数据集。
在一种可能的实现方式中，在图4所示知识管理系统10的基础上，该知识管理系统10还包括：协调节点和至少一个知识审批节点，协调节点和各知识审批节点均部署在区块链20 上；
协调节点用于为用户分配账号，不同的账号具有不同的权限，用户通过协调节点分配的账号登陆设备后，根据所登陆账号的权限不同，设备作为知识管理系统10中的知识管理节点11、应用节点13或知识审批节点；
协调节点还分别为每个知识管理节点11、每个应用节点13和每个知识审批节点分配不同的用户标识，以使知识管理节点11、应用节点13和知识审批节点之间可以根据彼此的用户标识进行信息交互。
在本申请实施例中，由于知识管理节点11和协调节点在整个知识管理系统10中处于核心地位，为了保证知识管理节点11和协调节点的安全，知识管理节点11和协调节点通过具有最高投票权限的智能合约选择，而且投票的结果可以按照配置的时间间隔进行更新。
在本申请实施例中，协调节点可以分别为每个知识管理节点11、每个应用节点13和每个知识审批节点分配不同的用户标识，以通过用户标识确定各个节点所属的组，以便于知识管理节点11能够根据用户标识分别与每个应用节点13和每个知识审批节点进行信息交互，保证整个知识管理系统10的有序、安全运行。
在本申请实施例中，如图5所示，知识管理节点11在获取知识创建者40上传的目标语义数据集后，一方面将目标语义数据集发送给数据存储节点12，由数据存储节点12将目标语义数据集存储到分布式存储系统30上，另一方面知识管理节点11会生成用于请求知识审批者对目标语义数据集进行审批的知识审批请求，并根据智能合约21将所生成的知识审批请求发送给至少一个知识审批节点14，接收到知识审批请求的知识审批节点14将知识审批者给出的审批结果发送给智能合约21后，智能合约21根据所接收到的各审批结果确定目标语义数据集是否可信，知识管理节点11在获得智能合约21确定目标语义数据集可信的结果后，将目标语义数据集的唯一标识存储到区块链20上。
如果智能合约根据所接收到的各审批结果确定目标语义数据不可信，知识管理节点11不会将目标语义数据集的唯一标识存储到区块链20上，以此来说明目标语义数据集是不可信的。更进一步地，在确定目标语义数据集不可信后，知识管理节点11可以根据智能合约将目标语义数据集从分布式存储系统30上删除，也可以在分布式存储系统30中保留目标语义数据集，但给目标语义数据集设置不可信的标记，以在用户查询目标语义数据集时给出不可信的提示。
在本申请实施例中，知识审批者是所属领域的专家，根据智能合约将知识审批请求发送给一个或多个知识审批节点，由相应的知识审批者对目标语义数据集进行审批并给出相应的审批结果，最终由智能合约根据各知识审批节点返回的审批结果确定目标语义数据集是否可信。智能合约可以根据各审批结果中审批通过的占比确定目标语义数据集是否可信，比如当 有大于80％的审批结果为审批通过时，确定目标语义数据集可信。
在本申请实施例中，知识管理节点11获取到的目标语义数据集可以是知识创建者创建的新的语义数据集，也可以是知识创建者针对已存储语义数据集的更新版本，从而该知识管理系统10既可以安全的对新创建的语义数据集进行管理，也可以安全的对知识创建者更新的不同版本的语义数据集进行管理。
另外，知识管理节点11还可以根据目标语义数据集生成目标知识图谱，并通过数据存储节点12将目标语义数据集和所生成的目标知识图谱一同存储到分布式存储系统30中。相应的，知识管理节点11还会生成用于请求对目标知识图谱进行审批的知识审批请求，并根据智能合约将该知识审批请求发送给至少一个知识审批节点，知识审批节点则可以将审批者对目标知识图谱进行审批的审批结果发送给智能合约，在智能合约根据审批结果确定目标知识图谱可信后，知识管理节点11将目标知识图谱的唯一标识存储到区块链20上。由此可见，本申请实施例提供的知识管理系统10不仅可以对语义数据集进行管理，以保证语义数据集的安全性，同时还能够对根据语义数据集所生成的知识图谱进行管理，以保证知识图谱的安全性。
需要说明的是，本申请实施例所提供知识管理系统10所包括的各个部分，可以单独针对语义数据集进行处理，也可以同时针对语义数据集和根据语义数据集生成的知识图谱进行处理。而且，在前述各实施例中，根据语义数据集所生成的知识图谱与语义数据集可以作为并列的知识管理对象，知识管理系统10所包括的各个部分可以采用对语义数据集进行管理的相同方式对知识图谱进行管理。
在一种可能的实现方式中，在上述各个实施例所提供知识管理系统10的基础上，区块链20可以是私有链。由于索引文件的唯一标识存储到区块链20上，而且智能合约也存储在区块链20上，而为了安全考虑，索引文件和智能合约并不希望所有用户都能够访问，通过私有链便可以限制对区块链20上所存储内容的访问权限，更适合本申请所提供知识管理系统10的应用场景。
