CN111708892A - Database system based on depth knowledge graph - Google Patents
Database system based on depth knowledge graph Download PDFInfo
- Publication number
- CN111708892A CN111708892A CN202010334955.6A CN202010334955A CN111708892A CN 111708892 A CN111708892 A CN 111708892A CN 202010334955 A CN202010334955 A CN 202010334955A CN 111708892 A CN111708892 A CN 111708892A
- Authority
- CN
- China
- Prior art keywords
- authority
- data
- association
- concept
- attribute
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/31—Indexing; Data structures therefor; Storage structures
- G06F16/316—Indexing structures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
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)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a database system based on a depth knowledge map, which comprises: a concept segmentation module for segmenting the whole entity objects of the knowledge graph into a plurality of subsets according to the entity object concepts; the association creating module is used for creating associations among the entity objects according to the plurality of divided subsets and generating an association object set; and the association calculation module is used for grouping the association object sets according to the association types and performing access calculation on the association object sets by using the independent memory database. By implementing the method, the problem that the traditional knowledge map database has insufficient information expression capability and cannot effectively support background data service requirements of most industrial and commercial application systems is solved. The computing efficiency in the ultra-large knowledge graph is improved, and the millisecond-level full-range text search is realized through the improvement in the algorithm.
Description
Technical Field
The invention relates to the field of computer science and information science, in particular to a database system based on a deep knowledge map.
Background
The knowledge graph is an information expression technology which is concerned about in the fields of artificial intelligence and knowledge question and answer in recent years. The traditional knowledge graph generally constructs and manages a knowledge structure in a certain field based on entities and the association between the entities, so that the purpose of knowledge inquiry or man-machine conversation is achieved.
Theoretically, the knowledge graph is universal to the expression of knowledge, and the principle of physical entity and abstraction is modeled into an entity, however, at present, the traditional knowledge graph is difficult to be used as an information chassis to support the development of a practical large-scale information system. The problems mainly come down as follows: 1) at present, the information expression capability of a knowledge map database is still low, and the requirement of developing any general information system is difficult to adapt. 2) With the gradual development of the mobile internet, in recent years, multimedia data, especially video stream and audio stream data, become indispensable components in information systems, and the conventional knowledge map database lacks an integrated solution for storing and extracting mass multimedia data. 3) The bottleneck of the calculation performance is reduced due to the expansion of the data volume as the spectrum scale is expanded. Such that the basic operation like the association query takes time beyond the acceptable range (in milliseconds) for general applications. 4) The prior knowledge map database is biased to express surface knowledge, and has no interface convention for detailed database operation.
Based on the limited information expression, the limitation of the computational efficiency, the fact of the universal query protocol and other reasons, no commercial knowledge map database which is mature enough can provide products with stable performance, rich enough information expression capability and convenient interface use for various information systems so far.
Disclosure of Invention
Therefore, the technical problem to be solved by the present invention is to overcome the limitations of information expression, computational efficiency, lack of multimedia data integration scheme, and lack of general query protocol in the prior art, thereby providing a database system based on a deep knowledge graph.
Therefore, the embodiment of the invention provides the following technical scheme:
the embodiment of the invention provides a database system based on a depth knowledge graph, which comprises: the system comprises a concept segmentation module, an association creation module and an association calculation module, wherein the concept segmentation module is used for segmenting all entity objects of a knowledge graph into a plurality of subsets according to entity object concepts; the association creating module is used for creating associations among the entity objects according to the plurality of divided subsets and generating an association object set; and the association calculation module is used for grouping the association object sets according to the association types and performing access calculation on the association object sets by using the independent memory database.
In one embodiment, an association computation module includes: the association type submodule is used for appointing the association type between any two entity objects and inquiring and extracting according to different association types in association calculation; and the association strength submodule is used for depicting the association strength between any two entity objects through the real number of a preset interval.
In one embodiment, the structure of the entity object includes: basic data, attribute tag data and extended attribute data, wherein the basic data comprises identification information, concept classification, name and associated information of an entity object; attribute tag data is attached to an entity object in a key value pair mode to describe the characteristics of the entity object; the extended attribute data includes information of all data structures subordinate to the entity object.
In one embodiment, the information of the extended attribute data structure includes: the data type and bear the weight of the structural container of the data type, wherein, the data type includes: a character data type and a binary data type; the structural container includes: a list structure, a time series structure, and a mapping table structure.
In one embodiment, the database system based on the depth knowledge-graph further comprises: the operation function module is used for creating, deleting, modifying and inquiring the concept, the entity object, the attribute tag and the extended attribute; the extended attribute data operation module is used for performing data adding, modifying, screening, converting, paging and calculating operations on data information in the extended attributes; the searching module is used for searching the entity objects of the database based on a reverse index mode; the file management module is used for embedding the multimedia resources into the knowledge graph through a multimedia file management framework; the file query module is used for extracting, converting and basically operating file data; the authority management module is used for controlling access authority among different entity objects through a preset authority management mechanism; and the interface module is used for establishing a database operation interface related to the concept, the entity object and the object user.
In one embodiment, the search module performs reverse index calculation based on three levels of information of the entity object within a range obtained by conceptual division by using a set function of the database.
In one embodiment, the three levels are an object level, an object attribute level, and a unit level within an attribute, respectively.
In one embodiment, the rights management mechanism includes: receiving a request sent by a user object, wherein the request comprises identification information of the user object; extracting a main body and an authority range of the operation according to the identification information of the user object; judging whether the user object has the authority for performing the current operation according to the main body of the operation and the authority range; and after judging that the user object has the authority for performing the current operation, executing the authority and returning an operation result to the user object.
