Scientific and technological innovation field knowledge graph construction method based on block chain

Abstract

The invention relates to a scientific and technological innovation field knowledge graph construction method based on a block chain. The method comprises the steps of obtaining data of a scientific and technological innovation field; performing knowledge extraction on the semi-structured data and the non-structured data by utilizing a knowledge extraction model, and extracting entities, attributes and relationships among the entities; fusing the structured data and the extracted knowledge to obtain structured data for constructing a knowledge graph; performing uplink on the structured data of the knowledge graph to construct the knowledge graph; decomposing the structured data, and encrypting the decomposed structured data to obtain an encryption key and a decryption key; and storing the encrypted structured data to a storage node, verifying the user when the user selects to update the storage node, and updating the data in the storage interval through the decryption key if the verification succeeds. According to the invention, the risk of data tampering in the construction and use process of the knowledge graph can be prevented.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain. Background Art

[0002] Technological innovation is currently in a state of rapid development. On the one hand, information about technological innovation is scattered across various data sources, which is not convenient for users to collectively consult and integrate for analysis. On the other hand, information about technological innovation involves many technical fields, and this information is growing explosively, which is not conducive to users quickly retrieving key information from it.

[0003] A knowledge graph is a knowledge base presented in a graphical way. It is based on entities, concepts and their relationships, and organizes different types of knowledge through semantic connections and relationship networks. Knowledge graphs are now applied to many fields, but for the field of scientific and technological innovation, the extraction of unstructured and semi-structured knowledge in the field of scientific and technological innovation is currently incomplete, not detailed, and inconsistent with the direction of objective needs. In addition, in the process of constructing the knowledge graph, the data is often unencrypted. If there is tampered data, it will cause inaccurate problems in the construction of the knowledge graph, and then cause significant deviations in the overall query results of the knowledge in the field of scientific and technological innovation. In addition, in the update of the knowledge graph, it is also possible to face the security threat of modifying, adding and deleting data at will. Therefore, the present invention provides a method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain. Summary of the invention

[0004] The purpose of the present invention is to provide a method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain. On the one hand, by encrypting the data for constructing the knowledge graph, the process of constructing the knowledge graph is safer. On the other hand, by encrypting the structured data of the knowledge graph, the knowledge graph can be prevented from being tampered with and authorized users can update it.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] A method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain, comprising:

[0007] Obtain data in the field of scientific and technological innovation, including structured data, semi-structured data and unstructured data;

[0008] Use the knowledge extraction model to extract knowledge from semi-structured and unstructured data, and extract entities, attributes, and relationships between entities;

[0009] Merge structured data and extracted knowledge to obtain structured data for building knowledge graphs;

[0010] Put the structured data of the knowledge graph on the chain to build a knowledge graph;

[0011] Decomposing the structured data, encrypting the decomposed structured data, and obtaining an encryption key and a decryption key;

[0012] The encrypted structured data is stored in the storage node. When the user chooses to update the storage node, the user is verified. If the verification is successful, the data in the storage interval is updated using the decryption key.

[0013] Optionally, using the knowledge extraction model to extract knowledge from semi-structured data and unstructured data includes:

[0014] Segment text information of semi-structured data and unstructured data to obtain segmentation results;

[0015] Perform feature extraction and re-encoding on the word segmentation results to obtain the re-encoding results;

[0016] Decode the re-encoding result to obtain a decoding result;

[0017] Calculate the label probability of the decrypted result, obtain the label sequence, reorder the label sequence, and obtain the entities, attributes, and relationships between entities.

[0018] Optionally, the text information of the semi-structured data and the unstructured data is segmented, and the segmentation results include:

[0019] Eliminate stop words in the text information and obtain the text without the stop words;

[0020] Performing forward matching on the text with stop words removed according to the science and technology innovation field dictionary and the preset word segmentation dictionary to generate a first phrase set;

[0021] Perform negative matching on the text with stop words removed according to the science and technology innovation field dictionary and the preset word segmentation dictionary to generate a second phrase set;

[0022] The word segmentation result is determined according to the number of phrases in the first phrase set and the second phrase set.

[0023] Optionally, determining as the word segmentation result according to the number of phrases in the first phrase set and the second phrase set includes: comparing the number of phrases in the first phrase set and the second phrase set, and taking the set containing more phrases as the word segmentation result;

[0024] Or, when the number of phrases in the second phrase set is greater than the number of phrases in the first phrase set, if the number of individual words in the first phrase set is greater than the number of individual words in the second phrase set, the second phrase set is determined as the word segmentation result; if the number of individual words in the first phrase set is less than the number of individual words in the second phrase set, the first phrase set is determined as the word segmentation result.

