CN112925921B

CN112925921B - DIKW atlas-based resource identification method, related device and readable medium

Info

Publication number: CN112925921B
Application number: CN202110431356.0A
Authority: CN
Inventors: 段玉聪; 胡婷
Original assignee: Hainan University
Current assignee: Hainan University
Priority date: 2021-04-21
Filing date: 2021-04-21
Publication date: 2022-02-22
Anticipated expiration: 2041-04-21
Also published as: CN112925921A

Abstract

The application provides a resource identification method based on a DIKW atlas, a related device and a readable medium, wherein the method comprises the steps of obtaining original resources of a target object and the intention of a user; traversing a resource map in a DIKW system by using the intention of a user and the original resource of a target object, and deducing and determining a plurality of derivation paths associated with the target object from the resource map; wherein, the resource map comprises: data, awareness, information, and knowledge maps; the derivation paths include a plurality of resources interconnected in a resource graph, and each derivation path includes an original resource of the target object and a new resource associated with the target object; new resources associated with the target object are obtained by traversing resource map derivation; because the hidden information in the target object can be deduced through a DIKW system, the resource recognition result of the user analyzed from each deduction path can be more comprehensive and accurate.

Description

DIKW atlas-based resource identification method, related device and readable medium

Technical Field

The application relates to the technical field of computer application, in particular to a resource identification method based on a DIKW atlas, a related device and a readable medium.

Background

With the continuous development of the field of artificial intelligence, the technology of intelligently recognizing information in information transfer tools such as audio, text, images, video and the like has been widely applied in a plurality of fields. For example, intelligently identifying entities in images, intelligently identifying semantics in audio information, and so forth.

However, because human beings have deviation when processing information such as voice, characters, images, etc., parts of interest can be automatically processed and parts of non-interest can be selectively ignored, so that a part of hidden information can be lost in a processing result obtained from information such as voice, characters, images, etc., and furthermore, the intelligent recognition technology designed by human beings can selectively ignore and process parts like a process of human beings processing information, so that the hidden information is lost.

Disclosure of Invention

Based on the above-mentioned shortcomings of the prior art, the present application provides a resource identification method based on a didw atlas, a related apparatus and a readable medium, so as to excavate a new resource related to a target object through a didw system.

To solve the above problems, the following solutions are proposed:

the application discloses a resource identification method based on a DIKW atlas in a first aspect, which comprises the following steps:

acquiring original resources of a target object and an intention of a user; wherein the target object is a tool for transferring information; the original resource of the target object comprises: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with the target object; the user's intention refers to the user's intention to obtain target information from the target object;

traversing a resource map in a DIKW system by using the intention of the user and the original resource of the target object, and deducing and determining a plurality of derivation paths associated with the target object from the resource map; wherein the resource map comprises: data, awareness, information, and knowledge maps; the derivation paths include a plurality of resources interconnected in the resource graph, and each of the derivation paths includes an original resource of the target object and a new resource associated with the target object; new resources associated with the target object are obtained by traversing the resource map derivation;

and analyzing resources meeting the intention of the user from each derivation path, and determining the resources meeting the intention of the user as the resource identification result of the user.

Optionally, in the method for identifying resources based on a didw atlas, the resource atlas further includes a value corresponding to each resource;

after acquiring the original resource of the target object and the intention of the user, the method further includes:

analyzing the intention of the user, and analyzing and calculating a value dimension and an input dimension of the intention of the user; the value dimension of the user's intention is used for explaining the value brought to the user by the accuracy and correctness of the target information acquired by the user from the target object; the input dimension of the user's intention is used for explaining the cost expected by the user to input in order to acquire target information from the target object;

wherein, traversing a resource map in a DIKW system by using the intention of the user and the original resource of the target object, and deducing and determining a plurality of derivation paths associated with the target object from the resource map, comprises:

traversing a resource map in a DIKW system by using the intention of the user, the original resource of the target object, the value dimension of the intention of the user and the input dimension, and deducing and determining a plurality of derivation paths related to the target object from the resource map; and the value corresponding to each resource included in the derivation path meets the requirements of the value dimension and the investment dimension of the user intention.

Optionally, in the method for identifying resources based on a didw graph, the traversing a resource graph in a didw system using the intention of the user and an original resource of the target object, and determining a plurality of derivation paths associated with the target object from the resource graph includes:

traversing a resource map in a DIKW system by using the intention of the user and the original resource of the target object, and deducing a deduction path where the original resource of the target object is located;

performing fusion processing on a plurality of resources in a derivation path where the original resource of the target object is located, and generating a new resource associated with the target object;

for each generated new resource associated with the target object, traversing a resource map in a DIKW system by using the new resource associated with the target object, the intention of the user and the original resource of the target object, and deriving a derivation path where the new resource associated with the target object is located.

Optionally, in the method for identifying resources based on a didw atlas, the fusing multiple resources in a derivation path where an original resource of the target object is located to generate a new resource associated with the target object includes:

fusing a plurality of data resources in a derivation path where the original resource of the target object is located to generate new data resources associated with the target object;

and/or performing fusion processing on the data resource in the derivation path where the original resource of the target object is located and the intention resource to generate a new information resource associated with the target object;

and/or performing fusion processing on the knowledge resource and the data resource in the derivation path where the original resource of the target object is located to generate a new data resource or a new knowledge resource associated with the target object;

and/or performing fusion processing on information resources and knowledge resources in a derivation path where the original resources of the target object are located to generate new knowledge resources associated with the target object;

and/or fusing a plurality of information resources in the derivation path where the original resource of the target object is located to generate a new information resource associated with the target object;

and/or fusing a plurality of knowledge resources in the derivation path where the original resource of the target object is located to generate a new knowledge resource associated with the target object.

Optionally, in the method for identifying resources based on a didw graph, the traversing a resource graph in a didw system by using the intention of the user and the original resource of the target object to derive a derivation path where the original resource of the target object is located includes:

traversing a resource map in a DIKW system by using the intention of the user and the original resource of the target object to find a derivation path where the original resource of the target object is located;

if the data resources are found to be missing in the derivation path of the original resources of the target object in the process of traversing the resource map in the DIKW system, deducing the missing data resources by using a data ontology model, an information logic model or a knowledge meta model;

if the information resource is found to be missing in the derivation path of the original resource of the target object in the process of traversing the resource map in the DIKW system, the missing information resource is derived by using a data ontology model or a knowledge meta model;

if knowledge resources are found to be missing in a derivation path where the original resources of the target object are located in the process of traversing a resource map in a DIKW system, using a body model of data or a logic model of information to derive the missing knowledge resources;

wherein, the traversing a resource map in a DIKW system by using the new resource associated with the target object, the user's intention and the original resource of the target object to derive a derivation path where the new resource associated with the target object is located includes:

traversing a resource map in a DIKW system by using the new resource associated with the target object, the intention of the user and the original resource of the target object to find a derivation path where the new resource associated with the target object is located;

if the data resource is missing in the derivation path of the new resource associated with the target object in the process of traversing the resource map in the DIKW system, deducing the missing data resource by using a data ontology model, an information logic model or a knowledge meta model;

if the information resource is missing in the derivation path of the new resource associated with the target object in the process of traversing the resource map in the DIKW system, deducing the missing information resource by using a data ontology model or a knowledge meta-model;

and if the missing knowledge resource is found in the derivation path of the new resource associated with the target object in the process of traversing the resource map in the DIKW system, using the ontology model of the data or the logic model of the information to derive the missing knowledge resource.

Optionally, in the method for identifying resources based on a didw atlas, if the target object is a target image, the acquiring an original resource of the target object includes:

carrying out entity identification on the target image to obtain entity information of an identified entity; wherein the entity information includes: status, positional relationship and directional relationship information;

converting the entity information of the identified entity into an original resource of the target image; wherein the new resources associated with the target object include new resources of unrecognized entities in the target image.