In one embodiment, the authority range is determined by attribute tag data and extended attribute data stored in the entity object, and includes: judging whether the user object has the authority for performing the current operation according to the main body and the authority range of the operation, wherein the judgment includes the judgment of whether the concept level has the authority for performing the current operation; when the concept level has the authority to carry out the current operation, executing the authority and returning the operation result to the user object; when the concept level has no authority for performing the current operation, judging whether the object level has the authority for performing the current operation; when the object level has the authority to carry out the current operation, executing the authority and returning the operation result to the user object; and when the object level has no authority for performing the current operation, judging that the user object has no authority for performing the current operation.
The technical scheme of the invention has the following advantages:
1. the database system based on the depth knowledge graph solves the problems that the traditional knowledge graph database has insufficient information expression capability and can not effectively support background data service requirements of most industrial and commercial application systems by introducing a depth knowledge graph mechanism. The bottom layer of the database supports various basic data operations to meet the requirements of different application scenarios.
2. The database system based on the deep knowledge graph allows the construction and the use of an ultra-large knowledge graph database, and the number and the scale of nodes can reach hundred million. By introducing a concept division mechanism into the knowledge graph, the problem of low calculation efficiency caused by excessive object nodes in the ultra-large knowledge graph is solved. The method can realize the relevance query at the millisecond level, and greatly improves the practicability of core operation in the knowledge graph.
3. According to the database system based on the deep knowledge graph, the management frame of the multimedia file is completely embedded in the deep knowledge graph database system, so that the multimedia management frame and the entity object data of the subject database can be seamlessly matched to jointly represent an entity object with any multiple multimedia information in the entity object, and the problem that a general knowledge graph database supports multimedia data difficultly is solved.
4. The database system based on the deep knowledge map improves the authority management mechanism and establishes the concept of the user authority proxy object. Access authority control among different objects is realized through authority association, and the authority control can be embodied to the attribute level inside the object, so that fine data reading and editing control is realized.
5. The database system based on the deep knowledge graph provided by the invention adopts a search method based on reverse index, and realizes a global text search algorithm in a knowledge graph database. The method realizes object level and intra-object level search, and realizes full-range text search in millisecond level through improvement in algorithm.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic block diagram of a specific example of a deep-knowledge-map-based database system in an embodiment of the present invention;
FIG. 2 is a tree structure diagram between different concepts in an embodiment of the invention;
FIG. 3 is a functional block diagram of a specific example of an association computation module in an embodiment of the invention;
FIG. 4 is a data structure diagram of a database system based on a depth knowledge map in an embodiment of the present invention;
FIG. 5 is a diagram of a structural container type in an embodiment of the present invention;
FIG. 6 is a schematic block diagram of another specific example of a depth-knowledge-map-based database system in an embodiment of the present invention;
FIG. 7 is a diagram of the relationship between the resource management subsystem framework and other components in an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In addition, the technical features involved in the different embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Examples
The embodiment of the invention provides a database system based on a depth knowledge graph, as shown in fig. 1, comprising: a concept segmentation module 1, an association creation module 2 and an association calculation module 3, wherein,
a concept segmentation module 1 for segmenting the whole entity objects of the knowledge graph into a plurality of subsets according to the entity object concepts.
In the embodiment of the invention, a definition of 'concept' is introduced into the knowledge graph. The concept divides all entity objects of the knowledge graph into a plurality of subsets according to respective basic properties, on one hand, the problem of data division under the condition of a large amount of data is solved, and on the other hand, a database user has clear expectation on logic of different entity objects. As shown in fig. 2, a tree structure with a definite hierarchical relationship is formed between different concepts, and the system is continuously differentiated from the most rooted concept to create more concepts to meet the requirements of different application scenarios. The concept differentiation and setting are defined by users, which means that the database system has general application. In addition, after a concept division mechanism is introduced, the calculation efficiency problem of the most common operation in the knowledge graph and the association extraction operation between the entities has a better solution. The knowledge graph is a graph structure at the top, and what has a large influence on association calculation in the graph structure is a 'super node', namely a network node with a very high association degree. Performing correlation calculations on super nodes involves the extraction of large amounts of data, often causing high delays in the calculations. After a concept segmentation mechanism is introduced, the association calculation can be carried out by concepts, so that the calculation amount and the information extraction amount of the association calculation of the large node are effectively controlled.
And the association creating module 2 is used for creating the association among the entity objects according to the plurality of divided subsets and generating an association object set. And the association calculation module 3 is used for grouping the association object sets according to the association types and performing access calculation on the association object sets by using the independent memory database.
In the embodiment of the invention, by introducing the definition of the concept, the correlation operation of the knowledge graph is greatly improved in both the calculation efficiency and the use simplicity, so that the database product can better support the development of an information system. The database improves the performance and the universality of the correlation calculation from three aspects.
First, by introducing concept definition, all entity objects are divided into a plurality of subsets according to their respective service properties, so that all object sets associated with any object can be extracted in groups according to respective concepts, thereby greatly accelerating the information extraction speed of the same kind of objects and simultaneously increasing the controllability of the result of the association calculation.
Second, by allowing custom relationship types to be added, the database supports extracting data in two dimensions, a concept type and an association type. The database manually encodes the relationship between the entity objects, and ensures that a user can strictly control the relationship property of the extracted result object in the associated data extraction process. The method aims to meet the requirement of classification management on the relationship between data in an information system, so that the function of a bottom-layer database is better exerted. Because the coding of the relation is user-defined, the method can completely adapt to the requirements of different services in different fields.
Third, the efficiency of the calculation of the association intersection of multiple entity objects is ensured by using an independent underlying data structure. Finding all entity object sets having specific associations with any plurality of entity objects is a common operation, in order to deal with such operations, the database accesses all object sets associated with the primary keys by using the memory database alone and realizes intersection operations of any plurality of sets, and meanwhile, the association intersections are grouped according to the associated types, so that efficient type-divided association intersection operations are realized.