[0025] Optionally, fusing the structured data and the extracted knowledge includes: calculating likelihood values ​​between different entities, and if the likelihood values ​​belong to a preset range, they are the same entity.

[0026] Optionally, encrypting the decomposed structured data, and obtaining an encryption key and a decryption key includes:

[0027] Obtaining a first encryption sequence and a first decryption sequence according to a data address corresponding to the decomposed structured data;

[0028] Obtaining a second encryption sequence and a second decryption sequence according to the structured data after the decomposition process;

[0029] Generate an encryption key according to the first encryption sequence and the second encryption sequence, and generate a decryption key according to the second decryption sequence and the second decryption sequence;

[0030] The verification key is sent to the blockchain center, and the decryption key is sent to the consensus node.

[0031] Optionally, storing the encrypted structured data in a storage node includes:

[0032] Obtaining an encryption key for the data address and obtaining a first encryption sequence;

[0033] The verification key of the blockchain center is retrieved to verify the first encryption sequence. If the verification passes, the structured data is allowed to be stored in the storage node; if the verification fails, it is discarded and the user of the address is locked.

[0034] Optionally, when a user chooses to update a storage node, user verification includes:

[0035] Obtain the user's basic information and match it with the user information of the storage node. If the match is successful, update the decomposed structured data in the data storage interval through the decryption key.

[0036] The beneficial effects of the present invention are as follows: the present invention extracts and integrates single and scattered knowledge in the field of scientific and technological innovation to form a scientific and technological innovation knowledge graph, thereby making a more comprehensive analysis of the field of scientific and technological innovation. The present invention uses blockchain technology to encrypt knowledge in the field of scientific and technological innovation to build a knowledge graph, making the process more secure. By encrypting the structured data of the knowledge graph, the knowledge graph can be prevented from being tampered with and can be updated by authorized users. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 This is a flow chart of a method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0041] Blockchain is a distributed, immutable digital ledger that records transactions and data and is shared across multiple nodes in a network. Each block contains a timestamp, transaction records, a hash of the previous block, and other data. When a new block is added to the chain, it is linked to the previous block, forming a continuous and immutable record.

[0042] like Figure 1 As shown, this embodiment provides a method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain, which uses blockchain as a basis to achieve data security and credibility assurance, and prevent the risk of data being tampered with during the construction and use of the knowledge graph, including:

[0043] Obtain data in the field of scientific and technological innovation, including structured data, semi-structured data and unstructured data;

[0044] Use the knowledge extraction model to extract knowledge from semi-structured and unstructured data, and extract entities, attributes, and relationships between entities;

[0045] Integrate structured data and extracted knowledge to achieve reference resolution, entity disambiguation, entity linking, etc., and obtain structured data for building knowledge graphs;

[0046] Upload the structured data of the knowledge graph to the blockchain to build a knowledge graph;

[0047] Decompose the structured data of the knowledge graph, encrypt the structured data of the decomposed knowledge graph, obtain encryption keys and decryption keys, and realize data security and non-tamperability through consensus algorithms; wherein, the structured data of the knowledge graph is decomposed according to the capsule grid algorithm;

[0048] The encrypted structured data is stored in the storage node. When the user chooses to update the storage node, the user is verified. If the verification is successful, the data in the storage interval is updated using the decryption key.

[0049] This embodiment combines blockchain with the construction of a knowledge graph, enabling users to securely update the knowledge graph.

[0050] In this embodiment, the data in the field of scientific and technological innovation can be any collected Chinese scientific and technological text files, where scientific and technological text files include but are not limited to patents, papers, and report data. If there are non-standard coding table characters in the unstructured data, including but not limited to formulas and special characters, placeholder characters are used to replace them, or unrecognizable characters are directly ignored; in addition, all English characters are lowercase.

[0051] In this embodiment, the entity data in the field of scientific and technological innovation mainly includes data generated through basic research, applied research, experimental development and other activities in the fields of natural sciences, engineering technology, etc., as well as data obtained through observation, monitoring, investigation, inspection and testing. Attributes of entity data: The attributes of entity data include but are not limited to the content, type, scale, quality, submission time, and name of the data management agency in the field of scientific and technological innovation. Relationships include cooperative relationships between data management agencies and associations between categories.

[0052] Furthermore, using the knowledge extraction model to extract knowledge from semi-structured and unstructured data includes:

[0053] Segment text information of semi-structured data and unstructured data to obtain segmentation results;

[0054] Perform feature extraction and re-encoding on the word segmentation results to obtain the re-encoding results;

[0055] Decode the re-encoding result to obtain a decoding result;

[0056] Calculate the label probability of the decrypted result, obtain the label sequence, reorder the label sequence, and obtain the entities, attributes, and relationships between entities.