The second aspect of the present application discloses a resource identification apparatus based on a didw atlas, including:

an acquisition unit configured to acquire an original resource of a target object and an intention of a user; wherein the target object is a tool for transferring information; the original resource of the target object comprises: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with the target object; the user's intention refers to the user's intention to obtain target information from the target object;

the derivation unit is used for traversing a resource map in a DIKW system by using the intention of the user and the original resource of the target object, and deriving and determining a plurality of derivation paths associated with the target object from the resource map; wherein the resource map comprises: data, awareness, information, and knowledge maps; the derivation paths include a plurality of resources interconnected in the resource graph, and each of the derivation paths includes an original resource of the target object and a new resource associated with the target object; new resources associated with the target object are obtained by traversing the resource map derivation;

and the first analysis unit is used for analyzing the resources meeting the intention of the user from each derivation path and determining the resources meeting the intention of the user as the resource identification result of the user.

Optionally, in the resource identification apparatus based on a didw atlas, the resource atlas further includes a value corresponding to each resource;

wherein, resource recognition device based on DIKW atlas still includes:

the second analysis unit is used for analyzing the intention of the user and analyzing and calculating a value dimension and an input dimension of the intention of the user; the value dimension of the user's intention is used for explaining the value brought to the user by the accuracy and correctness of the target information acquired by the user from the target object; the input dimension of the user's intention is used for explaining the cost expected by the user to input in order to acquire target information from the target object;

wherein the derivation unit includes:

a first derivation subunit, configured to traverse a resource graph in a didw system using the intent of the user, the original resource of the target object, a value dimension of the intent of the user, and an input dimension, and derive and determine multiple derivation paths associated with the target object from the resource graph; and the value corresponding to each resource included in the derivation path meets the requirements of the value dimension and the investment dimension of the user intention.

Optionally, in the above apparatus for identifying a resource based on a didw atlas, the derivation unit includes:

the second derivation subunit is configured to traverse a resource map in a didw system by using the intention of the user and the original resource of the target object, and derive a derivation path where the original resource of the target object is located;

a first fusion subunit, configured to perform fusion processing on multiple resources in a derivation path where an original resource of the target object is located, and generate a new resource associated with the target object;

and a third derivation subunit, configured to, for each generated new resource associated with the target object, traverse a resource graph in a didw system using the new resource associated with the target object, the intention of the user, and the original resource of the target object, and derive a derivation path where the new resource associated with the target object is located.

Optionally, in the above apparatus for identifying a resource based on a didw atlas, the first fusion subunit includes:

the second fusion subunit is configured to perform fusion processing on the multiple data resources in the derivation path where the original resource of the target object is located, and generate a new data resource associated with the target object;

Optionally, in the above apparatus for identifying a resource based on a didw atlas, the second derivation subunit includes:

the first traversal subunit is used for traversing a resource map in a DIKW system by using the intention of the user and the original resource of the target object so as to find a derivation path where the original resource of the target object is located;

a fourth derivation subunit, configured to, if a data resource is missing in a derivation path in which an original resource of the target object is located in a process of traversing a resource graph in a didw system, derive the missing data resource using a data ontology model, an information logic model, or a knowledge meta model;

a fifth derivation subunit, configured to, if information resources are missing in a derivation path in which the original resources of the target object are located in the process of traversing a resource graph in a didw system, derive the missing information resources using an ontology model of data or a meta model of knowledge;

a sixth derivation subunit, configured to, if a knowledge resource is missing in a derivation path in which an original resource of the target object is located in a process of traversing a resource graph in a didw system, derive the missing knowledge resource using a body model of data or a logical model of information;

wherein the third derivation subunit includes:

the second traversal subunit is used for traversing the resource map in the DIKW system by using the new resource associated with the target object, the intention of the user and the original resource of the target object to find a derivation path where the new resource associated with the target object is located;

a seventh derivation subunit, configured to, if a data resource is missing in a derivation path where a new resource associated with the target object is located in a process of traversing a resource graph in a didw system, derive the missing data resource using a data ontology model, an information logic model, or a knowledge meta model;

the eighth derivation subunit is configured to, if the information resource is missing in the derivation path where the new resource associated with the target object is located in the process of traversing the resource graph in the didw system, derive the missing information resource using the ontology model of the data or the meta model of the knowledge;

and the ninth derivation subunit is configured to, if a knowledge resource is found to be missing in a derivation path where a new resource associated with the target object is located in the process of traversing the resource graph in the didw system, derive the missing knowledge resource using the ontology model of the data or the logical model of the information.

Optionally, in the resource recognition apparatus based on a didw atlas, if the target object is a target image, the obtaining unit includes:

the identification subunit is used for carrying out entity identification on the target image to obtain entity information of the identified entity; wherein the entity information includes: status, positional relationship and directional relationship information;

a conversion subunit, configured to convert the entity information of the identified entity into an original resource of the target image; wherein the new resources associated with the target object include new resources of unrecognized entities in the target image.

A third aspect of the application discloses a computer readable medium having a computer program stored thereon, wherein the program when executed by a processor implements the method as described in any of the first aspects above.

The fourth aspect of the present application discloses a resource identification device based on a DIKW map, including:

one or more processors;

a storage device having one or more programs stored thereon;

the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method as in any one of the first aspects above.

It can be seen from the foregoing technical solutions that, in the resource identification method based on a didw atlas provided in an embodiment of the present application, original resources of a target object and an intention of a user are obtained, where the target object is a tool for transmitting information, and the original resources of the target object include: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with a target object, the user's intent referring to the user's intent to obtain target information from the target object. In the embodiment of the application, the intention of a user and the original resource of a target object are used for traversing a resource map in a DIKW system, a plurality of derivation paths associated with the target object are derived and determined from the resource map, each derivation path determined by derivation comprises the original resource of the target object and a new resource associated with the target object, the new resource associated with the target object is generated by traversing the resource map derivation, and the resource map comprises: the data map, the consciousness map, the information map and the knowledge map have derivation capacity, new resources related to a target object can be derived and generated, hidden information cannot be omitted selectively like the existing intelligent identification technology, then resources meeting the intention of a user are analyzed out from each derivation path finally, the resources meeting the intention of the user are determined as resource identification results of the user, the obtained resource identification results of the user are more accurate and comprehensive compared with the prior art, and the hidden information meeting the intention of the user cannot be lost.

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, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.

Fig. 1 is a schematic flow chart of a resource identification method based on a didw atlas according to an embodiment of the present application;

fig. 2 is a schematic flowchart of a method for obtaining an original resource of a target object according to an embodiment of the present application;

fig. 3 is a flowchart illustrating a method for determining a derivation path according to an embodiment of the present disclosure;

fig. 4 is a schematic processing flow diagram of a fusion process of data resources and intention resources according to an embodiment of the present application;

fig. 5 is a schematic processing flow diagram of fusion processing performed on knowledge resources and data resources according to an embodiment of the present application;

fig. 6 is a schematic flowchart of another resource identification method based on a didw atlas according to an embodiment of the present application;

fig. 7 is a schematic structural diagram of a resource identification apparatus based on a didw atlas according to an embodiment of the present application.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. 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.

Referring to fig. 1, the embodiment of the application discloses a resource identification method based on a DIKW atlas, which specifically includes the following steps:

s101, acquiring original resources of the target object and the intention of the user.

Wherein the target object is a tool for transferring information. Such as video, audio, images, text, and so forth. And the original resources of the target object comprise: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with the target object. The original resources of the target object are some resources extracted from the target object to describe the target object. And specifically may include at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with the target object. And the user's intention refers to the user's intention to acquire target information from the target object. For example, if the target object is a target image, the user's intent may be to identify the entities appearing in the target image, or to identify which containers are in the target image. For another example, if the target object is a consumption record text, the user's intention may be to estimate the gender and age of the consuming user in the consumption record text. Also for example, if the target object is a video, the user's intent may be to identify the category to which the video belongs.