在一种可能的实现方式中，在上述各实施例所提供给知识管理系统10的基础上，分布式存储系统30可以是星际文件系统(InterPlanetary File System，IPFS)，星际文件系统可以更好的与区块链20进行结合实现数据的安全存储，从而进一步保证本申请所提供知识管理系统10对于所管理知识的安全性。
下面介绍本申请实施例提供的知识管理方法，该方法可以基于前述的知识管理系统10来实现。如无特别声明，下述知识管理方法中涉及的分布式存储系统可为前述的分布式存储系统30，下述知识管理方法中涉及的知识管理节点可为前述的知识管理节点11，下述知识管理 方法中涉及的数据存储节点可为前述的数据存储节点12，下述知识管理方法中涉及的应用节点可以前述的应用节点13，下述知识管理方法中涉及的区块链可为前述的区块链20。
如图6所示，本申请一个实施例提供了一种知识管理方法60，该方法可以包括如下步骤：
601、知识管理节点获取知识创建者输入的目标语义数据集；
602、知识管理节点在目标语义数据集被存储后，获取目标语义数据集的唯一标识；
603、知识管理节点在根据区块链上存储的智能合约，确定目标语义数据集可信后，将目标语义数据集的唯一标识存储在区块链上。
在本申请实施例中，对于知识创建者上传的目标语义数据集，在通过智能合约验证目标语义数据集可信后，将目标语义数据集的唯一标识存储到区块链上，即对应唯一标识被存储在区块链上的语义数据集是经审批可信的，从源头上保证了目标语义数据集的正确性和安全性。另外，当用户查询语义数据集时，在获取到用户所需的语义数据集后，通过区块链上获取该语义数据集的唯一标识，通过获取到的唯一标识对该语义数据进行验证，以确定获取到的语义数据集是否被篡改，在确定该语义数据集未被篡改后发送给用户，从而可以确保发送给用户的语义数据集是未被篡改过的，从而可以进一步保证语义数据集的安全性。
在一种可能的实现方式中，在获取到目标语义数据集后，可以通过数据存储节点将目标语义数据集存储到分布式存储系统上。对目标语义数据集进行存储，具体可以通过如下步骤实现：
S1、知识管理节点将目标语义数据集发送给数据存储节点；
S2、数据存储节点将目标语义数据集存储在分布式存储系统上；
S3、数据存储节点生成用于对目标语义数据集中的本体文件进行索引的第一索引文件，其中，第一索引文件链接到第二索引文件，第二索引文件用于对上一次来自知识管理节点的语义数据集中的本体文件进行索引；
S4、数据存储节点对第一索引文件进行存储；
S5、数据存储节点计算第一索引文件的唯一标识，并将第一索引文件的唯一标识发送给知识管理节点；
S6、知识管理节点在确定目标语义数据集可信后，将第一索引文件的唯一标识存储到区块链上。
在本申请实施例中，在接收到知识创建者输入的目标语义数据集后，将目标语义数据集存储在分布式存储系统上，基于分布式存储系统的特性，能够增大目标语义数据集被篡改的难度，同时能够降低目标语义数据集由于意外原因而丢失的风险，以提高目标语义数据集的安全性。
在本申请实施例中，将目标语义数据集存储到分布式存储系统上之后，生成用于对目标语义数据集进行索引的第一索引文件，并将第一索引文件存储在分布式存储系统上，然后计算第一索引文件的唯一标识，在验证目标语义数据集可信后，将第一索引文件的唯一标识存储在区块链上，使得第一索引文件的唯一标识不会被篡改，从而可以根据区块链上的第一索引文件的唯一标识验证分布式存储系统中的目标语义数据集的存储路径是否被篡改，还能够通过第一索引文件的唯一标识从分布式存储系统上查找第一索引文件，从而更加方便从分布式存储系统上查找用户每次创建或更新后的语义数据集，用户在查阅语义数据集时，可以更加快速的向用户反馈所需的语义数据集，从而提高用户的使用体验。
在一种可能的实现方式中，在生成第一索引文件时，可以根据目标语义数据集涉及的至少一个业务领域，生成至少一个域文件夹，进而生成用于链接到各域文件夹和第二索引文件的第一索引文件，其中，每个业务领域对应一个域文件夹，不同业务领域对应不同域文件夹，每个域文件夹中记录有文件链接，文件链接为分布式存储系统中属于该域文件夹所对应业务领域且为最新版本的各本体文件的链接，本体文件包括目标语义数据集中该业务领域的数据。