[0057] Specifically, feature extraction and recoding of the word segmentation results include: vectorizing the word segmentation results, for the i-th word a in the word segmentation result A i , assuming that it consists of m words, each of which is composed of d1, ... d2, ... d n The characteristic value of the word is: Characteristic vector l = {q1, q2, ...q i , ..., q n}, use graph convolutional neural network to re-encode the feature vector. Get the re-encoding result. Then use LSTM to decode the text re-encoding result. Finally, calculate the label probability based on the softmax function.

[0058] In this embodiment, the knowledge graph structure is designed to include entities, attributes and relationships between entities, using a quadruple i , B i , C i , D i >Use the softmax function to calculate the label probability of the decoding result, A i is the word position label, indicating the position of the word in the sentence; B i is the indicator type label, indicating the attribute type of the word; C i It is the subject and object type label, indicating the part of speech of the word; D i is the relation type label, indicating the relation type of the triples in the sentence.

[0059] Furthermore, the text information of the semi-structured data and the unstructured data is segmented, and the segmentation results include:

[0060] Eliminate stop words in the text information and obtain the text without the stop words;

[0061] A forward match is performed on the text with stop words removed according to the dictionary in the field of scientific and technological innovation and the preset word segmentation dictionary to generate a first phrase set, wherein corpus samples in the field of scientific and technological innovation are collected and a dictionary in the field of patents is constructed based on the corpus samples.

[0062] Perform negative matching on the text with stop words removed according to the science and technology innovation field dictionary and the preset word segmentation dictionary to generate a second phrase set;

[0063] The word segmentation result is determined according to the number of phrases in the first phrase set and the second phrase set. ​

[0064] Specifically, according to the number of phrases in the first phrase set or the second phrase set, it is determined that the word segmentation result includes: when determining the first phrase set or the second phrase set as the word segmentation result, when the number of phrases in the first phrase set is greater than the number of phrases in the second phrase set, the first phrase set is determined as the word segmentation result; or when the number of phrases in the second phrase set is greater than the number of phrases in the first phrase set, the second phrase set is determined as the word segmentation result; or when the number of phrases in the second phrase set is greater than the number of phrases in the first phrase set, the number of single words in the first phrase set and the second phrase set is respectively counted; and when the number of single words in the first phrase set is greater than the number of single words in the second phrase set, the second phrase set is determined as the word segmentation result; or when the number of single words in the first phrase set is less than the number of single words in the second phrase set, the first phrase set is determined as the word segmentation result.

[0065] In this embodiment, common conjunctions in the scientific and technological innovation text information are used as stop words, such as "therefore", common adverbs in the scientific and technological innovation text information are used as stop words, such as "faster", common pronouns in the scientific and technological innovation text information are used as stop words, such as "it", common degree quantifiers in the scientific and technological innovation text information are used as stop words, such as "a kind of", common comparison verbs in the scientific and technological innovation document are used as stop words, such as "greater than", verbs that do not appear in key phrases in the scientific and technological innovation text information, such as "cannot", some nouns that do not appear in key phrases in the scientific and technological innovation text information, such as "limitations", and the quantifier regular expression is used as the stop pattern, such as "\d+ pieces".

[0066] Furthermore, the knowledge fusion of structured data and extracted knowledge includes: calculating the likelihood value between different entities, and if the likelihood value belongs to a preset range, they are the same entity.

[0067] Specifically, the calculation method of the likelihood value is as follows: Among them, the string corresponding to entity S is s, and the string corresponding to entity T is t.

[0068] Furthermore, the encryption of the decomposed structured data includes:

[0069] According to the data address corresponding to the decomposed structured data, obtain the first encryption sequence and the first decryption sequence;

[0070] According to the decomposed structured data, obtain the second encryption sequence and the second decryption sequence;

[0071] Generate an encryption key according to the first encryption sequence and the second encryption sequence, and generate a decryption key according to the second decryption sequence and the second decryption sequence;

[0072] The verification key is sent to the blockchain center, and the decryption key is sent to the consensus node.

[0073] Furthermore, storing the encrypted structured data in the storage node includes:

[0074] Obtaining an encryption key for the data address and obtaining a first encryption sequence;

[0075] The verification key of the blockchain center is retrieved to verify the first encryption sequence. If the verification passes, the structured data is allowed to enter the storage node; if the verification fails, it is discarded and the user of the address is locked.