According to the database system based on the depth knowledge graph, provided by the invention, by introducing a depth knowledge graph mechanism, all entity objects of the knowledge graph are divided into a plurality of subsets, and association among the entity objects is established according to the information structure of the entity objects and is subjected to association calculation, so that the problems that the traditional knowledge graph database is insufficient in information expression capacity and cannot effectively support background data service requirements of most industrial and commercial application systems are solved. The calculation efficiency in the ultra-large knowledge graph is improved.
In a specific embodiment, as shown in fig. 3, the association calculation module 3 includes:
and the association type submodule 31 is used for specifying the association type between any two entity objects and querying and extracting according to different association types in association calculation. In the embodiment of the invention, by creating specific types of associations among various different objects and performing association calculation, the database can provide a series of objects associated with specific objects for a user or find out the association among any one object set.
The association strength sub-module 32 is configured to depict the association strength between any two entity objects through the real number of a preset interval. In the embodiment of the invention, the association strength concept is introduced to further strengthen the accuracy of the association operation. The intensity between any two entity objects is characterized by a real number between 0 and 100, and different intensities can reflect different associated elements in different business contexts. In the case of two users, the strength of association may reflect the degree of closeness between the two users.
In a specific embodiment, the structure of the entity object includes: basic data, attribute tag data and extended attribute data, wherein the basic data comprises identification information, concept classification, name and associated information of an entity object; attribute tag data is attached to an entity object in a key value pair mode to describe the characteristics of the entity object; the extended attribute data includes information of all data structures subordinate to the entity object.
In the embodiment of the invention, the entity object is used as an atomic structure forming the knowledge graph, and the internal structure and information expression of the entity object are the core of the depth knowledge graph. The structure of the entity object is divided into three parts, basic data, attribute tag data and extended attribute data, which respectively have the following meanings:
basic data: and storing all core related data such as identification information, concept classification, name and associated information of the entity object. The identification information is referred to as an object primary key in the database, and the object primary key plays a role in uniquely identifying the object in the database. The concept classification describes the concept range to which the object belongs, and the names are divided into basic names and alternative names, wherein the basic name is the most standard natural language expression of the object, and the alternative names are the set of all possible language expression modes. The association information includes information summarizing the number of other objects associated with the object, and the main purpose is to accelerate association calculation.
Attribute tag data: any entity object can be hung on any plurality of attribute data in the form of key-value pairs, and the data can be of a character string type, a numerical value type or a composite type. For example, the people's republic of china is an object under the concept of a country, and the object has a series of attribute data of the territorial area of 960 ten thousand square kilometers, the capital city of Beijing, and the like. The user can create and edit any custom attribute data according to the service use requirement.
Extended attribute data: the information of all complex data structures belonging to the object is stored, including list class data and dictionary class data, and the information expression capability of the complex data structures is greatly enhanced because the complex data structures have strong information expression capability and the system does not limit the number of the extended attributes bound on a certain object. The extended attribute is divided into several basic types, each basic type is divided into several sub-types, and a specific data structure is shown in fig. 4. A common operation set based on a data structure is specified for each different type of extended attribute data structure, and is used for meeting the use requirements of most information systems. In all data types, the reference type list allows the extended attribute of any entity object to contain references to another series of objects, and this mechanism allows an object to logically contain an indefinite number of other objects and to literally specify the reason for the reference at the corresponding reference. Therefore, the deep knowledge map structure allows a user to describe the internal structure of an object with any structural complexity, and the complexity of any object can be recursively contained in other larger objects, so that a large tree information structure is automatically formed from the top-level object perspective, each node in the tree represents an object, and the connection between the nodes is the reference relation in the extended attribute. As a specific example, a cross-country large company A has 10 subsidiaries, a headquarters consists of 15 departments, and each of the subsidiaries consists of a plurality of subsidiary departments. In this case, the whole architecture of the multinational company a is a tree structure formed by branch companies, general company departments and branch company departments, each organization is represented by an entity object, and the reference attribute in the object describes the mutual membership between different entities.
The format protocol of the extended attribute data and the supported operation set are one of the core contents of the deep knowledge graph database system, wherein a reference attribute part related to an entity object and an entity association concept in a knowledge graph coexist, the former pays more attention to the information structure expression in the object, and the latter plays a larger role in association calculation of different entities.
In a specific embodiment, the information of the extended attribute data structure includes: the data type and bear the weight of the structural container of the data type, wherein, the data type includes: a character data type and a binary data type; the structural container includes: a list structure, a time series structure, and a mapping table structure.
In the embodiment of the invention, any object can contain any plurality of attributes, and the complexity and the universality of each attribute are determined by the combination of two factors of a structure container and a data type. The structure container is several data structure types for carrying specific data, and the data type is the data type actually stored in the container. Data types are divided into character data and binary data (files), the character data are divided into words, numerical values and object references according to the meaning of the representation, and the binary data are various and can be roughly divided into pictures, videos, audios, executable files, documents and the like.
And the structure container for holding these data, as shown in fig. 5, has three basic types of (a) list structure, (b) time-series structure, and (c) mapping table structure. The combination of container structures and data types generates a great number of combined data storage types for any information platform, thereby providing a rich use scene. Wherein the list is divided into: text lists, numerical lists, object reference lists, file lists (pictures, video, audio, executable files, documents, etc.). The time series is divided into: text time series, numerical time series, object reference time series, file time series (pictures, video, audio, executable files, documents, etc.). The mapping table is divided into: word maps, value maps, object reference maps, file maps (pictures, video, audio, executable files, documents, etc.).