在本申请实施例中，由于知识创建者可能同时创建或更新不同业务领域的语义数据集，在对知识创建者输入的目标语义数据集进行存储时，根据目标语义数据集所涉及的业务领域创建域文件夹，使得目标语义数据集涉及的不同业务领域对应不同的域文件夹，之后根据目标语义数据集中各本体文件所属的业务领域，将目标语义数据集中本体文件的链接记录到相应业务领域对应的域文件夹中，使得每一个域文件夹链接到其所对应业务领域的本体文件的最新版本，从而可以分领域、按版本对语义数据集进行存储，便于对语义数据集进行管理和版本追溯。
在一种可能的实现方式中，在图6所示知识管理方法60的基础上，在确定出第一索引文件链接到的各域文件夹后，可以计算每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件的唯一标识，在智能合约确定目标语义数据集可信后，将每个域文件夹的唯一标识和每个域文件夹中所记录的本体文件的唯一标识存储到区块链上。
在本申请实施例中，将第一索引文件的唯一标识存储到区块链上，当用户需要查询目标语义数据集时，通过触发智能合约可以获取到第一索引文件的唯一标识，根据第一索引文件的唯一标识，从分布式存储系统上获取第一索引文件，进而可以通过第一索引文件获取到目标语义数据集，以方便从分布式存储系统上查找用户所需的语义数据集。将第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识存储到区块链中，可以避免第一索引文件链接到的各域文件夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识被篡改，进而可以利用区块链中存储的第一索引文件链接到的各域文件 夹的唯一标识以及每个域文件夹所记录的各本体文件的唯一标识，验证相应的本体文件是否被篡改，保证语义数据集的安全性。
在一种可能的实现方式中，在将第一索引文件的唯一标识存储到区块链上时，首先利用知识创建者设定的密钥对第一索引文件的唯一标识进行加密，然后将经过加密的唯一标识存储到区块链上。
在本申请实施例中，由于第一索引文件的唯一标识用于从分布式存储系统上查找第一索引文件，通过对第一索引文件的唯一标识进行加密，仅有通过验证的用户才能够获取到第一索引文件的唯一标识，进而根据第一索引文件的唯一标识来访问分布式存储系统上的数据，从而可以保证分布式存储系统上数据的安全性。
在一种可能的实现方式中，在图6所示知识管理方法60的基础上，该知识管理方法60还包括：
应用节点获取用户查询目标语义数据集的知识查询请求；
应用节点将获取到的知识查询请求发送给知识管理节点；
在知识管理节点本地已存储目标语义数据集时，知识管理节点通过智能合约校验目标语义数据集可信后，将目标语义数据集发送给应用节点，以供应用节点对目标语义数据集进行可视化。
在本申请实施例中，将用户访问过的语义数据集缓存在知识管理节点的本地，当用户需要访问一个语义数据集时，首先检查知识管理节点的本地是否缓存有用户所要访问的语义数据集，如果知识管理节点的本地缓存有用户所要访问的语义数据集，则无需从分布式存储系统中获取数据，提高向用户反馈所需语义数据集的速度，从而可以提高用户的使用体验。
在一种可能的实现方式中，在图6所示知识管理方法60的基础上，该知识管理方法60还包括：
在知识管理节点本地未存储目标语义数据集时，知识管理节点根据知识查询请求从智能合约中获取第一索引文件的唯一标识；
知识管理节点将第一索引文件的唯一标识发送给数据存储节点；
数据存储节点根据接收到的第一索引文件的唯一标识，从分布式存储系统中获取第一索引文件；
数据存储节点根据获取到的第一索引文件，从分布式存储系统中获取第一索引文件链接到的各域文件夹；
数据存储节点根据获取到的各域文件夹中记录的各文件链接，从分布式存储系统中获取目标语义数据集，并将获取到的目标语义数据集发送给知识管理节点；
知识管理节点在通过智能合约校验获取到的目标语义数据集可信后，将目标语义数据集存储缓存到知识管理节点本地；
知识管理节点将目标语义数据集发送给应用节点，以供应用节点对目标语义数据集进行可视化。
在本申请实施例中，在用户需要访问目标语义数据集时，如果知识管理节点本地未存储目标语义数据集，则从分布式存储系统获取目标语义数据集，在验证目标语义数据集可信后，将目标语义数据集存储在知识管理节点的本地，并将目标语义数据集发送给应用节点进行可视化。知识管理节点在本地存储用户访问过的语义数据集(本体文件)，可以减少访问分布式存储系统的次数，可以降低知识管理系统的计算压力，还能够更快地向用户返回所需的语义数据集。
在一种可能的实现方式中，在图6所示知识管理方法60的基础上，区块链可以是私有链。由于索引文件的唯一标识存储到区块链上，而且智能合约也存储在区块链上，而为了安全考虑，索引文件和智能合约并不希望所有用户都能够访问，通过私有链便可以限制对区块链上所存储内容的访问权限。