[0076] Furthermore, when a user chooses to update a storage node, the user verification includes:

[0077] Obtain the user's basic information and match it with the user information of the storage node. If the match is successful, update the decomposed structured data in the data storage interval through the decryption key.

[0078] In this embodiment, user information has multi-level permission attributes; the graph data content specified by the graph data call request of user information with different levels of permission attributes is different. Therefore, this embodiment can be more conducive to improving the standardization of the use of knowledge graphs in the field of scientific and technological innovation by users with different levels of authority in the field of scientific and technological innovation; it can be conducive to improving the effectiveness of personnel training management, and avoiding the abuse of the value of graph knowledge, and has sufficient privacy protection.

[0079] The above embodiments are only descriptions of the preferred embodiments of the present invention, and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should fall within the protection scope determined by the claims of the present invention.

Claims

1. A method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain, characterized in that: include: Obtain data in the field of scientific and technological innovation, including structured data, semi-structured data and unstructured data; Use the knowledge extraction model to extract knowledge from semi-structured and unstructured data, and extract entities, attributes, and relationships between entities; Merge structured data and extracted knowledge to obtain structured data for building knowledge graphs; Put the structured data of the knowledge graph on the chain to build a knowledge graph; Decompose the structured data of the knowledge graph, encrypt the decomposed structured data, and obtain encryption keys and decryption keys; The encrypted structured data is stored in the storage node. When the user chooses to update the storage node, the user is verified. If the verification is successful, the data in the storage interval is updated using the decryption key.

2. The method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain according to claim 1 is characterized in that: Using knowledge extraction models to extract knowledge from semi-structured and unstructured data includes: Segment text information of semi-structured data and unstructured data to obtain segmentation results; Perform feature extraction and re-encoding on the word segmentation results to obtain the re-encoding results; Decode the re-encoding result to obtain a decoding result; Calculate the label probability of the decrypted result, obtain the label sequence, reorder the label sequence, and obtain the entities, attributes, and relationships between entities.

3. The method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain according to claim 2 is characterized in that: Segment text information of semi-structured data and unstructured data, and obtain segmentation results including: Eliminate stop words in the text information and obtain the text without the stop words; Performing forward matching on the text with stop words removed according to the science and technology innovation field dictionary and the preset word segmentation dictionary to generate a first phrase set; Perform negative matching on the text with stop words removed according to the science and technology innovation field dictionary and the preset word segmentation dictionary to generate a second phrase set; The word segmentation result is determined according to the number of phrases in the first phrase set and the second phrase set.

4. The method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain according to claim 1 is characterized in that: Determining the word segmentation result according to the number of phrases in the first phrase set and the second phrase set includes: comparing the number of phrases in the first phrase set and the second phrase set, and taking the set containing more phrases as the word segmentation result; Or, when the number of phrases in the second phrase set is greater than the number of phrases in the first phrase set, if the number of individual words in the first phrase set is greater than the number of individual words in the second phrase set, the second phrase set is determined as the word segmentation result; if the number of individual words in the first phrase set is less than the number of individual words in the second phrase set, the first phrase set is determined as the word segmentation result.

5. The method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain according to claim 1 is characterized in that: The fusion of structured data and extracted knowledge includes: calculating the likelihood values ​​between different entities. If the likelihood values ​​belong to a preset range, they are the same entity.

6. The method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain according to claim 1 is characterized in that: Encrypting the decomposed structured data and obtaining the encryption key and decryption key include: Obtaining a first encryption sequence and a first decryption sequence according to a data address corresponding to the decomposed structured data; Obtaining a second encryption sequence and a second decryption sequence according to the structured data after the decomposition process; Generate an encryption key according to the first encryption sequence and the second encryption sequence, and generate a decryption key according to the second decryption sequence and the second decryption sequence; The verification key is sent to the blockchain center, and the decryption key is sent to the consensus node.

7. The method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain according to claim 6 is characterized in that: Storing encrypted structured data in storage nodes includes: Obtaining an encryption key for the data address and obtaining a first encryption sequence; The verification key of the blockchain center is retrieved to verify the first encryption sequence. If the verification passes, the structured data is allowed to be stored in the storage node; if the verification fails, it is discarded and the user of the address is locked.

8. The method for constructing a knowledge graph in the field of scientific and technological innovation based on blockchain according to claim 7 is characterized in that: When a user chooses to update a storage node, the user verification process includes: Obtain the user's basic information and match it with the user information of the storage node. If the match is successful, update the decomposed structured data in the data storage interval through the decryption key.

Images