In a specific embodiment, the database system based on the depth knowledge graph, as shown in fig. 6, further includes:
and the operation function module 4 is used for creating, deleting, modifying and inquiring the concept, the entity object, the attribute tag and the extended attribute. In the embodiment of the invention, the operations of concept addition, deletion, modification and check, object level addition, deletion, modification and check and attribute level addition, deletion, modification and check are the basis of the database.
And the extended attribute data operation module 5 is used for performing data adding, modifying, screening, converting, paging and calculating operations on the data information in the extended attribute. In the embodiment of the invention, the creation, modification, query and operation of the data information in the extended attribute are specific operations of the depth knowledge graph.
And the searching module 6 is used for searching the entity objects of the database based on a reverse index mode. In the embodiment of the invention, a global text search algorithm is realized in a large-scale deep knowledge map database. The user can perform global or local text search according to the application requirements. The database system adopts a search method based on reverse index to realize object level and object inner level search, and realizes millisecond-level full-range text search through improvement in the algorithm.
And the file management module 7 is used for embedding the multimedia resources into the knowledge graph through a multimedia file management framework. In the embodiment of the invention, a multimedia file management framework is completely embedded in a deep knowledge map database system, and the multimedia management framework and the entity object data of a subject database can be seamlessly matched through a url technology to jointly represent an entity object with any multiple multimedia information in the entity object, so that the problem that a general knowledge map database is difficult to support multimedia data is basically solved.
In practical application, any text information is stored in an entity object in a deep knowledge graph in a specific type of attribute information, in order to store specific multimedia information in the attribute, a database compiles each unique resource type data into a url by using the concept of the url, and then an independent resource data management module embedded in a system converts the url corresponding to each resource into actual resource file data in a network service mode. In this way, various multimedia assets, including but not limited to pictures, video, audio, html documents, etc., are embedded in the deep knowledge graph in a uniform manner.
All files uploaded to the system by the user are finally transmitted to different resource management service frameworks through the main service framework. The framework of the resource management subsystem is shown in FIG. 7 in relation to other system components. First, the file asset management service is one or more that run independently for binary file uploads and downloads. In the database main service, all binary files are converted into a connection character string pointing to the resource, so that the object as a container can contain file data as common characters. In addition, to manage all files, there is a file explorer object in the main service framework, whose attributes store the actual locations of all file resource connections, enabling any file-oriented request to be directed to the corresponding file explorer service. The embodiment of the invention integrates two storage media of text data and file data, uniformly stores and manages the two types of information in the same knowledge map, and can use all hardware resources more concisely and efficiently.
And the file query module 8 is used for extracting, converting and basically operating file data. The embodiment of the invention comprises the operations of compressing, zooming, rotating and color changing of the picture file; compressing, fragmenting, merging, screenshot and converting the video file; audio compression, fragmentation, merging, conversion operations.
And the authority management module 9 is used for controlling access authority among different entity objects through a preset authority management mechanism. In the embodiment of the invention, the authority management of information is a necessary component for the normal operation of any large-scale information system, and particularly for a random read-write database with a large number of users, which data can be read and written by which part of users should be one of core mechanisms considered by the database.
The interface module 10 is used for establishing a database operation interface related to concepts, entity objects and object users.
In the embodiment of the invention, the concept related operation interface comprises: the system comprises a concept list interface, a concept detail interface, a concept creation interface, a concept moving interface, a concept deletion interface and a concept tree diagram interface, wherein the concept list interface is used for acquiring all object information of a specified concept. By passing the concept number parameter, the paging returns the basic information of all the objects under the concept, such as the total number of the objects, the standard name and the alternative name of each object. And the concept detail interface is used for acquiring basic information of the specified concept. Basic information such as concept names, concept numbers and the like related to specified concepts is received by transmitting the concept number parameters. A concept creation interface for concept creation. And transmitting the basic information and the upper concept number of the concept, and returning the success/failure state of the concept creation. And the concept moving interface is used for moving the concept A to the concept B. The concept number and the target concept number are passed and the status of the return concept move is success/failure. A concept deletion interface for deleting a specified concept from the database. The concept number is passed and the status of the returned concept deletion is success/failure. And the concept tree diagram interface is used for acquiring all the lower concepts of the specified concept under the specified project. And transmitting the concept number, and returning all lower concept names under the specified concept and basic concept information thereof, such as the concept name, the concept number, the concept direct lower concept name and the like.
In the entity object related operation interface, the object creation interface is used for creating the object, transmitting the main key of the object and the basic information of the object, and receiving the returned object creation success/failure state. And the object information modification interface is used for modifying the object basic information, transmitting the object primary key and the object basic information and receiving the returned object modification success/failure state. And the object detail information interface is used for acquiring an object detail page and transmitting the object main key to receive the returned basic information of the object and all the auxiliary information data, such as list attributes, character mapping attributes and number mapping attributes. And the related object query interface is used for checking the details of all related objects of the specified object under the specified concept, transmitting the main key and the target concept number of the object, and receiving the object which is returned by paging and related to the object under the target concept and the basic information thereof, such as the standard name and the alternative name of the object. And the list additional attribute screening interface is used for screening the data through some time conditions and position index conditions, transmitting the object main key, the additional attribute name and the screening conditions, and receiving the returned data meeting the screening conditions.
And the short attribute creating interface is used for creating data of the specified object short attribute, transmitting the object main key, the short attribute name and the short attribute value and receiving the returned short attribute creating state. And the short attribute modification interface is used for modifying the value of the short attribute of the specified object, transmitting the main key, the name and the value of the short attribute of the object and receiving the returned short attribute modification state. And the short attribute deleting interface is used for deleting the short attribute of the specified object, transmitting the main key and the name of the short attribute of the specified object and receiving the returned short attribute deleting state.