在一种可能的实现方式中，在图6所示知识管理方法60的基础上，分布式存储系统可以是星际文件系统(InterPlanetary File System，IPFS)，星际文件系统可以更好的与区块链进行结合实现数据的安全存储，从而进一步保证对所管理知识的安全性。
需要说明的是，上述各实施例提供的知识管理方法基于前述各实施例提供的知识管理系统实现，该知识管理方法与知识管理系统基于同一发明构思，该知识管理方法的具体内容可以参见前述知识管理系统实施例中的描述，在此不再进行赘述。
图7是本申请实施例提供的一种知识管理装置700的示意图，该知识管理装置700包括用于实现上述方法实施例中知识管理节点11所执行操作的模块，具体地，如图7所示，该知识管理装置700包括：
数据获取模块701，用于获取知识创建者输入的目标语义数据集；
标识获取模块702，用于在数据获取模块701获取到的目标语义数据集被存储后，获取目标语义数据集的唯一标识；
存储模块703，用于在根据区块链上存储的智能合约，确定目标语义数据集可信后，将目标语义数据集的唯一标识存储到区块链上。
在本申请实施例中，数据获取模块701获取到目标语义数据集后，标识获取模块702获取目标语义数据集的唯一标识，在目标语义数据集经智能合约验证可信后，存储模块703将 目标语义数据集的唯一标识存储到区块链。由于区块链的不可篡改性，对应唯一标识被存储在区块链上的语义数据集是可信的，从源头上保证了目标语义数据集的正确性和安全性。另外，存储模块703将目标语义数据集的唯一标识存储到区块链上，当用户查询目标语义数据集时，可以通过区块链上存储的唯一标识对目标语义数据集进行验证，以确定目标语义数据集是否被篡改，在确定目标语义数据集未被篡改后，再将目标语义数据集发送给用户，从而可以确保发送给用户的目标语义数据集是未被篡改的，从而可以进一步保证语义数据集的安全性。
可选地，图8是本申请实施例提供的另一种知识管理装置700的示意图，如图8所示，该知识管理装置700还包括：
收发模块704，用于将目标语义数据集发送给数据存储节点，由数据存储节点将目标语义数据集存储到分布式存储系统上，并接收来自数据存储节点的第一索引文件的唯一标识，将第一索引文件的唯一标识发送给存储模块703，其中，第一索引文件用于对分布式存储系统中目标语义数据集包括的本体文件进行索引，第一索引文件链接到第二索引文件，第二索引文件用于在分布式存储系统中对数据获取模块701上一次获取到的语义数据集中的本体文件进行索引；
存储模块703，还用于在确定目标语义数据集可信后，将第一索引文件的唯一标识存储到区块链上。
可选地，如图8所示，收发模块704还用于接收来自数据存储节点的每个域文件夹的唯一标识，以及每个域文件夹中记录的本体文件的唯一标识，并将接收到的每个域文件夹的唯一标识和每个域文件夹中记录的本体文件的唯一标识发送给所述存储模块703，存储模块703还用于在确定目标语义数据集可信后，将每个域文件夹的唯一标识和每个域文件夹中记录的本体文件的唯一标识存储到区块链上。
可选地，如图8所示，存储模块703还用于在将第一索引文件的唯一标识存储到区块链上时，利用知识创建者设定的密钥对第一索引文件的唯一标识进行加密，并将经过加密的第一索引文件的唯一标识存储到区块链上。
可选地，如图8所示，收发模块704还用于接收来自应用节点的知识查询请求，若知识管理节点的本地存储有知识查询请求所要访问的目标语义数据集，在通过智能合约校验知识管理节点本地存储的目标语义数据集可信后，将目标语义数据集发送给应用节点，以供应用节点对目标语义数据集进行可视化。
可选地，如图8所示，收发模块704还用于接收来自应用节点的知识查询请求，若知识管理节点的本地未存储知识查询请求所要访问的目标语义数据集，则根据知识查询请求从区 块链上获取第一索引文件的唯一标识，并将获取到的第一索引文件的唯一标识发送给数据存储节点。收发模块704还用于接收来自数据存储节点的目标语义数据集，在通过智能合约校验目标语义数据集可信后，将目标语义数据集存储到知识管理节点的本地，并将目标语义数据集发送给应用节点，以供应用节点对目标语义数据集进行可视化。
需要说明的是，上述知识管理装置实施例中各个模块之间的交互与前述知识管理系统实施例和知识管理方法实施例基于同一发明构思，具体内容可以参见前述知识管理系统实施例和知识管理方法实施例中的描述，在此不再赘述。
图9是本申请实施例提供的一种电子设备的示意图，本申请具体实施例并不对电子设备的具体实现做限定。参见图9，本申请实施例提供的电子设备900包括：处理器(processor)902、通信接口(Communications Interface)904、存储器(memory)906、以及通信总线908。其中：