Data resources are basic individual items of digital or other types of information obtained by observing a target object, which have no meaning in themselves without a contextual context. The data resource is mainly used for expressing the attribute of the entity in the target object, and specifically may include the attribute content and frequency of the entity in the target object. Where the frequency may be the frequency at which the entity occurs in different structures, time and space. Attributes of an entity may be the positional relationship, structural relationship, basic structure, etc. of the record entity. The data resource may further include the number of interactions (i.e., frequency count) between entities in different structures, time and spaces, and the ratio of the interaction frequency count of a group of entities to the sum of all the interaction frequency counts is the interaction frequency of the group of entities.

It is intended that the resources be clearly aware of the need for the goals and methods to be achieved, and that there be a need, i.e., an expression goal, in nature. The user' S intention acquired in step S101 may be analyzed and converted into an intention resource.

An information resource is a state obtained by assigning a specific purpose to an entity, and it can be simply understood that the information resource is a data addition purpose (intention). In the case of known information resources and corresponding data resources, the corresponding intent resources can be deduced, whereas in the case of known information resources and purposes, the corresponding data resources can be deduced.

Knowledge resources are built based on probabilistic computing or inductive, deductive or causal reasoning, describing the integrity abstract relationships that exist between content at the type/class level, which can be summarized as a piece of knowledge rules. Knowledge resources can be deduced from data resources and information resources through structured and formal deduction. Knowledge resources can define two semantic relationships between entities, namely an identical relationship and an opposite relationship, wherein the identical relationship means that in a certain scene, two entities are exactly identical in a certain attribute or relationship. An inverse relationship means that in a certain scenario, two entities are exactly opposite in a certain attribute or relationship. For example, apples and pears can both obtain their relationship to the ground through the law of universal gravitation, and the vehicles in the east-west direction and the vehicles in the north-south direction at the intersection have opposite movement states. Hidden information which cannot be observed when the target object is observed can be deduced through the semantic relation in the knowledge resources.

In the embodiment of the present application, the original resource of the target object refers to a resource that has not undergone derivation processing by the resource map in step S102. The original resources of the target object can be observed manually or can be obtained by identifying the target object by using an artificial intelligent identification technology. The original resources of the target object can be obtained in many ways, including but not limited to what is proposed in the embodiments of the present application.

Optionally, referring to fig. 2, in an embodiment of the present application, if the target object is a target image, an implementation of obtaining an original resource of the target object is performed, where the implementation includes:

s201, carrying out entity identification on the target image to obtain entity information of the identified entity, wherein the entity information comprises: status, positional relationship, and directional relationship information.

In this embodiment of the application, the obtained intention of the user may be to identify an entity in the target image. In step S201, before the resource map is not used, the entity in the target image is identified by observation or artificial intelligence identification, so as to obtain entity information of the identified entity.

The entity information of the identified entity includes status information (i.e., stationary or moving) of the identified entity, a location relationship between the identified entity and other entities (the other entities include other identified entities except the identified entity and unidentified entities), and direction relationship information between the identified entity and other entities.

The identified entity refers to an entity that has identified entity information, and the unrecognized entity refers to an entity that has not completely identified entity information, or an entity that has not identified entity information.

S202, converting the entity information of the identified entity into the original resource of the target image.

Wherein the new resources associated with the target object include new resources of unrecognized entities in the target image. An original source of a target image, comprising: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with the target image.

Attribute information in the entity information of the identified entity regarding the frequency of interaction between the identified entity and other entities, as well as the state, structure, etc. of the identified entity may be converted into a data resource associated with the target image. And the interaction relationship between the identified entity and other entities may be converted into an information resource associated with the target image, and then the entity information of the identified entity is converted into an original resource of the target image for performing step S102 shown in fig. 1, and then the new resource associated with the target object included in the obtained derived path is a new resource of the entity that is not identified in the target image. Since the user intends to identify the entity in the target image, after the entity information of the identified entity is converted into the original resource of the target image for performing step S102 shown in fig. 1, a new resource of the unrecognized entity is obtained by deriving the path, and what the unrecognized entity is can be determined by the new resource of the unrecognized entity.

S102, traversing a resource map in a DIKW system by using the intention of a user and the original resource of a target object, and deducing and determining a plurality of derivation paths associated with the target object from the resource map, wherein the resource map comprises: the system comprises a data graph, an awareness graph, an information graph and a knowledge graph, wherein the derivation paths comprise a plurality of resources which are mutually connected in the resource graph, each derivation path comprises an original resource of a target object and a new resource which is associated with the target object, and the new resource which is associated with the target object is obtained by traversing the resource graph and deriving and generating.

And traversing a resource map in a system related to data information knowledge and intelligence (DIKW) by using the original resource of the target object and the intention of the user, and deducing according to the deduction direction of the resource which can meet the intention of the user from the original resource of the target object through the resource map to obtain a plurality of deduction paths associated with the target object. Each derivation path obtained by traversing the resource graph in step S102 includes an original resource of the target object and a new resource associated with the target object, that is, of all resources included in the obtained derivation path, both the original resource of the target object and the new resource associated with the target object exist. The partially derived path may contain only the original resources, while the partially derived path may contain both the original resources and the new resources, or the partially derived path may contain only the new resources.

The new resource associated with the target object refers to a resource which is not originally contained in the resource map and is generated by performing calculation derivation on the original resource and the intention of the user. Because the resource map comprises the data map, the consciousness map, the information map and the knowledge map, the derivation capability is provided, the hidden information which is not discovered originally in the target object can be mined, and the mined hidden information is embodied in the new resource which is included in the derivation path and is related to the target object.

The data map in the resource map includes a plurality of data resources, if the original resources acquired in step S101 include data resources associated with the target object, the data resources may be used to traverse the data map for searching, and if the data resources can be found, the mapping relationship (or connection relationship) between the data resources and other resources in the resource map may be used to obtain the derivation path. And if the data is not found, storing the data into the data map, and deriving a derivation path by using the model in the resource map through derivation. Specifically, the data map may record a basic structure of the data entity, and record a position relationship and a structural relationship between child nodes of the data entity. However, the data map can only be used for static analysis, and cannot express the interaction relationship between data entities. The data map may also record the frequency of occurrence of the structure included in the data entity, specifically including the frequency of the three levels of structure, time and space. When a data map in a DIKW system is constructed, after a body system of data is established according to a plurality of data resources, the interaction times between entities distributed in the system in different structures, time and spaces are called frequency counts, and the ratio of the frequency counts of each group to the sum of all the frequency counts is called frequency or proportion. The sum of all the complete interaction paths between entity 1 and entity 2 entities is the interaction frequency of entity 1 and entity 2. When the interaction frequency between the two entities is lower than a threshold value, the indirect interaction relationship is regarded to exist, and otherwise, the direct interaction relationship is regarded to exist. And then data cleaning is carried out on an information map in a DIKW system, redundant data is eliminated, and only the direct interaction relation between entities can be recorded.

Optionally, an ontology model of the data may be included in the didw system. The ontology model of the data adopts an inheritance structure relationship of an object-oriented structure, and the inheritance structure relationship is a parent-child inclusion relationship of the object. Each object has a specific inclusion relationship, and there is a phenomenon of inheritance property, characteristic and the like between objects having the inclusion relationship, and the phenomenon can also be understood as object-oriented parent-child relationship.

The ontology model of the data can be represented in different structures, such as a tree, a graph, and the like, that is, the concrete representation structure of the ontology model of the data is not limited as long as the parent-child inclusion relationship of the object can be clearly represented. For example, if a data ontology model is represented by a directed graph, specific meanings represented by arrow directions should be defined, and the arrow directions in the directed graph are included. Ontology models of object-oriented structured data typically automatically extract concepts, concept hierarchies, and relationships between concepts from data graphs, information graphs, and knowledge graphs in a bottom-up approach. The abstract concept is to build a data ontology model with a layer-by-layer recursive structure. The data resources in the data map can establish an ontology model of the corresponding data according to different intentions. For example, if the cup is used for drinking water and the intention is to quench thirst, the body model of the data corresponding to the cup is a container, the main attribute of the container is closed, the intention is possibly threatened, the cup is used for attacking the other side, and the corresponding body model is a weapon, and the main attribute is hardness.