And the text mapping additional attribute creating interface is used for creating an additional attribute as a text mapping table, transmitting the main key of the object, the name of the text mapping attribute and related information, and receiving a returned text mapping attribute creating state. And the text mapping additional attribute deleting interface is used for deleting the text mapping additional attribute of a specified object, transmitting the main key of the object, transmitting the name of the text mapping additional attribute and receiving the deleted state of the returned text mapping attribute. And the text mapping additional attribute data adding interface is used for adding data into the specified text mapping attribute, transmitting the object main key, the attribute name and the key value and receiving the addition state of the returned text mapping data. And the text mapping additional attribute data deleting interface is used for deleting data from the specified text mapping attribute, transmitting the object main key, the attribute name and the key value and receiving the deleted state of the returned text mapping attribute data. And the text mapping additional attribute data modification interface is used for modifying the value corresponding to the specified key in the specified text attribute, transmitting the object main key, the attribute name and the key value and receiving the modification state of the returned text mapping data. The text-mapping additional attribute data text import interface is used for adding data of text-mapping additional attributes in batches through text import, transmitting an object main key, an attribute name and a text file and receiving the import state of returned text-mapping additional attribute data.
And the digital mapping additional attribute creating interface is used for creating an additional attribute as a digital mapping table, transmitting the main key of the object, the name of the digital mapping attribute and related information, and receiving the returned digital mapping attribute creating state. And the digital mapping additional attribute deleting interface is used for deleting the digital mapping additional attribute of a specified object, transmitting the main key of the object and the name of the digital mapping additional attribute and receiving the returned deleting state of the digital mapping attribute. And the digital mapping additional attribute data adding interface is used for adding data into the designated digital mapping attributes, transmitting the object main key, the attribute name, the key and the number, and receiving the addition state of the returned digital mapping data. And the digital mapping additional attribute data deleting interface is used for deleting data from the designated digital mapping attribute, transmitting the object main key, the attribute name and the number, and receiving the deleting state of the returned digital mapping attribute data. And the digital mapping additional attribute data modification interface is used for modifying the value corresponding to the designated key in the designated digital attribute, transmitting the object primary key, the attribute name and the key value and receiving the modification state of the returned digital mapping data. The digital mapping additional attribute data text import interface adds data of digital mapping additional attributes in batches through text import, transmits object main keys, attribute names and text files, and receives the import state of returned digital mapping additional attribute data.
And the list mapping additional attribute creating interface is used for creating an additional attribute as a list mapping table, transmitting the object main key, the list mapping attribute name and related information, and receiving a returned list mapping attribute creating state. And the list mapping additional attribute deleting interface is used for deleting the list mapping additional attribute of a specified object, transmitting the main key of the object and the name of the list mapping additional attribute, and receiving the returned list mapping attribute deleting state. And the list mapping additional attribute data adding interface is used for adding data into the specified list mapping attribute, transmitting the object primary key, the attribute name and the key value and returning to the list mapping data adding state. And the list mapping additional attribute data deleting interface is used for deleting data from the specified list mapping attribute, transmitting the object main key, the attribute name and the key value and receiving the returned list mapping attribute data deleting state. And the list mapping additional attribute data modification interface is used for modifying the value corresponding to the specified key in the specified list attribute, transmitting the object primary key, the attribute name and the key value and receiving the returned list mapping data modification state. And the time sequence text import interface is used for adding the time sequence data in batches through text import, transmitting the object main key, the attribute name and the text file and receiving the returned time sequence text import state.
And the association creation interface is used for creating an association for the two object main keys, transmitting the two object main keys and the associated numbers and receiving a returned association creation state. And the association deletion interface is used for releasing the association information of the two object main keys, transmitting the two object main keys and receiving the returned association deletion state. And acquiring all concept list interfaces related to the designated object, extracting a concept list related to the designated object and other concepts, transmitting a main key of the object, and receiving the returned concept list and basic information of each concept, such as concept name, concept number and related quantity. And the object associated tag adding interface is used for adding an associated tag for the specified object, transmitting the main key and tag data of the object and receiving the returned object associated tag adding state. And the object associated tag deleting interface is used for deleting the associated tag for the specified object, transmitting the main key and tag data of the object and receiving the returned object associated tag deleting state. And the object associated tag modification interface is used for modifying the associated tag for the specified object, transmitting the main key and tag data of the object and receiving the returned object associated tag modification state.
And in the operation interfaces related to the object user, a login interface is used for logging in an account password, transmitting a user name and a password and receiving a returned login state and session information. And the exit interface is used for exiting the account number, transmitting the sessionid and receiving the returned exit state of the user. And the user creating interface is used for creating a user under the login of the super user, transmitting and inputting a specified user name, a password, a mailbox and a mobile phone number, and receiving a returned user creating state. And the user password modification interface is used for appointing a user to modify the password, transmitting and inputting an appointed user name and the password and receiving a returned user password modification state. And the concept permission adding interface is used for adding permission of a designated concept for a designated user, transmitting a concept number and permission and receiving a returned permission modification adding state. And the concept permission modification interface is used for modifying the permission of the specified concept for the specified user, transmitting the concept number and the permission and receiving the returned permission modification state. And the concept permission deleting interface is used for deleting the permission of the appointed concept for the appointed user, transmitting the concept number and the permission and receiving the returned permission deleting state. And the object authority adding interface is used for adding the authority of the appointed concept for the appointed user, transmitting the main key and the authority of the object and receiving the returned authority adding state. And the object permission modification interface is used for modifying the permission of the designated concept for the designated user, transmitting the main key and the permission of the object and receiving the returned permission modification plus state. And the object permission deleting interface is used for deleting the permission of the designated concept for the designated user, transmitting the main key and the permission of the object and receiving the returned permission deleting state. The most important external service protocol of the database is an operation interface, the operation specification and the function of the database are definitely given by the knowledge-graph-based database provided by the embodiment of the invention, and interface convention which is refined aiming at the database operation is carried out.