处理器902、通信接口904、以及存储器906通过通信总线908完成相互间的通信。
通信接口904，用于与其它电子设备或服务器进行通信。
处理器902，用于执行程序910，具体可以执行前述任一知识管理方法实施例中知识管理节点11执行的操作。
具体地，程序910可以包括程序代码，该程序代码包括计算机操作指令。
处理器902可能是中央处理器CPU，或者是特定集成电路ASIC(Application Specific Integrated Circuit)，或者是被配置成实施本申请实施例的一个或多个集成电路。智能设备包括的一个或多个处理器，可以是同一类型的处理器，如一个或多个CPU；也可以是不同类型的处理器，如一个或多个CPU以及一个或多个ASIC。
存储器906，用于存放程序910。存储器906可能包含高速RAM存储器，也可能还包括非易失性存储器(non-volatile memory)，例如至少一个磁盘存储器。
程序910具体可以用于使得处理器902执行前述任一知识管理方法实施例中知识管理节点11执行的操作。
程序910中各步骤的具体实现可以参见前述任一知识管理方法实施例中的相应步骤和单元中对应的描述，在此不赘述。所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的设备和模块的具体工作过程，可以参考前述方法实施例中的对应过程描述，在此不再赘述。
通过本申请实施例的电子设备，对于知识创建者上传的目标语义数据集，在通过智能合约验证目标语义数据集可信后，将目标语义数据集的唯一标识存储到区块链上，即对应唯一标识被存储在区块链上的语义数据集是经审批可信的，从源头上保证了目标语义数据集的正 确性和安全性。另外，当用户查询语义数据集时，在获取到用户所需的语义数据集后，通过区块链上获取该语义数据集的唯一标识，通过获取到的唯一标识对该语义数据进行验证，以确定获取到的语义数据集是否被篡改，在确定该语义数据集未被篡改后发送给用户，从而可以确保发送给用户的语义数据集是未被篡改过的，从而可以进一步保证语义数据集的安全性。
本申请还提供了一种计算机可读存储介质，存储用于使一机器执行前述任一知识管理方法实施例中知识管理节点11执行的操作的指令。具体地，可以提供配有存储介质的系统或者装置，在该存储介质上存储着实现上述实施例中任一实施例的功能的软件程序代码，且使该系统或者装置的计算机(或CPU或MPU)读出并执行存储在存储介质中的程序代码。
在这种情况下，从存储介质读取的程序代码本身可实现上述实施例中任何一项实施例的功能，因此程序代码和存储程序代码的存储介质构成了本申请的一部分。
用于提供程序代码的存储介质实施例包括软盘、硬盘、磁光盘、光盘(如CD-ROM、CD-R、CD-RW、DVD-ROM、DVD-RAM、DVD-RW、DVD+RW)、磁带、非易失性存储卡和ROM。可选择地，可以由通信网络从服务器计算机上下载程序代码。
此外，应该清楚的是，不仅可以通过执行计算机所读出的程序代码，而且可以通过基于程序代码的指令使计算机上操作的操作系统等来完成部分或者全部的实际操作，从而实现上述实施例中任意一项实施例的功能。
此外，可以理解的是，将由存储介质读出的程序代码写到插入计算机内的扩展板中所设置的存储器中或者写到与计算机相连接的扩展模块中设置的存储器中，随后基于程序代码的指令使安装在扩展板或者扩展模块上的CPU等来执行部分和全部实际操作，从而实现上述实施例中任一实施例的功能。
需要说明的是，上述各流程和各系统结构图中不是所有的步骤和模块都是必须的，可以根据实际的需要忽略某些步骤或模块。各步骤的执行顺序不是固定的，可以根据需要进行调整。上述各实施例中描述的系统结构可以是物理结构，也可以是逻辑结构，即，有些模块可能由同一物理实体实现，或者，有些模块可能分由多个物理实体实现，或者，可以由多个独立设备中的某些部件共同实现。
以上各实施例中，硬件模块可以通过机械方式或电气方式实现。例如，一个硬件模块可以包括永久性专用的电路或逻辑(如专门的处理器，FPGA或ASIC)来完成相应操作。硬件模块还可以包括可编程逻辑或电路(如通用处理器或其它可编程处理器)，可以由软件进行临时的设置以完成相应操作。具体的实现方式(机械方式、或专用的永久性电路、或者临时设 置的电路)可以基于成本和时间上的考虑来确定。
上文通过附图和优选实施例对本申请进行了详细展示和说明，然而本申请不限于这些已揭示的实施例，基与上述多个实施例本领域技术人员可以知晓，可以组合上述不同实施例中的代码审核手段得到本申请更多的实施例，这些实施例也在本申请的保护范围之内。

Claims (19)

  1. 一种知识管理系统(10)，其特征在于，包括：至少一个知识管理节点(11)，所述至少一个知识管理节点部署在区块链(20)上；
    所述知识管理节点，用于执行如下处理：
    获取知识创建者输入的目标语义数据集；
    在所述目标语义数据集被存储后，获取所述目标语义数据集的唯一标识；
    在根据所述区块链(20)上存储的智能合约，确定所述目标语义数据集可信后，将所述目标语义数据集的唯一标识存储在所述区块链(20)上。
  2. 根据权利要求1所述的系统，其特征在于，所述知识管理系统(10)还包括：数据存储节点(12)；
    所述数据存储节点(12)部署在分布式存储系统上；
    所述知识管理节点(11)，还用于将所述目标语义数据集发送给所述数据存储节点(12)；
    所述数据存储节点(12)，用于将所述目标语义数据集存储在所述分布式存储系统上，生成用于对所述目标语义数据集中的本体文件进行索引的第一索引文件，对所述第一索引文件进行存储，计算所述第一索引文件的唯一标识，并将所述第一索引文件的唯一标识发送给所述知识管理节点(11)，其中，所述第一索引文件链接到第二索引文件，所述第二索引文件用于对上一次来自所述知识管理节点(11)的语义数据集中的本体文件进行索引；
    所述知识管理节点(11)，还用于在确定所述目标语义数据集可信后，将所述第一索引文件的唯一标识存储到所述区块链(20)上。
  3. 根据权利要求2所述的系统，其特征在于，所述数据存储节点(12)用于执行如下处理：
    根据所述目标语义数据集涉及的至少一个业务领域，生成至少一个所述域文件夹，其中，每个业务领域对应一个域文件夹，不同业务领域对应不同域文件夹，每个域文件夹中记录有文件链接，所述文件链接为所述分布式存储系统中属于该域文件夹所对应业务领域且为最新版本的各本体文件的链接，所述本体文件包括所述目标语义数据集中该业务领域的数据；
    生成用于链接到各所述域文件夹和所述第二索引文件的所述第一索引文件。
  4. 根据权利要求3所述的系统，其特征在于，
    所述数据存储节点(12)，还用于计算每个所述域文件夹的唯一标识和每个所述域文件夹中所记录的本体文件的唯一标识，并将每个所述域文件夹的唯一标识和每个所述域文件夹中所记录的本体文件的唯一标识发送给所述知识管理节点(11)；
    所述知识管理节点(11)，还用于在确定所述目标语义数据集可信后，将每个所述域文件夹的唯一标识和每个所述域文件夹中所记录的本体文件的唯一标识存储到所述区块链(20) 上。
  5. 根据权利要求2所述的系统，其特征在于，
    所述知识管理节点(11)，还用于在将所述第一索引文件的唯一标识存储到所述区块链(20)上时，利用所述知识创建者设定的密钥对所述第一索引文件的唯一标识进行加密，并将经过加密的所述第一索引文件的唯一标识存储到所述区块链(20)上。
  6. 根据权利要求3所述的系统，其特征在于，所述知识管理系统(10)还包括：至少一个应用节点(13)；
    所述至少一个应用节点(13)部署在所述区块链(20)上；
    所述应用节点(13)，用于获取用户查询所述目标语义数据集的知识查询请求，并将获取到的所述知识查询请求发送给所述知识管理节点(11)；
    所述知识管理节点(11)，还用于在所述知识管理节点(11)本地已存储所述目标语义数据集时，根据所述知识查询请求，通过所述智能合约校验所述目标语义数据集可信后，将所述目标语义数据集发送给所述应用节点(13)，以供所述应用节点(13)对所述目标语义数据集进行可视化。
  7. 根据权利要求6所述的系统，其特征在于，
    所述知识管理节点(11)，还用于在所述知识管理节点(11)本地未存储所述目标语义数据集时，根据所述知识查询请求从所述区块链(20)上获取所述第一索引文件的唯一标识，并将所述第一索引文件的唯一标识发送给所述数据存储节点(12)；
    所述数据存储节点(12)，还用于根据接收到的所述第一索引文件的唯一标识，从所述分布式存储系统中获取所述第一索引文件，根据获取到的所述第一索引文件，从所述分布式存储系统中获取所述第一索引文件链接到的各所述域文件夹，根据获取到的各所述域文件夹中记录的各文件链接，从所述分布式存储系统中获取所述目标语义数据集，并将获取到的所述目标语义数据集发送给所述知识管理节点(11)；
    所述知识管理节点(11)，还用于在通过所述智能合约校验获取到的所述目标语义数据集可信后，将所述目标语义数据集存储到所述知识管理节点(11)本地，并将所述目标语义数据集发送给所述应用节点(13)，以供所述应用节点(13)对所述目标语义数据集进行可视化。
  8. 根据权利要求1至7中任一所述的系统，其特征在于，
    所述区块链(20)为私有链，和/或，所述分布式存储系统为星际文件系统IPFS。
  9. 一种知识管理方法，其特征在于，由知识管理系统(10)执行，其中所述知识管理系统(10)包括部署在区块链(20)上的至少两个知识管理节点(11)，所述知识管理方法包括：
    所述知识管理节点(11)获取知识创建者输入的目标语义数据集；
    所述知识管理节点(11)在所述目标语义数据集被存储后，获取所述目标语义数据集的唯一标识；
    所述知识管理节点(11)在根据所述区块链(20)上存储的智能合约，确定所述目标语义数据集可信后，将所述目标语义数据集的唯一标识存储在所述区块链(20)上。
  10. 根据权利要求9所述的方法，其特征在于，所述知识管理系统(10)还包括数据存储节点(12)，所述数据存储节点(12)部署在分布式存储系统上，所述方法还包括：
    所述知识管理节点(11)将所述目标语义数据集发送给所述数据存储节点(12)；
    所述数据存储节点(12)将所述目标语义数据集存储在所述分布式存储系统上；