The intention map includes a plurality of intention resources, the intention of the user acquired in step S101 and, if the original resource includes an intention resource associated with the target object, the intention resource and the intention of the user may be used to search through the intention map, and if the intention resource can be found, the mapping relationship (or the connection relationship) between the intention resource and other resources in the resource map may be used to obtain the derivation path. And if not, storing the data into the intention map, and using the model in the resource map to obtain a derivation path through derivation. The intention graph can record the logic relationship between intentions, including child intentions and composition intentions, clearly express the containment relationship existing between different intentions through the hierarchical expression form of a tree structure, a parent node represents the semantic coverage range and the composition intentions with larger interpretable range, and a child node connected below the parent node represents a more specific child intention.

Optionally, an intention model may also be included in the DIKW system. The intention includes sub-intents, component intents and inclusion intents, and an exclusion relationship exists among the sub-intents, the component intents and the inclusion intents, wherein one of the sub-intents is selected more, and two of the sub-intents and the component intents are selected more. Further, the deep-level connotation of intent is related to value. The intention model clearly expresses the existing relation among different intentions through a hierarchical expression form of a tree structure, a parent node expresses a semantic coverage range and a composition intention with a larger interpretable range, and a child node connected below the parent node expresses a more specific child intention. The parent node and the child node are connected by a straight line without an arrow or a straight line with an arrow pointing from the child node to the parent node, and the direction of pointing of the arrow indicates 'yes'.

The information map includes a plurality of information resources, and if the information resources associated with the target object are included in the original resources obtained in step S101, the information resources may be used to search through the information map, and if the information resources can be searched, the mapping relationship (or connection relationship) between the information resources and other resources in the resource map may be used to obtain the derivation path. And if the information is not found, storing the information into the information map, and using the model in the resource map to obtain a derivation path through derivation. Specifically, the information map may record the interaction relationship between the entities, including direct interaction relationship and indirect interaction relationship, or may express the interaction relationship between the entities by using a multi-component group. Generally, the more elements contained in the tuple, the more accurate the obtained interaction relationship between the entities, but as the number of the elements increases, the atlas may have different judgments on the interaction relationship, so that the judgment priority needs to be determined by more examples, so as to derive the most correct interaction relationship.

Alternatively, the relationship between the entities recorded in the information graph may be a directional relationship and a topological relationship (or a positional relationship). Topological relationships between entities are defined as those relationships that are invariant under topological transformations, including rotation, scaling, and translation. The topological relationship between two entities is: intersecting, separating, tangent, containing, overlapping, and overlapping. Each element in the information graph represents different topological relations and describes a series of parent-child relations in the position relations, and the topological relations are inherited from a data ontology model in the data graph and can represent the proximity and the relevance between the entities.

Optionally, a logical model of the information may also be included in the DIKW architecture. Logical modeling of information is essentially a logical relationship model approach to linking and passing inferences between information. The information is related by the basic association methods including, relating, and/or not, and these association methods together form the hierarchical relationship of the information network structure.

The knowledge graph includes a plurality of knowledge resources, in step S101, if the original resources include knowledge resources associated with the target object, the knowledge resources may be searched in the traversal information graph, and if the knowledge resources may be searched, the mapping relationship (or connection relationship) between the knowledge resources and other resources in the resource graph may be used to obtain the derivation path. And if not, storing the data into the knowledge graph, and using the model in the resource graph to obtain a derivation path through derivation. The knowledge graph is summarized with a plurality of rules, and the rules are mainly used for defining two semantic relations between entities, namely an identical relation and an opposite relation. These two relationships mean that in a certain scenario, two entities are exactly the same or opposite in a certain attribute or relationship. For example, apples and pears can both get their relationship to the ground through the law of universal gravitation, and the movement states of vehicles in the east-west direction and vehicles in the north-south direction at the intersection are opposite. The hidden information (namely the new resources associated with the target object) is found on the knowledge graph through knowledge reasoning by utilizing the two basic semantic relations.

Optionally, meta-models of knowledge may also be included in the DIKW architecture. Knowledge is defined to be derivable from data resources and information resources via a structured, formalized deductive, built based on probabilistic computing or inductive, deductive or reasoning, describing the existence of an integrity abstract relationship between content at the type/class level, which can be summarized as a knowledge rule. Explicit knowledge is objective knowledge, can be expressed in language, expressed in words and numbers, easily collated, stored, and converted into information and encoded, communicated and shared in data form, also referred to as expressible knowledge or encodable knowledge. Knowledge engineering is a technology for acquiring knowledge and information with high efficiency and large capacity by using modern scientific and technical means according to specified knowledge rules. The basic meta-model of knowledge is just an indispensable component, and the basic meta-models are connected and integrated through specific rules to form a complete knowledge engineering, so that the ability of efficiently identifying and acquiring information is realized.

It should be noted that the didw system in the embodiment of the present application may be obtained by pre-constructing resources of a plurality of target objects, and in the process of executing the embodiment shown in fig. 1, a derivation path is determined by continuous derivation, and the constructed didw system is also continuously updated, so that the derivation function of the didw system is further improved. For example, extracting image concepts may be done in advance from the image dataset by processing the data and attribute information. From the information of the image dataset, concepts of image object classes (e.g. "highway" and "transportation warehouse") are constructed and connected to the ontology of data. Then, in each layer of the ontology model of the data, the attribute field information of the image relationship data is used to extract subclass concepts of the entity object, so that semantic relationships between the concepts extracted from the information can be enriched. For example, categories include "train station" and "bus station," and these subcategories belong to the concept of "transportation warehouse. Meanwhile, entities can be extracted from the image data set. The image entity extraction rule based on the hierarchy and attribute domain information is provided aiming at the phenomenon that an image entity is divided into a plurality of features and abstracted through a point, line, surface and geometric set. The entities in the image dataset are stored in a data table of a pre-constructed spatial location database. Wherein each row of the data table includes a normal field and a geometric field. Each row of the data table corresponds to one entity, the common field of each row represents the attribute of the entity corresponding to the row, and the geometric field of each row represents the position of the entity corresponding to the row in the image. Namely, an ontology model and a data map of data in a DIKW system are constructed by extracting image concepts and entities from an image dataset.

Optionally, in order to extract all relevant equivalence relations between entities of the image dataset and entities in the database of the search engine, an encyclopedia is opened and searched using a searcher to find out which image dataset entities' hundreds of entities should be linked. If the searcher retrieves a specified web page containing tags and information boxes describing geographic entities, the same relationship will be established to link these entities.

The DIKW system is constructed by using resources related to various target objects, and simultaneously, the resources with derivation association relations in each layer of atlas are connected in the DIKW system, so that the construction of the DIKW system is completed. Specifically, the data resource plus the intention resource obtains a corresponding information resource, and the data resource plus the information resource deduces a corresponding knowledge resource.

In the process of executing step S102, the resource map in the DIKW system is traversed using the user 'S intention and the original resource of the target object, and a plurality of derivation paths associated with the user' S intention and the original resource of the target object in the resource map are found. The derivation path includes a plurality of resources connected to each other in the resource map, a derivation association relationship exists between the plurality of resources connected to each other, and the resources included in all the determined derivation paths include not only the original resources in step S101, but also resources related to the user 'S intention, new resources related to the target object, which are derived after traversing the resource map using the user' S intention and the original resources of the target object, and possibly existing resources in the DIKW system.

It should be noted that the new resource associated with the target object is a newly generated resource that originally does not exist in the didw system, and is derived from the original resource and the user's intention by traversing the resource map. The original resources in step S101 can be obtained by human observation or artificial intelligence recognition technology. However, because human beings have deviation when processing information such as voice, characters, images, etc., parts of interest can be automatically processed and parts of non-interest can be selectively ignored, so that a part of hidden information can be lost in a processing result obtained from information such as voice, characters, images, etc., and furthermore, the intelligent recognition technology designed by human beings can selectively ignore and process parts like a process of human beings processing information, so that the hidden information is lost.