In a specific embodiment, the search module performs calculation of reverse indexes based on information of three levels of entity objects within a range obtained by conceptual division by using a set function of the database.
In the embodiment of the invention, the retrieval of the content is completed by adopting a reverse index-based mode. Firstly, independent retrieval query is allowed to be carried out in any range through concept division, so that the problem of low calculation efficiency caused by excessive later-stage screening of retrieval results only in a global range is solved; secondly, the calculation of the reverse index is completed by using a set function in a memory database redis realized, so that the calculation speed is ensured; thirdly, decomposing and splitting the whole information of the object into three layers of information and respectively establishing a searching mechanism. These three levels are the object level, the object attribute level, and the unit level within the attributes, respectively. The three-layer searching mechanism ensures that the user can search the text contents of the relation of the user on different layers and accurately position the corresponding object part.
In a specific embodiment, the rights management mechanism includes: receiving a request sent by a user object, wherein the request comprises identification information of the user object; extracting a main body and an authority range of the operation according to the identification information of the user object; judging whether the user object has the authority for performing the current operation according to the main body of the operation and the authority range; and after judging that the user object has the authority for performing the current operation, executing the authority and returning an operation result to the user object.
In the embodiment of the invention, the management of the user authority is completed through the user object. The user concept is a special concept reserved in a database, and the following objects are user objects. A user object represents a real user and the system needs to one-to-one binding sessionid (a key for login) to the user object at the bottom level. Since a user may log on multiple devices, sessionid can only represent the log-on of a client in the system, and a user object may correspond to multiple sessionids, i.e. clients, which all share the rights attribute of the user. Therefore, a natural person user can log in the database through a plurality of clients, and each client binds with the user object in the database in a one-to-one mode through sessionid so as to achieve the purpose of representing the authority of the natural user by using the virtual user object.
The basic mode of rights management is that when any user object attempts to access or modify information of another object, a rights attribute system inside the object dynamically determines whether the operation can be authorized. First, sessionkey contains identity information during all interactions with the database. According to the identity information contained in the sessionkey, the system determines the main body of the operation as which user object, and extracts the authority range. And checking whether the corresponding user main body has the authority for performing the current operation or not according to the type of the operation, the main body of the operation and the authority range. If the authority for carrying out the current operation exists, executing, and if the authority does not exist, returning no authority information. If the authority authorization is successful, executing the action and returning data; if the authorization is not successful, the operation is returned without enough authority information.
In one embodiment, the permission range is determined by attribute tag data and extended attribute data stored in the entity object, and includes: judging whether the user object has the authority for performing the current operation according to the main body and the authority range of the operation, wherein the judgment includes the judgment of whether the concept level has the authority for performing the current operation; when the concept level has the authority to carry out the current operation, executing the authority and returning the operation result to the user object; when the concept level has no authority for performing the current operation, judging whether the object level has the authority for performing the current operation; when the object level has the authority to carry out the current operation, executing the authority and returning the operation result to the user object; and when the object level has no authority for performing the current operation, judging that the user object has no authority for performing the current operation.
In the embodiment of the invention, the information authority of the user is divided into the authority at the entity object level and the authority at the attribute level. Object level rights control whether a user can perceive the presence of an object, while attribute level rights manage whether a user has the right to read or write for different attributes. The object level viewing rights describe the most basic rights of any user for an object, i.e. having a viewing right is the basis for having other information rights, whereas the absence of a viewing right means that the object is completely invisible to the user, neither in the list nor in the search results, who cannot perceive the presence of the object and therefore does not have any rights to read and write. If the user has viewing rights to the entity object itself, then the next step is whether certain operational rights are present for the particular object attributes.
The attribute authority of the object is divided into reading and writing, and the two information authorities of different users to different object attributes are recorded in the object attributes of the users. This attribute is called the user's rights attribute, where the objects the user has read and write rights to and the corresponding attribute names are recorded in the form of a list of references.
The rights of the user object are stored by attribute data within the user object representing the user. The authority is divided into: the concept level rights and the object level rights indicate the user's access rights to the concept and object levels, respectively. As shown in the following table, the two levels of access right range descriptions for concept and object are as follows:
the database checks whether a user object has a certain authority to a specific concept or object, and judges from macro to micro, firstly judges from the concept level, and then judges the object level. For example, an object a belongs to the concept X, and now a user corresponding to a client is U, and when the client wants to read a system attribute of a, the client needs to view the authority. Then the database will start to judge from the authority at the conceptual level, and the logic is as follows:
1) and finding the authority representation corresponding to the concept X from the concept authority mapping table of the user U, and judging whether the authority contains a read attribute. If yes, the operation is considered to be possible; if not, entering the permission viewing of the object level;
2) and finding the authority representation corresponding to the object A from the concept authority mapping table of the user U, and judging whether the authority contains a read attribute. If yes, the operation is considered to be possible; if not, the authority is determined to be insufficient.
Therefore, the expression of the rights has a progressive nature, and all the rights information in the above table appears in the conceptual rights table as well as the object rights table. If a concept authority table of an object authorizes a certain authority of the object, the user is considered to have corresponding authority for all objects under the concept.