    所述数据存储节点(12)生成用于对所述目标语义数据集中的本体文件进行索引的第一索引文件，其中，所述第一索引文件链接到第二索引文件，所述第二索引文件用于对上一次来自所述知识管理节点(11)的语义数据集中的本体文件进行索引；
    所述数据存储节点(12)对所述第一索引文件进行存储；
    所述数据存储节点(12)计算所述第一索引文件的唯一标识，并将所述第一索引文件的唯一标识发送给所述知识管理节点(11)；
    所述知识管理节点(11)在确定所述目标语义数据集可信后，将所述第一索引文件的唯一标识存储到所述区块链(20)上。
  11. 根据权利要求10所述的方法，其特征在于，所述生成用于对所述目标语义数据集中的本体文件进行索引的第一索引文件，包括：
    根据所述目标语义数据集涉及的至少一个业务领域，生成至少一个所述域文件夹，其中，每个业务领域对应一个域文件夹，不同业务领域对应不同域文件夹，每个域文件夹中记录有文件链接，所述文件链接为所述分布式存储系统中属于该域文件夹所对应业务领域且为最新版本的各本体文件的链接，所述本体文件包括所述目标语义数据集中该业务领域的数据；
    生成用于链接到各所述域文件夹和所述第二索引文件的所述第一索引文件。
  12. 根据权利要求11所述的方法，其特征在于，所述方法还包括：
    所述数据存储节点(12)计算每个所述域文件夹的唯一标识和每个所述域文件夹中所记录的本体文件的唯一标识；
    所述数据存储节点(12)将每个所述域文件夹的唯一标识和每个所述域文件夹中所记录的本体文件的唯一标识发送给所述知识管理节点(11)；
    所述知识管理节点(11)在确定所述目标语义数据集可信后，将每个所述域文件夹的唯一标识和每个所述域文件夹中所记录的本体文件的唯一标识存储到所述区块链(20)上。
  13. 根据权利要求10所述的方法，其特征在于，所述将所述第一索引文件的唯一标识存储到所述区块链(20)上，包括：
    利用所述知识创建者设定的密钥对所述第一索引文件的唯一标识进行加密；
    将经过加密的所述第一索引文件的唯一标识存储到所述区块链(20)上。
  14. 根据权利要求11所述的方法，其特征在于，所述知识管理系统(10)还包括至少一个应用节点(13)，所述应用节点(13)部署在所述区块链(20)上，所述方法还包括：
    所述应用节点(13)获取用户查询所述目标语义数据集的知识查询请求；
    所述应用节点(13)将获取到的所述知识查询请求发送给所述知识管理节点(11)；
    在所述知识管理节点(11)本地已存储所述目标语义数据集时，所述知识管理节点(11)根据所述知识查询请求，通过所述智能合约校验所述目标语义数据集可信后，将所述目标语义数据集发送给所述应用节点(13)，以供所述应用节点(13)对所述目标语义数据集进行可视化。
  15. 根据权利要求14所述的方法，其特征在于，所述方法还包括：
    在所述知识管理节点(11)本地未存储所述目标语义数据集时，所述知识管理节点(11)根据所述知识查询请求从所述智能合约中获取所述第一索引文件的唯一标识；
    所述知识管理节点(11)将所述第一索引文件的唯一标识发送给所述数据存储节点(12)；
    所述数据存储节点(12)根据接收到的所述第一索引文件的唯一标识，从所述分布式存储系统中获取所述第一索引文件；
    所述数据存储节点(12)根据获取到的所述第一索引文件，从所述分布式存储系统中获取所述第一索引文件链接到的各所述域文件夹；
    所述数据存储节点(12)根据获取到的各所述域文件夹中记录的各文件链接，从所述分布式存储系统中获取所述目标语义数据集，并将获取到的所述目标语义数据集发送给所述知识管理节点(11)；
    所述知识管理节点(11)在通过所述智能合约校验获取到的所述目标语义数据集可信后，将所述目标语义数据集存储缓存到所述知识管理节点(11)本地；
    所述知识管理节点(11)将所述目标语义数据集发送给所述应用节点(13)，以供所述应用节点(13)对所述目标语义数据集进行可视化。
  16. 根据权利要求9至15中任一所述的方法，其特征在于，
    所述区块链(20)为私有链，和/或，所述分布式存储系统为星际文件系统IPFS。
  17. 一种知识管理装置，其特征在于，包括用于实现权利要求9-16中任一项所述方法中 所述知识管理节点(11)所执行操作的模块。
  18. 一种电子设备，其特征在于，包括：处理器(902)、通信接口(904)、存储器(906)和通信总线(908)，所述处理器(902)、所述存储器(906)和所述通信接口(904)通过所述通信总线(908)完成相互间的通信；
    所述存储器(906)用于存放至少一可执行指令，所述可执行指令使所述处理器(902)执行如权利要求9-16中任一项所述方法中所述知识管理节点(11)执行的操作。
  19. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质上存储有计算机指令，所述计算机指令在被处理器执行时，使所述处理器执行权利要求9-16中任一项所述方法中所述知识管理节点(11)执行的操作。
PCT/CN2021/109880 2021-07-30 2021-07-30 知识管理系统、方法、装置、电子设备和存储介质 WO2023004807A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2021/109880 WO2023004807A1 (zh) 2021-07-30 2021-07-30 知识管理系统、方法、装置、电子设备和存储介质
CN202180100338.2A CN117716349A (zh) 2021-07-30 2021-07-30 知识管理系统、方法、装置、电子设备和存储介质