In the embodiment of the present application, the resource map included in the DIKW system may be derived to find multiple derived paths associated with the target object, and new resources included in the derived paths and associated with the target object have some hidden information lost by human beings during the processing. Therefore, in the embodiment of the application, a plurality of derivation paths which are derived through the resource map in the DIKW system and are associated with the target object are used for mining the omitted hidden information, so that more and more complete resources meeting the intention of the user can be found.

Optionally, in a specific embodiment of the present application, the resource graph further includes a value corresponding to each resource, where after the step S101 is executed, the method further includes:

analyzing the intention of the user, and analyzing and calculating the value dimension and the input dimension of the intention of the user. The value dimension of the user's intention is used for explaining the value brought to the user by the accuracy and the correctness of the target information acquired from the target object by the user, and the investment dimension of the user's intention is used for explaining the cost expected to be invested by the user in order to acquire the target information from the target object. Wherein, executing an implementation manner of step S102 includes: traversing a resource map in a DIKW system by using the intention of the user, the original resources of the target object, the value dimension and the input dimension of the intention of the user, and deducing and determining a plurality of derivation paths associated with the target object from the resource map, wherein the value corresponding to each resource included in the derivation paths meets the requirements of the value dimension and the input dimension of the intention of the user.

The value corresponding to the data resource is defined as the frequency of the data entity appearing in different structures, time and spaces, and the higher the value corresponding to the data resource is, the more accurate the resource meeting the user's intention finally obtained by deriving the data resource is (i.e. the more accurate the derivation result is). The calculation formula of the value corresponding to the data resource may be defined as f (ed) ═ f (edstr) + f (edtime) + f (edspace). Where f (ed) identifies the value of the data entity, f (edstr) represents the frequency of occurrence of the data entity in different structures, f (edtime) represents the frequency of occurrence of the data entity in different times, and f (edspace) represents the frequency of occurrence of the data entity in different spaces. Where f (edstr) ═ f (EDSon1) + f (EDSon2) + … + f (EDSon), { EDSon1, EDSon2, …, EDSon } is the N subclasses of the inherited data entity ED, and where f (EDSon1) ═ f (EDSon1-son1) + f (EDSon1-son2) + … + f (EDSon1-sonM), { EDSon1-son1, EDSon1-son2, …, and EDSon1-sonM } are the M subclasses of the inherited data entity EDSon1, f (EDSon2), …, f (EDSon) may also be subdivided as above until the lowest level relationship is not reachable. Wherein the relationship between sibling subclasses such as EDSon1 and EDSon2 is not repeatable.

The value corresponding to the information resource may include an information length value determined based on the information length in the information graph and an information breadth value determined based on the information breadth in the information graph. The information length refers to the length of data intention association condition obtained through longitudinal association analysis of the data entities and the intention in the information and the frequency of the data appearing in different structures, time and spaces, and the information breadth refers to the breadth of data intention association condition obtained through transverse association analysis of the data entities and the intention in the information. The value corresponding to the information resource can be obtained by evaluating the length of the track information. There are many analysis means for the corresponding value of the information resource, including but not limited to those provided in the embodiments of the present application. Similarly, the larger the corresponding value of the information resource is, the more accurate the resource meeting the user's intention finally obtained by deriving using the information resource is (i.e. the more accurate the derivation result is).

The value corresponding to the knowledge is used for reflecting the accuracy of a specific reasoning result, and the possibility or relevance of the accuracy of the predicted reasoning result can be judged according to the value.

For example, in a special case, the gender of the person is confidential data and cannot be directly obtained. But specific consumption data and consumption record data of a specific person can be obtained, the consumption data is subjected to information fusion to obtain specific information, namely that the male proportion of buying cigarettes is 80% ", and the value of the specific information for judging the sex of the specific person to be male is ValueI 1. Assuming that the acquiring specific person purchases a pack of cigarettes, the value of the consumption information for determining that the gender of the specific person is male is valueii 2, and assuming that the acquiring specific person purchases a male product, the value of the consumption information for determining that the gender of the specific person is male is valueii 3. It is readily determined by common sense that a particular person who purchases both cigarettes and a male product is more likely to be a male. However, if a large number of cigarettes are purchased at one time, the possibility that the specific purchaser is male is rather reduced, or if the frequency of purchase deviates from the frequency of use of the article to be too great, the possibility that the specific purchaser is male is also reduced.

And performing fusion processing by using values corresponding to the data, the information and the knowledge to obtain a weighted value.

And the value dimension of the user's intention refers to the value that the accuracy and correctness of the target information obtained from the target object by the user bring to the user. Wherein the accuracy is determined by a value driven model and the correctness is determined according to the confidence level. The precision of the obtained data needs to be set and adjusted when the required data resources are processed. If the result requires a higher accuracy for the user's intended needs, resources with higher value are required. If this data resource is processed, the accuracy of the resulting data is driven by the fact that higher values are not needed, but only by the approximate need for the corresponding data results. At that time, the value requirements of the resources involved in the derivation path can be reduced. Correctness depends on the confidence level. The confidence interval exhibits the extent to which the true value of the parameter has a certain probability of falling around the measurement. The confidence interval indicates the confidence with which the measured value of the measured parameter is plausible. Confidence is required when defining and determining the correctness of the position.

By analyzing the intention of the user, the value dimension of the intention of the user is analyzed and calculated, and then the precision and the confidence coefficient in the derivation process by using the resource map can be set. The set precision and confidence determine how much valuable resources are needed to be used for derivation in the derivation process.

By analyzing the intention of the user, the input dimension of the intention of the user is analyzed and calculated to be used for explaining the cost expected by the user to input in order to obtain the target information from the target object. The cost (time and money) invested by the user is in a proportional function relation with the obtained target information, namely the more the time and money are invested, the higher the correctness and precision of the obtained target information content are, and vice versa.

After the value dimension and the input dimension of the user intention are obtained through intention analysis, when step S102 is executed, the user intention, the original resource of the target object, the value dimension and the input dimension of the user intention are used, a resource map in a didw system is traversed, a plurality of derivation paths associated with the target object are derived and determined from the resource map, and the value corresponding to each resource included in the derivation paths meets the requirements of the value dimension and the input dimension of the user intention, so that the resource identification result of the user finally obtained in step S103 can meet the requirements of the value dimension and the input dimension of the user intention.

Optionally, in a specific embodiment of the present application, the intention of the user may be analyzed by using an intention model, and then the value dimension and the investment dimension of the intention of the user may be calculated by using an analysis result obtained after the intention model is analyzed.

Alternatively, in a specific embodiment of the present application, cost calculation may be performed before constructing the DIKW system mentioned in step S102. The scene of a DIKW system needing to be used is analyzed and predicted, a scheme for building a corresponding DIKW system model is put forward, and then the cost required for building the DIKW system is estimated, wherein the cost comprises time cost, technical cost, capital cost, intelligence cost, physical cost and the like. And then, according to the cost which can be actually provided, the construction scheme of the DIKW system is adjusted, and the DIKW system meeting the cost requirement is constructed according to the finally adjusted scheme.

Optionally, referring to fig. 3, in an embodiment of the present application, an implementation of step S102 is performed, including:

s301, traversing a resource map in a DIKW system by using the intention of the user and the original resource of the target object, and deducing a deduction path where the original resource of the target object is located.

Specifically, the method includes traversing a DIKW system by using the intention of a user and the original resource of a target object, finding out a derivation path where the original resource is located, and deriving the original resource by using the intention of the user and various models to obtain a complete derivation path where the original resource is located, wherein the complete derivation path is formed by connecting the resources of each layer of the atlas.