The database system based on the deep knowledge graph provided by the invention divides the whole entity object of the knowledge graph into a plurality of subsets, establishes the association between the entity objects according to the information structure of the entity objects and carries out association calculation, and adopts a search method based on reverse index to realize the search of object level and object inner level. By introducing a deep knowledge map mechanism, the problems that the traditional knowledge map database has insufficient information expression capability and cannot effectively support background data service requirements of most industrial and commercial application systems are solved. A concept division mechanism is introduced into the knowledge graph, so that the problem of low calculation efficiency caused by too many object nodes in the ultra-large knowledge graph is solved. The method can realize the relevance query at the millisecond level, and greatly improves the practicability of core operation in the knowledge graph. A management framework of a multimedia file is completely embedded in a deep knowledge map database system, so that the problem that a general knowledge map database supports multimedia data difficultly is solved. Access authority control among different objects is realized through authority association, and the authority control can be embodied to the attribute level inside the object, so that fine data reading and editing control is realized. The searching method based on reverse index is adopted to realize the searching of object level and object inner level, and the whole-range text searching of millisecond level is realized through the improvement in the algorithm.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (9)
1. A deep knowledge graph-based database system, comprising: a concept segmentation module, an association creation module, and an association computation module, wherein,
a concept segmentation module for segmenting the whole entity objects of the knowledge graph into a plurality of subsets according to the entity object concepts;
the association creating module is used for creating associations among the entity objects according to the plurality of divided subsets and generating an association object set;
and the association calculation module is used for grouping the association object sets according to the association types and performing access calculation on the association object sets by using the independent memory database.
2. The deep knowledge graph-based database system of claim 1, wherein the association calculation module comprises:
the association type submodule is used for appointing the association type between any two entity objects and inquiring and extracting according to different association types in association calculation;
and the association strength submodule is used for depicting the association strength between any two entity objects through the real number of a preset interval.
3. The deep knowledge-graph-based database system of claim 1, wherein the structure of the entity object comprises: base data, attribute tag data, and extended attribute data, wherein,
the basic data comprises identification information, concept classification, name and associated information of the entity object;
attribute tag data is attached to an entity object in a key value pair mode to describe the characteristics of the entity object;
the extended attribute data includes information of all data structures subordinate to the entity object.
4. The deep knowledge-graph-based database system of claim 3, wherein the information of the extended attribute data structure comprises: a data type and a structure container carrying the data type, wherein,
the data types include: a character data type and a binary data type;
the structural container includes: a list structure, a time series structure, and a mapping table structure.
5. The depth-knowledgegraph-based database system of claim 1, further comprising:
the operation function module is used for creating, deleting, modifying and inquiring the concept, the entity object, the attribute tag and the extended attribute;
the extended attribute data operation module is used for performing data adding, modifying, screening, converting, paging and calculating operations on data information in the extended attributes;
the searching module is used for searching the entity objects of the database based on a reverse index mode;
the file management module is used for embedding the multimedia resources into the knowledge graph through a multimedia file management framework;
the file query module is used for extracting, converting and basically operating file data;
the authority management module is used for controlling access authority among different entity objects through a preset authority management mechanism;
and the interface module is used for establishing a database operation interface related to the concept, the entity object and the object user.
6. The deep knowledge graph-based database system according to claim 5, wherein the search module performs the calculation of the reverse index based on the information of the three levels of the entity object within a range obtained by the concept division by using the aggregation function of the database.
7. The deep-knowledge-graph-based database system according to claim 6, wherein the three levels are an object level, an object attribute level, and a unit level within an attribute, respectively.
8. The deep knowledge graph-based database system of claim 5, wherein the rights management mechanism comprises: receiving a request sent by a user object, wherein the request comprises identification information of the user object; extracting a main body and an authority range of the operation according to the identification information of the user object; judging whether the user object has the authority for performing the current operation according to the main body of the operation and the authority range; and after judging that the user object has the authority for performing the current operation, executing the authority and returning an operation result to the user object.
9. The deep knowledge graph-based database system according to claim 8, wherein the authority scope is determined by attribute tag data and extended attribute data stored by the entity object, and comprises: judging whether the user object has the authority for performing the current operation according to the main body and the authority range of the operation, wherein the judgment includes the judgment of whether the concept level has the authority for performing the current operation; when the concept level has the authority to carry out the current operation, executing the authority and returning the operation result to the user object; when the concept level has no authority for performing the current operation, judging whether the object level has the authority for performing the current operation; when the object level has the authority to carry out the current operation, executing the authority and returning the operation result to the user object; and when the object level has no authority for performing the current operation, judging that the user object has no authority for performing the current operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010334955.6A CN111708892B (en) | 2020-04-24 | 2020-04-24 | Database system based on depth knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010334955.6A CN111708892B (en) | 2020-04-24 | 2020-04-24 | Database system based on depth knowledge graph |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111708892A true CN111708892A (en) | 2020-09-25 |