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/109880 WO2023004807A1 (zh) 2021-07-30 2021-07-30 知识管理系统、方法、装置、电子设备和存储介质

Publications (1)

Publication Number Publication Date
WO2023004807A1 true WO2023004807A1 (zh) 2023-02-02

Family

ID=85087405

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/109880 WO2023004807A1 (zh) 2021-07-30 2021-07-30 知识管理系统、方法、装置、电子设备和存储介质

Country Status (2)

Country Link
CN (1) CN117716349A (zh)
WO (1) WO2023004807A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117150050A (zh) * 2023-10-31 2023-12-01 卓世科技(海南)有限公司 一种基于大语言模型的知识图谱构建方法及系统

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180129955A1 (en) * 2016-11-09 2018-05-10 Cognitive Scale, Inc. Hybrid Blockchain Data Architecture for use Within a Cognitive Environment
CN111177080A (zh) * 2019-12-31 2020-05-19 西安理工大学 一种基于区块链和ipfs的知识图谱存储与验证方法
US20200250295A1 (en) * 2019-01-31 2020-08-06 Salesforce.Com, Inc. Systems, methods, and apparatuses for protecting consumer data privacy using solid, blockchain and ipfs integration
CN111917861A (zh) * 2020-07-28 2020-11-10 广东工业大学 基于区块链和知识图谱的知识存储方法、系统及其应用
WO2020236832A1 (en) * 2019-05-23 2020-11-26 Healthpointe Solutions, Inc. System and method for using a blockchain to manage knowledge in a healthcare ecosystem
CN112348717A (zh) * 2020-11-25 2021-02-09 山东师范大学 一种基于区块链的知识上链与获取方法及系统

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180129955A1 (en) * 2016-11-09 2018-05-10 Cognitive Scale, Inc. Hybrid Blockchain Data Architecture for use Within a Cognitive Environment
US20200250295A1 (en) * 2019-01-31 2020-08-06 Salesforce.Com, Inc. Systems, methods, and apparatuses for protecting consumer data privacy using solid, blockchain and ipfs integration
WO2020236832A1 (en) * 2019-05-23 2020-11-26 Healthpointe Solutions, Inc. System and method for using a blockchain to manage knowledge in a healthcare ecosystem
CN111177080A (zh) * 2019-12-31 2020-05-19 西安理工大学 一种基于区块链和ipfs的知识图谱存储与验证方法
CN111917861A (zh) * 2020-07-28 2020-11-10 广东工业大学 基于区块链和知识图谱的知识存储方法、系统及其应用
CN112348717A (zh) * 2020-11-25 2021-02-09 山东师范大学 一种基于区块链的知识上链与获取方法及系统

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117150050A (zh) * 2023-10-31 2023-12-01 卓世科技(海南)有限公司 一种基于大语言模型的知识图谱构建方法及系统
CN117150050B (zh) * 2023-10-31 2024-01-26 卓世科技(海南)有限公司 一种基于大语言模型的知识图谱构建方法及系统

Also Published As

Publication number Publication date
CN117716349A (zh) 2024-03-15

Similar Documents

Publication Publication Date Title
JP6865851B2 (ja) ブロックチェーンワールド状態マークルパトリシアトライ木サブツリーの構築
CN114365133A (zh) 利用秘密共享和对读取的共识在元数据驱动的区块链上实现被遗忘的权利的系统或方法
JP2021527274A (ja) ブロックチェーンの分散アクセス、記憶、及び転送
CN114514732A (zh) 用于区块链dag结构的共识协议
US20140189118A1 (en) Maintaining consistent globally unique identifiers via an asynchronous interface
US10664451B1 (en) Systems and methods for encrypting data in backend storage caches shared by multiple decentralized applications
US9613044B2 (en) Maintaining concurrency and consistency of globally unique identifiers
US8930698B2 (en) Encrypting globally unique identifiers at communication boundaries
JP7502812B2 (ja) 共有クラウドプラットフォーム内のエンティティ中心のリソース指向データベースとして実装されたデータ管理者のための方法及びシステム
US11283595B1 (en) Systems and methods for securing cached data stored off-chain in a blockchain-based network
US20230224174A1 (en) File verification system and method
US20230342437A1 (en) Blockchain-based system and method for publishing an operating system
KR20230034931A (ko) 암호 데이터 입력 블록체인 데이터 구조
WO2021198750A1 (en) System and method to manage information and documents on a native blockchain network system including permissioned blockchain, storage, sharing, organisation, porting and various applications
US20240171407A1 (en) Improved methods & systems for signature verification in blockchain-implemented data applications
CN117043772A (zh) 区块链数据分离
Shah et al. M-DPS: a blockchain-based efficient and cost-effective architecture for medical applications
WO2023004807A1 (zh) 知识管理系统、方法、装置、电子设备和存储介质
US10657139B2 (en) Information processing apparatus and non-transitory computer readable medium for distributed resource management
Xu et al. A blockchain-based log storage model with efficient query
Khaleelullah et al. Verifcation of academic records using hyperledger fabric and ipfs
Gupta et al. Blockchain Enabled Hadoop Distributed File System Framework for Secure and Reliable Traceability
Li et al. Improved Efficient Privacy‐Preserving Certificateless Provable Data Possession Scheme for Cloud Storage
Ramachandran et al. Blockchain and Data Integrity Authentication Technique for Secure Cloud Environment
Jahan et al. Utilizing Hyperledger-Based Private Blockchain to Secure E-Passport Management

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21951397

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 202180100338.2

Country of ref document: CN

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21951397

Country of ref document: EP

Kind code of ref document: A1