It should be noted that, in the process of deriving using the intention of the user and the original resource of the target object, the original resource in the DIKW system may be used to form a complete derivation path where the original resource of the target object is located, or the model may be used to derive and generate a complete derivation path where the original resource of the target object is located by using a new resource associated with the target object that does not exist in the DIKW system.

Optionally, in a specific embodiment of the present application, an implementation manner of executing step S301 includes:

and traversing the resource map in the DIKW system by using the intention of the user and the original resource of the target object to find the derivation path of the original resource of the target object. If the data resource is missing in the derivation path of the original resource of the target object in the process of traversing the resource map in the DIKW system, the missing data resource is derived by using a data ontology model, an information logic model or a knowledge meta model. If the information resource is missing in the derivation path where the original resource of the target object is found in the process of traversing the resource map in the DIKW system, the missing information resource is derived by using the ontology model of the data or the meta model of the knowledge. If the knowledge resources are missing in the derivation path of the original resources of the target object in the process of traversing the resource map in the DIKW system, the missing knowledge resources are derived by using the ontology model of the data or the logic model of the information.

Specifically, in the process of traversing the resource map in the DIKW system by using the intention of the user and the original resource of the target object to find the derivation path where the original resource of the target object is located, a situation that the derivation path where the original resource of the target object is located cannot be formed by using the original resource in the DIKW system may occur. Specifically, it may be found that data resources, information resources, or knowledge resources are missing in the derivation path where the original resources of the target object are located.

And in the process of traversing a resource map in a DIKW system, data resources are found to be missing in a derivation path where original resources of a target object are located, and the data resources can be derived by using a data ontology model. For example, most students in a class are 15 years old, and only the xiaoming is 17 years old, then the xiaoming can be uniquely inferred according to the age attribute. Alternatively, the data resources may also be derived using a logical model of the information. Specifically, since the data resource and the intention model are combined to obtain the logical model of the information, the data resource can be obtained by separating the logical model of the information from the intention model. Alternatively, the data resources may also be derived using meta-models of knowledge. If the meta-model of knowledge is derived from the data ontology model where the missing data resources are located, the missing data resources can be obtained by performing reverse-reasoning on the meta-model of knowledge. Future data resources may also be inferred from meta-models of already existing knowledge. Specifically, the following results are obtained through statistical analysis: an adult of 175 cm typically has a two step separation of 70cm, a 170 cm component has a two step separation of 65cm and a 180 cm component of 75 cm. Now that the step interval of the small and clear is known to be 67cm, the approximate height of the small and clear is known to be about 172 cm, and the missing data resource is deduced.

If the information resource is missing in the derivation path where the original resource of the target object is found in the process of traversing the resource map in the DIKW system, the missing information resource is derived by using the ontology model of the data or the meta model of the knowledge. Specifically, if the generation of the local model of the data is related to the intent, the ontology model of the data and one or more specific intents are linked to reflect the motivation of the behavior or the information resource, so that the missing information resource can be deduced through the ontology model of the data. For example: the teacher lets students close the door, pull the curtain, turn off the light, then open the projector, can deduce that this class is a film class. And the next behavior intention of the entity can be conjectured according to the existing knowledge meta-model of the abstract relation between the entities, so that the missing information resource can be obtained. For example: xiaoming drinks water after each exercise, and Bob is supposed to select mineral water for drinks and mineral water after running.

If the knowledge resources are missing in the derivation path of the original resources of the target object in the process of traversing the resource map in the DIKW system, the missing knowledge resources are derived by using the ontology model of the data or the logic model of the information. After the ontology model of the data reaches a certain scale, the knowledge resources can be generated in a reasoning way, the overall development or change trend of things can be expressed, and the future behavior of the entity can be presumed. For example: xiaoming fails to reach the excellent English achievement in the science achievement with the first grade and the second grade, and can learn that a great amount of time is spent in learning English in other days. And the logical model of the information can be deduced to obtain knowledge resources through abstract analysis. For example: the Xiaoming can be a person loving life by taking regular gymnasium exercise every week and enjoying fruit, not loving to stay up all night and other information.

And deducing and generating the missing new resources (namely the new resources associated with the target object) in the deducing path of the original resources through the ontology model of the data, the logic model of the information or the meta model of the knowledge to obtain the complete deducing path of the original resources.

S302, fusing a plurality of resources in the derivation path where the original resource of the target object is located, and generating a new resource associated with the target object.

If the resource identification result meeting the intention of the user cannot be inferred by the resource included in the derivation path in which the original resource of the target object is located, the multiple resources in the derivation path in which the original resource of the target object is located can be subjected to fusion processing to generate a new resource associated with the target object, and then the new resource associated with the target object is continuously used for derivation to attempt to derive the resource identification result meeting the intention of the user.

Optionally, in a specific embodiment of the present application, an implementation manner of executing step S302 includes:

fusing a plurality of data resources in a derivation path in which an original resource of a target object is positioned to generate a new data resource associated with the target object, and/or fusing a data resource in the derivation path in which the original resource of the target object is positioned and an intention resource to generate a new information resource associated with the target object, and/or fusing a knowledge resource and a data resource in the derivation path in which the original resource of the target object is positioned to generate a new data resource or a new knowledge resource associated with the target object, and/or fusing an information resource and a knowledge resource in the derivation path in which the original resource of the target object is positioned to generate a new knowledge resource associated with the target object, and/or fusing a plurality of information resources in the derivation path in which the original resource of the target object is positioned, and generating a new information resource associated with the target object, and/or fusing a plurality of knowledge resources in the derivation path where the original resource of the target object is located to generate the new knowledge resource associated with the target object.

Specifically, the process of fusing a plurality of data resources in the derivation path where the original resource of the target object is located into a new data resource associated with the target object includes: because the relational data resources refer to data types for describing external relations among data entities, parent-child inheritance relationship exists among the relational data resources, while the discrete data resources have no specific semantics, but the parent-child inheritance relationship also exists among the data entities. Therefore, the data resources can be fused with each other, and the discrete data resources and the relational data resources can be combined to generate new data resources associated with the target object.

The data resource in the derivation path where the original resource of the target object is located is fused with the specific intention resource, and a new information resource associated with the corresponding target object can be generated in a combined manner. For example, referring to fig. 4, the ontology model of the data includes four entities, apple, strawberry, cherry tomato, and wax apple, belonging to the fruit category. And the intention model connected with the ontology model of the data has the intention of identifying whether the fruit is red or not and whether the fruit is sweet or not. The purpose of judging whether the fruit is red or not is to judge whether the fruit is bad or not, and the purpose of judging whether the fruit is hard or not is to judge whether the fruit is large or not. Finally, data resources in the ontology model of the data and intention resources in the intention model are used to fuse a logic model of the information, and new information resources deduced from the logic model of the information indicate that the fruit is red if the fruit is not bad, large if the fruit is not hard, and sweet if the fruit is large and red.

And performing fusion processing on the knowledge resource and the data resource in the derivation path where the original resource of the target object is located, so as to generate a new data resource or a new knowledge resource associated with the target object. For example, a distance between two steps of an adult of 175 cm is generally 70cm, a distance between 170 cm and 180 cm is generally 65cm, a distance between 180 cm and a step of a small degree is generally 67cm, and a height of the small degree is approximately 172 cm according to knowledge resources obtained through statistical analysis, so that new data resources are obtained. For another example, referring to FIG. 5, the meta-model of knowledge provides the knowledge resources needed for life to intake energy, which is supplied from carbohydrates. The ontology model of the data shows that lotus root, sweet potato and rice all belong to carbohydrates. Through the fusion of data resources and knowledge resources in the ontology model of the data, the derived new knowledge resources are: the tee ingests energy by eating rice.

And fusing the information resources in the derivation path where the original resources of the target object are located with the existing knowledge resources to generate new knowledge resources which can be predicted to develop and are associated with the target object. For example, knowledge resource K1: the south region is 6-7 months in plum rain season, the information resource I is that the south region is in Hainan, and a new knowledge resource K is obtained by combining K1 and I: hainan often rains in 6-7 months.

And performing fusion processing on the information resources and the information resources in the derivation path where the original resources of the target object are located, so as to generate new information resources associated with the target object. For example, the known information resource I1: xiaoming liking drinking, information resource I2: xiaoming likes sports. New information resource I3 was generated by combining I1 and I2: xiaoming likes drinking water after exercise.

A plurality of knowledge resources in the derivation path where the original resource of the target object is located can be fused with each other to generate a new knowledge resource. For example, knowledge resource K1: mouse has a habit of grinding teeth, knowledge resource K2: the mouse incisors can grow for a lifetime. The researchers summarized the new knowledge resource K3 according to experimental observation by fusing K1 and K2: the mouse molars are used for obtaining survival and ensuring growth balance.

And S303, for each generated new resource associated with the target object, traversing the resource map in the DIKW system by using the new resource associated with the target object, the intention of the user and the original resource of the target object, and deducing a derivation path where the new resource associated with the target object is located.

Optionally, an implementation of step S303 is performed, including:

and traversing the resource map in the DIKW system by using the new resource associated with the target object, the intention of the user and the original resource of the target object to find a derivation path where the new resource associated with the target object is located, and if the data resource is found to be missing in the derivation path where the new resource associated with the target object is located in the process of traversing the resource map in the DIKW system, deducing the missing data resource by using a data ontology model, an information logic model or a knowledge meta model. If the information resource is missing in the derivation path where the new resource associated with the target object is located in the process of traversing the resource map in the DIKW system, the missing information resource is derived by using the ontology model of the data or the meta model of the knowledge. And if the knowledge resources are missing in the derivation path of the new resources associated with the target object in the process of traversing the resource map in the DIKW system, deducing the missing knowledge resources by using the ontology model of the data or the logic model of the information.

The specific principle and process of deriving the path in step S303 are similar to those in step S301, and reference is made to this, which is not described herein again.

S103, analyzing the resources meeting the intention of the user from each derivation path, and determining the resources meeting the intention of the user as the resource identification result of the user.

In all the determined derivation paths associated with the target object, resources which can meet the intention of the user and can acquire target information from the target object exist, the resources which meet the intention of the user can be analyzed by analyzing the resources included in each derivation path, and then the resources which meet the intention of the user are determined as the resource identification result of the user.

All the derivation paths include resources which are new resources associated with the target object besides the original resources. The new resources associated with the target object are ignored by the artificial intelligence technology in the prior art (namely, hidden information), and the derived path derived by using a DIKW system is mined in the embodiment of the application, so that the hidden information cannot be lost in the resource identification result of the user determined in the embodiment of the application finally, and the identification result is more accurate and comprehensive compared with the existing identification result obtained by artificial observation or artificial intelligence identification.

Referring to fig. 6, the specific resource identification process based on the embodiment shown in fig. 1 is as follows: when the target object is a target image and the intention of the user is to identify all entities from the target image, a DIKW system is constructed by using data resources, information resources, intention resources and knowledge resources of a plurality of images in advance. The DIKW system comprises a resource map, and the resource map comprises a data map, a consciousness map, an information map and a knowledge map. And then observing the target image, acquiring entity information of the entity in the target image, and dividing the entity in the target image into an identified entity and an unidentified entity. The entity information comprises entity states, relationship positions among the entities and direction relationships among the entities. An unidentified entity refers to an entity for which the complete entity information for the entity has not yet been specified. Converting the obtained entity information into an original resource of a target image, wherein the original resource of the target image comprises: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with the target image. And classifying the data resources related to the target image to obtain discrete data resources and relational data resources. The method comprises the steps of traversing a resource map by using the intention of a user and the original resource of a target image, specifically, traversing an information map according to the information resource associated with an identified entity, deducing a derivation path where the information resource associated with the identified entity is located, obtaining relationship information among entities in the target image by using a knowledge map, finding an unidentified entity having the same or opposite relationship with the identified entity, deducing a corresponding derivation path in the knowledge map based on the state, position relationship and direction relationship of the identified entity, and traversing a data map to obtain entity information of the unidentified entity. And finally, finding out the resources meeting the intention of the user from all the derivation paths, and outputting the resources as a resource identification result. The resource identification result includes entity information of entities in all the target images.

The resource identification method based on the DIKW atlas, provided by the embodiment of the application, includes the steps of obtaining original resources of a target object and intentions of a user, wherein the target object is a tool for transmitting information, and the original resources of the target object include: and at least one of a data resource, an intention resource, an information resource, and a knowledge resource associated with the target object, the intention of the user referring to the intention of the user to acquire the target information from the target object. In the embodiment of the application, the intention of a user and the original resource of a target object are used for traversing a resource map in a DIKW system, a plurality of derivation paths associated with the target object are derived and determined from the resource map, each derivation path determined by derivation comprises the original resource of the target object and a new resource associated with the target object, the new resource associated with the target object is generated by traversing the resource map derivation, and the resource map comprises: the data map, the consciousness map, the information map and the knowledge map have derivation capacity, new resources related to a target object can be derived and generated, hidden information cannot be omitted selectively like the existing intelligent identification technology, then resources meeting the intention of a user are analyzed out from each derivation path finally, the resources meeting the intention of the user are determined as resource identification results of the user, the obtained resource identification results of the user are more accurate and comprehensive compared with the prior art, and the hidden information meeting the intention of the user cannot be lost.

Referring to fig. 7, based on the resource identification method based on the didw atlas shown in fig. 1, the embodiment of the present application correspondingly discloses a resource identification apparatus based on the didw atlas, which includes: an acquisition unit 701, a derivation unit 702, and a first analysis unit 703.

An obtaining unit 701, configured to obtain an original resource of a target object and an intention of a user. The target object is a tool for transmitting information, and the original resource of the target object comprises: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with the target object, the user's intent referring to the user's intent to obtain the target information from the target object.

Optionally, in an embodiment of the present application, if the target object is a target image, the obtaining unit 701 includes: an identification subunit and a conversion subunit.

And the identification subunit is used for carrying out entity identification on the target image to obtain entity information of the identified entity. Wherein, the entity information includes: status, positional relationship, and directional relationship information.

The conversion subunit is used for converting the entity information of the identified entity into an original resource of the target image; wherein the new resources associated with the target object include new resources of unrecognized entities in the target image.

And the derivation unit 702 is configured to traverse a resource map in a didw system by using the intention of the user and the original resource of the target object, and derive and determine a plurality of derivation paths associated with the target object from the resource map. Wherein, the resource map comprises: the system comprises a data graph, an awareness graph, an information graph and a knowledge graph, wherein the derivation paths comprise a plurality of resources which are mutually connected in the resource graph, each derivation path comprises an original resource of a target object and a new resource which is associated with the target object, and the new resource which is associated with the target object is obtained by traversing the resource graph and deriving and generating.

Optionally, in an embodiment of the present application, the derivation unit 702 includes: a second derivation subunit, a first fusion subunit, and a third derivation subunit.

And the second derivation subunit is used for traversing the resource map in the DIKW system by using the intention of the user and the original resource of the target object, and deriving a derivation path where the original resource of the target object is located.

Optionally, in an embodiment of the present application, the second derivation subunit includes: the first traversal subunit, the fourth derivation subunit, the fifth derivation subunit, and the sixth derivation subunit.

And the first traversal subunit is used for traversing the resource map in the DIKW system by using the intention of the user and the original resource of the target object so as to find the derivation path where the original resource of the target object is located.

And the fourth derivation subunit is used for, if the data resource is missing in the derivation path where the original resource of the target object is found in the process of traversing the resource map in the DIKW system, deriving the missing data resource by using the ontology model of the data, the logical model of the information, or the meta model of the knowledge.

And the fifth derivation subunit is used for, if the information resource is missing in the derivation path where the original resource of the target object is found in the process of traversing the resource map in the DIKW system, deriving the missing information resource by using the ontology model of the data or the meta model of the knowledge.

And the sixth derivation subunit is used for, if the knowledge resource is missing in the derivation path where the original resource of the target object is found in the process of traversing the resource map in the DIKW system, using the ontology model of the data or the logic model of the information to derive the missing knowledge resource.

And the first fusion subunit is used for performing fusion processing on the plurality of resources in the derivation path where the original resource of the target object is located to generate a new resource associated with the target object.

Optionally, in a specific embodiment of the present application, the first fusion subunit includes:

and the second fusion subunit is used for performing fusion processing on the plurality of data resources in the derivation path where the original resource of the target object is located, and generating a new data resource associated with the target object. And/or performing fusion processing on the data resource in the derivation path where the original resource of the target object is located and the intention resource to generate a new information resource associated with the target object. And/or fusing the knowledge resources and the data resources in the derivation path where the original resources of the target object are located to generate new data resources or new knowledge resources associated with the target object. And/or performing fusion processing on the information resources and the knowledge resources in the derivation path where the original resources of the target object are located to generate new knowledge resources associated with the target object. And/or fusing a plurality of information resources in the derivation path where the original resource of the target object is located to generate a new information resource associated with the target object. And/or fusing a plurality of knowledge resources in the derivation path where the original resource of the target object is located to generate a new knowledge resource associated with the target object.

And the third derivation subunit is used for traversing the resource map in the DIKW system by using the new resource associated with the target object, the intention of the user and the original resource of the target object aiming at each generated new resource associated with the target object, and deriving a derivation path where the new resource associated with the target object is located.

Optionally, in an embodiment of the present application, the third derivation subunit includes:

and the second traversing subunit is used for traversing the resource map in the DIKW system by using the new resource associated with the target object, the intention of the user and the original resource of the target object so as to find the derivation path where the new resource associated with the target object is located.

And the seventh derivation subunit is used for, if the data resource is missing in the derivation path where the new resource associated with the target object is located in the process of traversing the resource map in the DIKW system, deriving the missing data resource by using the ontology model of the data, the logical model of the information, or the meta model of the knowledge.

And the eighth derivation subunit is used for, if the information resource is missing in the derivation path where the new resource associated with the target object is located in the process of traversing the resource map in the DIKW system, deriving the missing information resource by using the ontology model of the data or the meta model of the knowledge.

And the ninth derivation subunit is used for, if the knowledge resource is found to be missing in the derivation path where the new resource associated with the target object is located in the process of traversing the resource map in the DIKW system, using the ontology model of the data or the logical model of the information to derive the missing knowledge resource.

A first analyzing unit 703 for analyzing the resource satisfying the user's intention from each of the derivation paths, and determining the resource satisfying the user's intention as the resource identification result of the user.

Optionally, in a specific embodiment of the present application, the resource map further includes a value corresponding to each resource. Wherein, resource identification device based on DIKW atlas still includes:

and the second analysis unit is used for analyzing the intention of the user and analyzing and calculating the value dimension and the input dimension of the intention of the user. The value dimension of the user's intention is used for explaining the value brought to the user by the accuracy and the correctness of the target information acquired from the target object by the user, and the investment dimension of the user's intention is used for explaining the cost expected to be invested by the user in order to acquire the target information from the target object. The derivation unit 702 includes:

and the first derivation subunit is used for traversing the resource map in the DIKW system by using the intention of the user, the original resource of the target object, the value dimension of the intention of the user and the input dimension, and deriving and determining a plurality of derivation paths associated with the target object from the resource map. The value corresponding to each resource included in the derivation path meets the requirements of the value dimension and the investment dimension of the intention of the user.

The specific principle and the execution process of each unit in the resource identification device based on the didw atlas disclosed in the embodiment of the present application are the same as those of the resource identification method based on the didw atlas disclosed in the embodiment of the present application, and reference may be made to corresponding parts in the resource identification method based on the didw atlas disclosed in the embodiment of the present application, which are not described herein again.

The resource identification apparatus based on the didw atlas provided by the embodiment of the application obtains the original resource of the target object and the intention of the user through the obtaining unit 701, where the target object is a tool for transferring information, and the original resource of the target object includes: at least one of a data resource, an intent resource, an information resource, and a knowledge resource associated with a target object, the user's intent referring to the user's intent to obtain target information from the target object. In the embodiment of the present application, the derivation unit 702 traverses a resource map in a didw system using an intention of a user and an original resource of an object, and derives and determines a plurality of derivation paths associated with a target object from the resource map, each derivation path determined by derivation includes the original resource of the target object and a new resource associated with the target object, the new resource associated with the target object is generated by traversing the resource map derivation, and the resource map includes: the data map, the consciousness map, the information map and the knowledge map have derivation capacity, new resources related to a target object can be derived and generated, hidden information cannot be omitted selectively like the existing intelligent identification technology, the first analysis unit 703 can finally analyze the resources meeting the intention of the user from each derivation path, the resources meeting the intention of the user are determined as resource identification results of the user, the obtained resource identification results of the user are more accurate and comprehensive compared with the prior art, and the hidden information meeting the intention of the user cannot be lost.

The embodiment of the application discloses a computer readable medium, on which a computer program is stored, wherein the program is executed by a processor to implement the resource identification method based on the DIKW atlas according to any one of the above embodiments.

The embodiment of the application discloses resource identification equipment based on DIKW atlas includes: one or more processors, a storage device, on which one or more programs are stored. When the one or more programs are executed by the one or more processors, the one or more processors implement the method for resource identification based on the didw atlas according to any of the embodiments.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

It is further noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.

Claims

1. A resource identification method based on a DIKW map is characterized by comprising the following steps:

analyzing resources meeting the intention of the user from each derivation path, and determining the resources meeting the intention of the user as a resource identification result of the user;

the resource map further includes a value corresponding to each resource, wherein after the original resource of the target object and the intention of the user are obtained, the method further includes:

analyzing the intention of the user, and analyzing and calculating a value dimension and an input dimension of the intention of the user; the value dimension of the user's intention is used for explaining the value brought to the user by the accuracy and correctness of the target information acquired by the user from the target object; the investment dimension of the user's intent is used to illustrate the cost that the user expects to invest in order to obtain target information from the target object.

2. The method of claim 1, wherein traversing a resource graph in a DIKW system using the user's intent and the original resources of the target object, and deriving from the resource graph a plurality of derived paths associated with the target object comprises:

3. The method of claim 1, wherein traversing a resource graph in a DIKW system using the user's intent and the target object's original resources to determine a plurality of derived paths associated with the target object from the resource graph comprises:

4. The method according to claim 3, wherein the fusing the plurality of resources in the derivation path where the original resource of the target object is located to generate a new resource associated with the target object comprises:

5. The method of claim 3, wherein the step of using the user's intention and the original resource of the target object to traverse the resource map in the DIKW system to derive the derivation path of the original resource of the target object comprises:

6. The method of claim 1, wherein if the target object is a target image, the obtaining of the original resource of the target object comprises:

7. A resource recognition device based on DIKW atlas is characterized by comprising:

a first analysis unit, configured to analyze, from each of the derivation paths, a resource that satisfies an intention of the user, and determine, as a resource identification result of the user, the resource that satisfies the intention of the user;

the second analysis unit is used for analyzing the intention of the user and analyzing and calculating a value dimension and an input dimension of the intention of the user; the value dimension of the user's intention is used for explaining the value brought to the user by the accuracy and correctness of the target information acquired by the user from the target object; the investment dimension of the user's intent is used to illustrate the cost that the user expects to invest in order to obtain target information from the target object.

8. The apparatus of claim 7, wherein the resource graph further comprises a value corresponding to each resource;

wherein, resource recognition device based on DIKW atlas still includes:

wherein the derivation unit includes:

9. A computer-readable medium, on which a computer program is stored, wherein the program, when executed by a processor, implements the method of any one of claims 1 to 6.

10. A DIKW map-based resource identification device is characterized by comprising:

one or more processors;

a storage device having one or more programs stored thereon;

the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-6.