| CN111708892B CN111708892B (en) | 2021-08-03 |
Family
ID=72536741
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010334955.6A Active CN111708892B (en) | 2020-04-24 | 2020-04-24 | Database system based on depth knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111708892B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112988946A (en) * | 2021-05-07 | 2021-06-18 | 泰瑞数创科技(北京)有限公司 | Geographic entity database user customization method |
| CN113111135A (en) * | 2021-04-22 | 2021-07-13 | 北京金山数字娱乐科技有限公司 | Knowledge graph construction method and device |
| CN113296891A (en) * | 2021-05-25 | 2021-08-24 | 和美（深圳）信息技术股份有限公司 | Multi-scene knowledge graph processing method and device based on platform |
| CN113779313A (en) * | 2021-01-12 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Knowledge management method and system based on graph database |
| CN114301725A (en) * | 2021-12-24 | 2022-04-08 | 珠海格力电器股份有限公司 | Device control method, device, electronic device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103488724A (en) * | 2013-09-16 | 2014-01-01 | 复旦大学 | Book-oriented reading field knowledge map construction method |
| CN107220321A (en) * | 2017-05-19 | 2017-09-29 | 重庆邮电大学 | Solid threedimensional embodies in a kind of literary scape conversion method and its system |
| CN107526799A (en) * | 2017-08-18 | 2017-12-29 | 武汉红茶数据技术有限公司 | A kind of knowledge mapping construction method based on deep learning |
| CN107609052A (en) * | 2017-08-23 | 2018-01-19 | 中国科学院软件研究所 | A kind of generation method and device of the domain knowledge collection of illustrative plates based on semantic triangle |
| US20180197088A1 (en) * | 2017-01-10 | 2018-07-12 | International Business Machines Corporation | Discovery, characterization, and analysis of interpersonal relationships extracted from unstructed text data |
| CN108427735A (en) * | 2018-02-28 | 2018-08-21 | 东华大学 | Clinical knowledge map construction method based on electronic health record |
| CN110008353A (en) * | 2019-04-09 | 2019-07-12 | 福建奇点时空数字科技有限公司 | A kind of construction method of dynamic knowledge map |
-
2020
- 2020-04-24 CN CN202010334955.6A patent/CN111708892B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103488724A (en) * | 2013-09-16 | 2014-01-01 | 复旦大学 | Book-oriented reading field knowledge map construction method |
| US20180197088A1 (en) * | 2017-01-10 | 2018-07-12 | International Business Machines Corporation | Discovery, characterization, and analysis of interpersonal relationships extracted from unstructed text data |
| CN107220321A (en) * | 2017-05-19 | 2017-09-29 | 重庆邮电大学 | Solid threedimensional embodies in a kind of literary scape conversion method and its system |
| CN107526799A (en) * | 2017-08-18 | 2017-12-29 | 武汉红茶数据技术有限公司 | A kind of knowledge mapping construction method based on deep learning |
| CN107609052A (en) * | 2017-08-23 | 2018-01-19 | 中国科学院软件研究所 | A kind of generation method and device of the domain knowledge collection of illustrative plates based on semantic triangle |
| CN108427735A (en) * | 2018-02-28 | 2018-08-21 | 东华大学 | Clinical knowledge map construction method based on electronic health record |
| CN110008353A (en) * | 2019-04-09 | 2019-07-12 | 福建奇点时空数字科技有限公司 | A kind of construction method of dynamic knowledge map |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113779313A (en) * | 2021-01-12 | 2021-12-10 | 北京沃东天骏信息技术有限公司 | Knowledge management method and system based on graph database |
| CN113111135A (en) * | 2021-04-22 | 2021-07-13 | 北京金山数字娱乐科技有限公司 | Knowledge graph construction method and device |
| CN112988946A (en) * | 2021-05-07 | 2021-06-18 | 泰瑞数创科技(北京)有限公司 | Geographic entity database user customization method |
| CN112988946B (en) * | 2021-05-07 | 2021-08-31 | 泰瑞数创科技(北京)有限公司 | Geographic entity database user customization method |
| CN113296891A (en) * | 2021-05-25 | 2021-08-24 | 和美（深圳）信息技术股份有限公司 | Multi-scene knowledge graph processing method and device based on platform |
| CN113296891B (en) * | 2021-05-25 | 2024-04-19 | 和美（深圳）信息技术股份有限公司 | Platform-based multi-scene knowledge graph processing method and device |
| CN114301725A (en) * | 2021-12-24 | 2022-04-08 | 珠海格力电器股份有限公司 | Device control method, device, electronic device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111708892B (en) | 2021-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111708892B (en) | Database system based on depth knowledge graph | |
| Krishnaraj et al. | An efficient radix trie‐based semantic visual indexing model for large‐scale image retrieval in cloud environment | |
| US10403052B2 (en) | Mixing virtual image data and physical image data | |
| CN112131449B (en) | Method for realizing cultural resource cascade query interface based on ElasticSearch | |
| US7860852B2 (en) | Systems and apparatuses for seamless integration of user, contextual, and socially aware search utilizing layered approach | |
| CN112116488A (en) | Water conservancy big data comprehensive maintenance system | |
| US20050203943A1 (en) | Personalized classification for browsing documents | |
| CN105808753B (en) | A kind of regionality digital resources system | |
| WO2012070935A1 (en) | Method and system for handling a digital sample | |
| CN111274294B (en) | Universal distributed heterogeneous data integrated logic convergence organization, release and service method and system | |
| CN111552797B (en) | Name prediction model training method and device, electronic equipment and storage medium | |
| US11126682B1 (en) | Hyperlink based multimedia processing | |
| CN105007314B (en) | Towards the big data processing system of magnanimity readers ' reading data | |
| CN106599111A (en) | Data management method and storage system | |
| CN104794150A (en) | Cloud storage model and management method based on space knowledge cloud environment | |
| Herrera-Cubides et al. | Abstraction of linked data’s world | |
| Werbrouck et al. | Data patterns for the organisation of federated linked building data. | |
| KR101683138B1 (en) | Apparatus for searching information, and control method thereof | |
| KR101068676B1 (en) | Ontology modeling apparatus and method for connecting person to resource | |
| CN107291875B (en) | Metadata organization management method and system based on metadata graph | |
| Kashyap | Information brokering over heterogeneous digital data: A metadata-based approach | |
| CN114546982A (en) | Data processing method and device, computer readable medium and electronic equipment | |
| KR102623386B1 (en) | Super meatverse operating system | |
| CN112689157B (en) | Method for processing cloud computing resources and wireless communication device | |
| Beitner | Microcosm++: the development of a loosely coupled object based architecture for open hypermedia systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |