Dynamic ontology object construction method and device based on capability problem backstepping, electronic equipment and storage medium

Abstract

The application discloses a dynamic ontology object construction method and device based on capability problem backstepping, equipment and a storage medium. The method comprises the following steps: acquiring a to-be-processed document, analyzing content expression relationship, structure organization relationship and knowledge association relationship, and determining a plurality of knowledge expression objects; determining content supplement relationship and task support relationship between the knowledge expression objects according to a task processing target corresponding to task demand information and knowledge use demand, and forming a target object unit; performing structured processing according to a knowledge content range corresponding to the target object unit and a field organization requirement, forming a knowledge field, and establishing an association relationship between the knowledge field and the target object unit, so as to generate a dynamic ontology object. The application can organize knowledge content according to task demand, and improve the adaptability of knowledge object construction.

Description

Technical Field

[0001] This application relates to the field of document information processing technology, and in particular to a method, apparatus, electronic device and storage medium for constructing dynamic ontology objects based on capability problem inversion. Background Technology

[0002] With the development of information technology, a large amount of knowledge content recorded in document form has been generated in various business activities, such as technical documents, business specifications, operation guides, analysis reports, and process records. Since this knowledge content is usually organized according to the chapter structure, descriptive order, and business background during the writing process, the relationships between different knowledge contents are often scattered within the document. To improve the efficiency of knowledge utilization, it is necessary to parse and organize the knowledge content in the document, transforming the scattered knowledge content into dynamic ontology objects that can be invoked by the business processing process, thereby supporting subsequent knowledge query, analysis, and application.

[0003] In existing technologies, the processing of document knowledge content typically relies on the document's own structure or the relationships between knowledge content as the organizational basis. This involves extracting, dividing, associating, or structurally representing the text content to form corresponding knowledge representations. For example, the content organization can be determined based on the hierarchical relationships within the document, or the connections between knowledge content can be identified based on the relationships between text elements, thus forming a structured knowledge expression. However, these methods primarily focus on how to extract and express knowledge content from the document. Their organizational process usually depends on the document's original information structure, making it difficult to further analyze the interrelationships between knowledge content in conjunction with specific task processing objectives. In practical applications, different knowledge content within the same document may serve different task requirements. Some knowledge content needs to participate in task processing with other knowledge content to form complete support, while other knowledge content, although related, may not necessarily fall within the scope of knowledge required by the current task. Therefore, forming knowledge objects solely based on the document structure or the relationships between knowledge content itself is insufficient to accurately reflect the correspondence between knowledge content and specific task requirements.

[0004] Furthermore, when dealing with complex business tasks, the effectiveness of a knowledge object depends not only on the quantity of knowledge content it contains, but also on whether different knowledge contents can form a reasonable combination relationship around the task processing objectives. Existing knowledge organization methods lack a knowledge reconstruction process oriented towards task requirements, making it difficult to determine the scope of knowledge content involved in task processing based on specific task objectives, and also difficult to further determine the supplementary relationships, supporting relationships, and corresponding combination boundaries between different knowledge contents. This results in dynamic ontology objects that may contain knowledge content unrelated to the current task, or lack the relevant knowledge content required to complete the current task. Therefore, how to organize and combine knowledge content in a document in a task-driven manner according to task processing objectives and knowledge usage needs, determine the knowledge composition relationships that meet task requirements, and form dynamic ontology objects adapted to specific task scenarios has become an urgent technical problem to be solved. In addition, as knowledge application scenarios change, knowledge objects not only need to meet the current task processing requirements, but also need to support knowledge invocation under subsequent different capability requirements. Existing technologies typically construct knowledge objects based on preset data structures or fixed knowledge categories, making it difficult to reverse-determine the organizational scope and expression form of knowledge objects according to actual capability requirements, and also making it difficult to dynamically adjust knowledge objects when knowledge content changes. Summary of the Invention

[0005] This application aims to address the problem that in existing document information processing, knowledge content is usually organized based on the document's own organizational structure or the relationships between knowledge content. This makes it difficult to determine the scope of knowledge content involved in the task processing and the combination relationships between different knowledge contents in conjunction with the specific task processing objectives, resulting in a low degree of adaptability between the generated dynamic ontology object and the actual task requirements.

[0006] Specifically, when performing structured processing of document content, existing technologies typically focus on the extraction and expression of the knowledge content itself, but lack a processing mechanism to reorganize the knowledge content according to task requirements. They cannot determine the content supplementation relationship and task support relationship between different knowledge contents based on the task processing objectives, and it is also difficult to further determine the scope and composition boundary of the knowledge content combination used to form dynamic ontology objects.

[0007] Therefore, this application provides a method, apparatus, electronic device, and storage medium for constructing dynamic ontology objects based on capability problem inversion, so as to realize the adaptive organization of knowledge content in documents according to task requirements and form dynamic ontology objects that meet task processing requirements.

[0008] To achieve the above objectives, this application provides a method for constructing dynamic ontology objects based on capability problem inversion, including: Obtain the document to be processed, parse the content expression relationship, structural organization relationship and knowledge association relationship in the document to be processed, and determine multiple knowledge expression objects; Obtain task requirement information, determine the content supplementation relationship and task support relationship between multiple knowledge expression objects based on the task processing objectives and knowledge usage requirements corresponding to the task requirement information, determine the combination of knowledge expression objects participating in the construction of dynamic ontology objects based on the content supplementation relationship and task support relationship, determine the combination range and composition boundary corresponding to the combination of knowledge expression objects, and organize the combination of knowledge expression objects according to the combination range and composition boundary to form target object units; Based on the knowledge content scope and field organization requirements corresponding to the target object unit, the target object unit is structured to form knowledge fields, and the association between the knowledge fields and the target object unit is established to generate a dynamic ontology object.

[0009] Optionally, the step of parsing the content expression relationships, structural organization relationships, and knowledge association relationships in the document to be processed to determine multiple knowledge expression objects includes: The scope of knowledge description in the document to be processed is determined based on the content expression relationship; The hierarchical relationship between different knowledge description scopes is determined based on the structural organization relationship, and the association relationship between different knowledge description scopes is determined based on the knowledge association relationship; The boundaries of knowledge content are adjusted according to the scope, hierarchical relationship, and association relationship of the knowledge description to form multiple knowledge expression objects.

[0010] Optionally, adjusting the knowledge content division boundaries based on the knowledge description scope, hierarchical relationship, and association relationship includes: The scope of associated content corresponding to the scope of knowledge description is determined based on the hierarchical relationship; Determine the associated knowledge content corresponding to the scope of associated content based on the knowledge association relationship; The knowledge content division boundaries are adjusted according to the scope of the associated content and the associated knowledge content to form the knowledge expression object.

[0011] Optionally, determining the content supplementation relationship and task support relationship among the multiple knowledge representation objects includes: Determine the knowledge content required to complete the corresponding task based on the task processing objectives; Based on the knowledge content, determine the corresponding knowledge representation object from among the multiple knowledge representation objects; Based on the knowledge content deficiencies corresponding to the knowledge representation object and its supporting role in the task processing objective, the content supplementation relationship and task support relationship are determined.

[0012] Optionally, determining the content supplementation relationship and task support relationship based on the knowledge content deficiencies corresponding to the knowledge representation object and its supporting role in the task processing objective includes: Based on the knowledge integrity requirements corresponding to the task processing objectives, determine the scope of knowledge content that needs to be covered to complete the task processing objectives; Based on the scope of the knowledge content, analyze the missing content areas and corresponding supplementary content among multiple knowledge expression objects; Based on the missing content areas and the corresponding supplementary content, a combination of knowledge expression objects that jointly cover the scope of the knowledge content is determined.

[0013] Optionally, the step of combining the knowledge representation objects to form the target object unit includes: The combination boundary is determined based on the knowledge content coverage of the combination of knowledge representation objects. The composition range corresponding to the combination of knowledge representation objects is determined based on the combination boundary. The knowledge representation objects are associated and organized according to the defined composition range to form the target object unit.

[0014] Optionally, after forming the target object unit, the method further includes: Based on the content association relationships between the knowledge representation objects that form the target object unit, determine the content connection relationships between different knowledge representation objects; Adjust the organization of knowledge content corresponding to the target object unit according to the content connection relationship.

[0015] Optionally, the method further includes: Obtain information on changed task requirements or newly added knowledge representation objects; Based on the changed task requirements information or the newly added knowledge representation objects, the content supplementation relationship and task support relationship between knowledge representation objects are redefined. The knowledge representation object combination is adjusted according to the redefined content supplementation relationship and task support relationship, and the target object unit is re-formed.

[0016] Optionally, adjusting the corresponding dynamic ontology object based on the re-formed target object unit includes: The changed knowledge content is determined based on the re-formed target object unit; Determine the corresponding knowledge fields based on the changed knowledge content; Update the dynamic ontology object based on the knowledge fields.

[0017] Optionally, the step of structuring the target object unit according to the knowledge content scope and field organization requirements corresponding to the target object unit to form knowledge fields includes: The field division range corresponding to the target object unit is determined based on the scope of the knowledge content. Determine the field association method corresponding to the field division range based on the field organization requirements; The knowledge fields are formed based on the field division scope and field association methods.

[0018] Optionally, the method further includes: The knowledge representation object that forms the target object unit is determined based on the field content corresponding to the knowledge field. The association between the knowledge field and the knowledge representation object is established based on the knowledge content source corresponding to the knowledge representation object.

[0019] Optionally, the method further includes: Obtain content change information corresponding to the knowledge representation object that forms the target object unit; Based on the content change information, determine the affected target object units and the corresponding range of change; Adjust the target object unit according to the range of change, and update the corresponding dynamic ontology object.

[0020] Optionally, determining the affected target object unit and the corresponding range of change based on the content change information includes: Based on the information about the changes in the content, determine the knowledge representation objects that have changed; The scope of influence corresponding to the target object unit is determined based on the content supplementation relationship and task support relationship corresponding to the changed knowledge expression object; The knowledge content that needs to be adjusted is determined based on the scope of the impact.

[0021] Optionally, the method further includes: Obtain information on the requirements of multiple tasks; Determine the corresponding combination of knowledge expression objects based on the task processing objectives and knowledge usage requirements corresponding to different task requirements; Different knowledge representation objects are combined to form corresponding target object units, and corresponding dynamic ontology objects are generated.

[0022] Optionally, determining the corresponding combination of knowledge representation objects based on the task processing objectives and knowledge usage requirements corresponding to different task requirement information includes: Adjust the range of combinations of knowledge expression objects according to the knowledge content requirements corresponding to different task requirements; Determine the corresponding composition boundaries based on the adjusted combination range; The corresponding target object unit is formed based on the composition boundary.

[0023] This application also provides a dynamic ontology object construction apparatus based on capability problem inversion, including: The knowledge parsing module is used to acquire the document to be processed, parse the content expression relationship, structural organization relationship and knowledge association relationship in the document to be processed, and determine multiple knowledge expression objects; The object construction module is used to obtain task requirement information, determine the content supplementation relationship and task support relationship between multiple knowledge expression objects based on the task processing goal and knowledge usage requirements corresponding to the task requirement information, determine the combination of knowledge expression objects participating in the construction of dynamic ontology objects based on the content supplementation relationship and task support relationship, determine the combination range and composition boundary corresponding to the combination of knowledge expression objects, and organize the combination of knowledge expression objects according to the combination range and composition boundary to form a target object unit; The structure generation module is used to perform structured processing on the target object unit according to the knowledge content scope and field organization requirements corresponding to the target object unit, form knowledge fields, and establish the association between the knowledge fields and the target object unit to generate a dynamic ontology object.

[0024] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, it implements the steps of the above method.

[0025] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described method.

[0026] Compared with the prior art, this application has at least the following beneficial effects: This application analyzes the content expression relationships, structural organization relationships, and knowledge association relationships in the document to be processed, forming multiple knowledge expression objects. This enables the knowledge content in the document to be organized as independent knowledge expression units, providing a foundation for subsequent knowledge reconstruction based on task requirements.

[0027] Furthermore, based on the task processing objectives and knowledge usage requirements, this application analyzes the content supplementation relationship and task support relationship between different knowledge expression objects, and determines the combination of knowledge expression objects and the corresponding combination range and composition boundary based on the determined relationship, so that the organization process of knowledge content no longer depends on the original document structure, but can form adaptive target object units according to specific task requirements.

[0028] Furthermore, this application performs structuring processing on the target object unit to form corresponding knowledge fields and establishes the association between the knowledge fields and the target object unit, so that the generated dynamic ontology object can maintain the association between the knowledge organization results corresponding to the task requirements and improve the matching degree between the dynamic ontology object and the actual task processing requirements.

[0029] Therefore, this application can realize the reorganization of knowledge content according to task requirements, and improve the adaptability and flexibility of the dynamic ontology object construction process. Attached Figure Description

[0030] Figure 1 A flowchart illustrating the dynamic ontology object construction method based on capability problem inversion provided in this application; Figure 2 This is a schematic diagram of a dynamic ontology object construction device based on capability problem inversion, which is provided for this application.

[0031] Figure 3 A schematic diagram of an electronic device structure is provided in this application; Figure 4 This is a schematic diagram of a computer-readable storage medium structure provided in this application. Detailed Implementation

[0032] The following describes in detail, with reference to the accompanying drawings, the dynamic ontology object construction method based on capability problem inversion provided in the embodiments of this application.

[0033] The method provided in this application does not directly generate a dynamic ontology object based on the original structure of the document to be processed. Instead, after obtaining the document, it first parses the knowledge content carried in the document to form knowledge representation objects that can reflect independent knowledge scopes. Further, it determines the combination relationships between different knowledge representation objects according to specific task requirements, so that the resulting knowledge organization can adapt to the current task processing goal. Finally, it performs structured processing based on the formed target object units to generate a dynamic ontology object. In this application, a dynamic ontology object refers to a structured knowledge object formed by determining the knowledge content scope according to task or capability requirements and organizing it based on multiple knowledge representation objects. The dynamic ontology object is not a pre-set fixed data structure, but can be adjusted according to changes in task requirements, changes in knowledge content, or the addition of new knowledge representation objects to achieve dynamic updates of the knowledge object content scope and organizational relationships.

[0034] like Figure 1 As shown in the figure, the present application provides a method for constructing dynamic ontology objects based on capability problem inversion, which includes the following steps: S1, Obtain the document to be processed, parse the content expression relationship, structural organization relationship and knowledge association relationship in the document to be processed, and determine multiple knowledge expression objects.

[0035] In this embodiment, the document to be processed is first obtained. This document can be an electronic document containing business knowledge, technical knowledge, or rule knowledge, such as business specification documents, technical description documents, operation guidance documents, and management system documents. Since the document to be processed is usually arranged according to the writer's descriptive habits and document organization methods, the knowledge content required for different tasks may be scattered in different locations. Therefore, this embodiment does not directly divide the knowledge content based on chapters or paragraphs in the document. Instead, it analyzes the content expression relationships, structural organization relationships, and knowledge association relationships in the document to be processed to determine the knowledge expression objects that can independently express knowledge content.

[0036] Specifically, the first step is to analyze the content expression relationships within the document to be processed. These content expression relationships reflect the types and meanings of knowledge content expressed by different content sections in the document. Specifically, the scope of knowledge expressed by a given content section can be determined based on the objects, attributes, processes, and results described in each section. For example, in a technical document describing equipment operation and management, information describing the equipment's components expresses the equipment's structure, information describing operating parameters expresses the equipment's status, and information describing fault handling procedures expresses the anomaly handling. Although all of these contents reside within the same document, they express different knowledge content; therefore, the corresponding scope of knowledge content can be determined based on the content expression relationships.

[0037] Furthermore, the structural organization relationships within the document to be processed are analyzed. These structural organization relationships reflect the organization of different knowledge content within the document and their hierarchical relationships. Specifically, the organizational relationships between different content sections can be determined based on chapter relationships, title attribution relationships, content arrangement relationships, and citation relationships within the document. For example, when the document includes a general equipment introduction chapter and multiple component introduction chapters, the inclusion relationship between the overall equipment information and component information can be determined based on the chapter hierarchy; when one content section references definitions, parameters, or processing conditions from another content section, the organizational association between the two can be determined based on the citation relationship. By analyzing the structural organization relationships, the organization of different knowledge content within the original document can be determined, providing a basis for subsequently adjusting the scope of knowledge content division.

[0038] Furthermore, the knowledge relationships within the document to be processed are analyzed. These knowledge relationships reflect the connections between different knowledge contents based on the described object, processing procedure, and knowledge dependency. Specifically, knowledge relationships can be determined based on whether the objects corresponding to different content parts are consistent, whether the processing procedure is continuous, and whether one content part depends on another. For example, in equipment technical documents, equipment structure information and operating parameter information both describe the same equipment object, indicating an object relationship between them; similarly, fault phenomenon description information and fault handling process information have a processing relationship because the handling process is used to resolve the corresponding fault phenomenon. By analyzing these knowledge relationships, it can be determined whether different knowledge contents need to maintain an associated organization.

[0039] After obtaining the content expression relationships, structural organization relationships, and knowledge association relationships, the knowledge content in the document to be processed can be organized and analyzed based on these three types of relationships to determine the scope of content that forms the knowledge expression object. Specifically, the basic knowledge scope corresponding to each content part can be determined first based on the content expression relationships, so that the content corresponding to different objects, attributes, processes, or results can be identified; then, the structural organization relationships can be used to determine whether there are hierarchical inclusion, reference, or organizational relationships between different knowledge contents; and at the same time, the knowledge association relationships can be used to determine whether different knowledge contents need to maintain a connection based on a common descriptive object or processing process.

[0040] Based on the above analysis, the boundaries of knowledge content in the original document can be adjusted. Multiple content sections expressing the same knowledge object, having a continuous processing flow, or exhibiting clear knowledge dependencies can be organized into a single knowledge object. Content sections that are adjacent only due to document layout but describe different objects and have no knowledge connection can be formed into different knowledge objects.

[0041] For example, for a document containing object introduction, function description, operation process, and exception handling, the scope of knowledge content corresponding to the basic object information, function description, operation process, and exception handling can be determined based on the content expression relationship; the chapter organization relationship of the above content in the document can be determined based on the structural organization relationship; and the relationship between function description and operation process, and between abnormal phenomena and handling methods can be determined based on the knowledge association relationship. Based on the above relationship analysis results, knowledge expression objects for basic object information, function description, operation process, and exception handling can be formed.

[0042] In this embodiment, the knowledge representation object (BRE) represents a knowledge unit in the document to be processed that has a relatively complete knowledge content scope. It does not simply correspond to fixed chapters, paragraphs, or text fragments in the document, but is formed by determining the knowledge content scope based on content expression relationships and adjusting the content organization boundaries in conjunction with structural organization relationships and knowledge association relationships. Through this method, this embodiment converts continuous text content in the original document into multiple BREs with clearly defined knowledge content scopes. This allows subsequent steps to determine the combination of BREs participating in the construction of the dynamic ontology object based on the content supplementation and task support relationships between different BREs, according to specific task requirements, without needing to re-filter the knowledge content from the original document.

[0043] S2, Obtain task requirement information, and determine the content supplementation relationship and task support relationship between multiple knowledge expression objects based on the task processing objectives and knowledge usage requirements corresponding to the task requirement information. Determine the combination of knowledge expression objects participating in the construction of dynamic ontology objects based on the content supplementation relationship and task support relationship, and determine the combination range and composition boundary corresponding to the combination of knowledge expression objects. Associate and organize the combination of knowledge expression objects according to the combination range and composition boundary to form target object units.

[0044] In this embodiment, after forming multiple knowledge representation objects through step S1, task requirement information is further obtained, and the existing knowledge representation objects are organized and processed in a task-oriented manner according to the current task requirements. It should be noted that this application does not directly generate dynamic ontology objects based on the original document association relationships or fixed organization methods between knowledge representation objects after they are formed. Instead, it further combines the specific task processing objectives and knowledge usage requirements to determine which knowledge representation objects need to participate in the current task processing, thereby forming a knowledge organization result that matches the task requirements.

[0045] In this embodiment, the task requirement information describes the processing task that needs to be completed based on knowledge content and the corresponding knowledge organization requirements. The task requirement information does not simply represent the user's query content, but rather information used to determine the subsequent organization direction of the knowledge representation object. Specifically, the task requirement information may include the task processing purpose, the type of problem to be solved, the type of processing result to be output, and the knowledge content requirements needed to complete the corresponding processing.

[0046] For example, in an expense reimbursement scenario, the task requirement information could correspond to the processing task of "determining whether the accommodation expenses submitted by an employee meet the reimbursement requirements." The task objective is to determine if the accommodation expenses meet the reimbursement conditions, and the knowledge usage requirement is to acquire knowledge related to personnel level, business trip location, accommodation standards, and approval requirements. In a knowledge query scenario, the task requirement information could correspond to the processing task of "querying the accommodation standards corresponding to a specified region." The task objective is to obtain the standard information corresponding to the target region, and the knowledge usage requirement mainly involves knowledge related to the region's scope and accommodation limits. Because the task processing objectives and knowledge usage requirements corresponding to different task requirement information are different, even when faced with the same document to be processed and the same set of knowledge representation objects, different knowledge representation objects may need to be selected for combination. Therefore, this embodiment, by acquiring task requirement information, enables the subsequent knowledge organization process to determine the knowledge representation objects participating in the construction of the dynamic ontology object based on the actual needs of the current task, rather than organizing them according to the fixed relationships of the document content or the knowledge representation objects themselves.

[0047] Furthermore, in this embodiment, the task processing objective represents the processing purpose that the current task needs to achieve, and the knowledge usage requirement represents the scope of knowledge content required to achieve the task processing objective and the coordination requirements between different knowledge contents. Based on the task processing objective, the knowledge content to be covered to complete the current task can be determined, and the type of knowledge content to be invoked can be determined according to the knowledge usage requirement. For example, in the accommodation expense reimbursement scenario, when the task processing objective is to determine whether the accommodation expenses submitted by an employee meet the reimbursement requirements, completing this task requires not only obtaining the accommodation expenses themselves, but also combining knowledge content such as the accommodation standards corresponding to the employee's level, the accommodation limits corresponding to the business trip location, and the corresponding approval requirements. Therefore, it is necessary to determine the knowledge content that can jointly support the task processing objective from multiple knowledge representation objects.

[0048] Furthermore, based on the scope of knowledge content corresponding to the task processing objective and the knowledge usage requirements, the content complementarity relationships between multiple knowledge representation objects are determined. These content complementarity relationships represent the complementary relationships of knowledge content between different knowledge representation objects when meeting the current task requirements. Specifically, the correspondence between the knowledge content required to be covered by the task processing objective and the knowledge content contained in each knowledge representation object can be analyzed to determine whether different knowledge representation objects provide the different knowledge content required for the current task. When one knowledge representation object can only provide a portion of the knowledge content required for task processing, while another knowledge representation object can provide the corresponding missing content, a content complementarity relationship can be determined between the two.

[0049] For example, in a task involving accommodation expense reimbursement, a knowledge representation object (KRE) describing the accommodation standards corresponding to employee levels can provide a basis for judging personnel qualifications; a KRE for describing accommodation limits in different regions can provide a basis for judging regional restrictions; and a KRE for describing reimbursement approval requirements can provide a basis for judging the process. These KREs correspond to different judgment dimensions in the accommodation expense review process. While individually they cannot fully cover the knowledge required for reimbursement judgment, their combination can collectively meet the task processing requirements. Therefore, different KREs form a complementary relationship.

[0050] Based on the established content complementarity relationships, the task support relationships between knowledge representation objects are further determined according to the task processing objectives. These task support relationships represent the influence of different knowledge representation objects on achieving the current task processing objectives. Specifically, based on the current task processing procedure, it can be determined whether each knowledge representation object participates in the task judgment, task analysis, or task result determination process. In other words, content complementarity relationships focus on whether different knowledge representation objects can jointly provide complete knowledge content, while task support relationships focus on whether these knowledge representation objects truly serve the current task objectives.

[0051] For example, in a accommodation expense reimbursement task, the accommodation standard knowledge representation object is used to determine whether the expense exceeds the corresponding standard, the regional limit knowledge representation object is used to determine the expense restrictions in different regions, and the approval requirement knowledge representation object is used to determine whether further approval is needed. These knowledge representation objects participate in different processing stages of the reimbursement review process, thus collectively forming a task support relationship for the accommodation expense reimbursement task. By simultaneously determining the content supplementation relationship and the task support relationship, it is possible to avoid combining knowledge representation objects solely based on their content similarity or document position. For instance, although two knowledge representation objects describe similar topics or are arranged adjacently in the original document, if they cannot jointly cover the knowledge scope required for the task processing objective, or if one of the knowledge representation objects will not participate in the task processing process, then it is unnecessary to include them in the knowledge representation object combination corresponding to the current task.

[0052] After determining the content supplementation relationship and the task support relationship, the combination of knowledge representation objects participating in the construction of the dynamic ontology object can be determined based on the above relationships. Specifically, knowledge representation objects that can jointly cover the knowledge scope required for the task processing goal and jointly support the task processing process can be selected from multiple knowledge representation objects to form a corresponding combination of knowledge representation objects.

[0053] Furthermore, the corresponding combination scope and composition boundary are determined based on the identified combination of knowledge representation objects. The combination scope represents the range of knowledge content that the combination of knowledge representation objects needs to cover, i.e., the types of knowledge content required by the current task processing objective. The composition boundary limits the range of knowledge representation objects participating in the formation of the target object unit, i.e., determining which knowledge representation objects need to be included in the knowledge organization results corresponding to the current task, and which knowledge representation objects, although related, do not belong to the current task processing scope and do not need to participate. For example, in an accommodation expense reimbursement task, the combination scope may include knowledge content such as personnel conditions, regional standards, expense limits, and approval requirements; the composition boundary is used to determine the specific knowledge representation objects participating in the formation of the target object unit, such as including accommodation standard knowledge representation objects, regional limit knowledge representation objects, and approval requirement knowledge representation objects, but excluding knowledge representation objects related to transportation expenses.

[0054] After determining the scope and boundaries of the combination, the knowledge representation objects are associated and organized according to these scopes and boundaries to form target object units. These target object units represent the knowledge content organization results formed for a specific task processing objective. They are not simply a collection of multiple knowledge representation objects, but rather a knowledge organization result reorganized under task requirements based on the content supplementation and task support relationships between the knowledge representation objects. For example, for an accommodation expense reimbursement task, the knowledge representation objects corresponding to accommodation standards, regional limits, reimbursement conditions, and approval requirements can be associated and organized to form a target object unit supporting the accommodation expense review task. This target object unit fully reflects the knowledge content combination required for the current task processing, providing a foundation for subsequent field-based processing of the target object unit and the generation of dynamic ontology objects. It should be noted that the target object unit represents the knowledge organization result during the formation of dynamic ontology objects, and its corresponding knowledge content scope and boundaries can be adjusted according to changes in capability requirements. When new business processing capabilities or new knowledge application scenarios need to be supported, the knowledge representation object combination participating in the construction can be redefined, and the dynamic ontology object corresponding to the target object unit can be updated.

[0055] In the above manner, the embodiments of this application introduce task requirements into the knowledge object construction process, so that knowledge expression objects are no longer combined in a fixed way, but the scope of knowledge content participating in the organization is determined according to different task processing goals, and target object units adapted to the current task requirements are formed, thereby realizing the transformation from document knowledge content to task-driven dynamic ontology objects.

[0056] S3. Based on the knowledge content scope and field organization requirements corresponding to the target object unit, the target object unit is structured to form knowledge fields, and the association between the knowledge fields and the target object unit is established to generate a dynamic ontology object.

[0057] In this embodiment, after forming the target object unit through step S2, the target object unit is further processed into a structured form to generate a dynamic ontology object. It should be noted that the structured processing in this embodiment does not involve directly extracting fields from the original document content. Instead, it involves standardizing the organized knowledge content within the target object unit, based on the target object unit already determined according to task requirements. In other words, step S2 addresses the question of "which knowledge contents need to participate in the current task processing," while step S3 addresses the question of "how these determined knowledge contents can be formed into an identifiable and callable dynamic ontology object."

[0058] In this application, the task requirement information can be represented as a business capability problem or processing objective that needs to be completed using knowledge objects. By analyzing the information requirements corresponding to the capability problem, the scope of knowledge content required to complete the corresponding capability can be determined, and based on the scope of knowledge content, the combination of knowledge expression objects participating in the construction of the dynamic ontology object can be determined in reverse, so that the generated dynamic ontology object can match the actual capability requirements.

[0059] Specifically, the knowledge content that needs to be structured is first determined based on the knowledge content scope corresponding to the target object unit. The knowledge content scope corresponding to the target object unit represents the specific types of knowledge content contained within that target object unit and the organizational relationships between these knowledge contents. Specifically, the knowledge content in the target object unit that needs to be described in a field-based manner can be determined based on the object information, attribute information, process information, and result information corresponding to each knowledge representation object that forms the target object unit.

[0060] For example, in the scenario of accommodation expense reimbursement, after forming a target object unit for the task of "determining whether the accommodation expense meets the reimbursement requirements", the target object unit contains knowledge content such as the accommodation standard corresponding to the employee level, the accommodation limit corresponding to the business trip area, accommodation expense information, and approval requirements. Then, the structured expression content to be generated can be determined based on the above knowledge content scope, including personnel condition information, regional restriction information, expense standard information, and approval requirement information.

[0061] Furthermore, based on the field organization requirements corresponding to the target object unit, the knowledge content within the target object unit is divided and organized into fields. The field organization requirements represent the organizational method used when converting the knowledge content within the target object unit into knowledge fields. Specifically, the field organization requirements may include requirements such as the classification method of knowledge content, the correspondence between fields, and the data expression method between fields, used to determine which field format should be used to represent different knowledge content.

[0062] For example, for the target object unit corresponding to accommodation expense reimbursement, the knowledge content related to employee level can be organized into a "Personnel Level Field" according to the field organization requirements, the knowledge content related to regional limits can be organized into a "Regional Limit Field", the knowledge content related to accommodation standards can be organized into an "Accommodation Standard Field", and the knowledge content related to approval requirements can be organized into an "Approval Requirements Field". Among them, different fields correspond to different knowledge content ranges in the target object unit, enabling the dynamic ontology object to retain the task-related knowledge contained in the original target object unit.

[0063] After determining the scope of knowledge content and the requirements for field organization, the target object unit can be structured to form corresponding knowledge fields. Specifically, based on the division relationship between various knowledge contents in the target object unit, knowledge content with independent expression needs can be converted into corresponding fields, and the field values, field descriptions, or field-related content can be kept consistent with the knowledge content in the target object unit.

[0064] For example, in a device status analysis scenario, when the target object unit contains device composition information, operating parameter information, and anomaly handling information, device composition fields, operating parameter fields, and anomaly handling fields can be formed according to field organization requirements, so that the above fields correspond to different knowledge content ranges in the target object unit.

[0065] Furthermore, after forming the knowledge field, an association is established between the knowledge field and the target object unit. This association indicates which part of the knowledge content within the target object unit the knowledge field originates from, and which target object unit the knowledge field belongs to. Specifically, the correspondence between the knowledge field and the target object unit can be established based on the source of the knowledge content corresponding to the knowledge field, the field's scope, and the compositional relationship of the target object unit.

[0066] For example, when the target object unit is formed by the knowledge representation objects of personnel level accommodation standards, regional quotas, and approval requirements, an association can be established between the personnel level field and the corresponding accommodation standard knowledge content, and an association can be established between the regional quota field and the regional restriction knowledge content. Furthermore, the correspondence between the above fields and the target object unit can be maintained, so that the task knowledge organization scope corresponding to the field can be determined when the dynamic ontology object is called later.

[0067] By establishing the association between knowledge fields and target object units, we can avoid losing the original knowledge organization relationship by only forming mutually independent field information. In other words, the dynamic ontology object generated in this embodiment not only contains the fieldized knowledge content, but also retains the correspondence between fields and target object units formed by task requirements, enabling the dynamic ontology object to reflect the knowledge content combination process.

[0068] For example, in the task of reimbursing accommodation expenses, the generated dynamic ontology object not only includes the accommodation standard field, the regional limit field, and the approval requirement field, but also can determine that the above fields jointly serve the accommodation expense review task through the association between the fields and the target object unit, rather than originating from unrelated policy text content.

[0069] Through the above methods, the embodiments of this application perform structuring processing on the basis of the target object unit, transforming the knowledge content formed by task-oriented organization into a dynamic ontology object with clear field expression relationships, thereby realizing the transformation from task-adaptive knowledge organization results to structured knowledge expression results.

[0070] Preferably, based on the above embodiments, this application also provides a specific implementation method for determining the knowledge expression object based on content expression relationship, structural organization relationship and knowledge association relationship.

[0071] In this embodiment, the process of forming knowledge representation objects in step S1 is mainly used to convert the original content in the document to be processed into knowledge representation objects with a clear knowledge content scope. This allows subsequent steps to determine the knowledge content that needs to be processed together based on the content supplementation relationship and task support relationship between different knowledge representation objects, according to task requirements. Specifically, after obtaining the document to be processed, the scope of knowledge content expressed by different content parts can be determined first based on the content representation relationship in the document. The content representation relationship is used to reflect the knowledge information such as objects, attributes, processes, and results described by the document content itself. Specifically, the knowledge representation scope corresponding to different content parts can be determined by analyzing the text descriptions, table content, parameter descriptions, and process descriptions in the document. For example, in a corporate travel management system document, the content part describing accommodation standards expresses knowledge of accommodation cost restrictions, the content part describing the approval process expresses knowledge of reimbursement review process, and the content part describing the personnel level corresponding standards expresses knowledge of personnel conditions. Although the above content exists in the same system document, they correspond to different knowledge content scopes.

[0072] Furthermore, the organizational relationships between different knowledge content areas are determined based on the structural organization. These structural organization relationships reflect the arrangement and hierarchical relationship of knowledge content within the document. Specifically, the presence of inclusion, dependency, or succession relationships between different knowledge contents can be determined based on chapter levels, title attribution, content citations, and preceding and following descriptions. For example, if a policy document has a "Accommodation Fee Standard" chapter with two sub-chapters, "Standards for Different Personnel Levels" and "Standards for Different Regions," then the organizational relationship between the accommodation fee standard, personnel level standard, and regional standard can be determined based on the chapter hierarchy. Similarly, if a process description section cites a judgment condition from another rule description section, then the organizational relationship between the two can be determined based on the content citation relationship.

[0073] Furthermore, the relationships between different knowledge content scopes are determined based on knowledge association relationships. These knowledge association relationships reflect the connections formed between different knowledge contents due to the described object, processing procedure, or information dependency. Specifically, this can be determined based on whether different content sections describe the same object, whether there is a sequential relationship in the processing procedure, and whether they need to be used in combination. For example, accommodation standard information and personnel level information are both used to determine the scope of accommodation reimbursement for employees. Although they are in different chapters, they jointly affect the determination of accommodation expenses, therefore there is a knowledge association relationship between them. Similarly, there is a knowledge association relationship between the description of a fault phenomenon and the description of fault handling measures because the handling measures are formed in response to the corresponding fault phenomenon.

[0074] After determining the content expression relationships, structural organization relationships, and knowledge association relationships, the knowledge content in the document to be processed can be reorganized according to these relationships to determine the knowledge expression objects. Specifically, the basis for knowledge content division can be determined based on the content expression relationships, the scope of content that needs to maintain its organizational relationship can be determined based on the structural organization relationships, and the related content that needs to be retained can be determined based on the knowledge association relationships. Content that expresses the same object, attribute, process, or result and can collectively form a complete knowledge meaning can be classified into the same knowledge expression object; content that is adjacent only due to document layout but describes different objects and has no knowledge association can be classified into different knowledge expression objects.

[0075] For example, for a travel policy document containing content such as "accommodation standards," "regional limits," and "approval processes," the scope of knowledge content corresponding to the above content can be determined according to the content expression relationship; the organization method between policy chapters can be determined according to the structural organization relationship; and the relationship between accommodation standards, regional limits, and approval processes in reimbursement tasks can be determined according to the knowledge association relationship, thereby forming knowledge expression objects for accommodation standards, regional limits, and approval processes.

[0076] The knowledge representation objects formed in the above way are not simply divided according to document chapters or text paragraphs, but are comprehensively determined based on the knowledge scope, organization, and relevance of the document content. This ensures that each knowledge representation object can carry relatively complete knowledge content, providing a foundation for subsequent knowledge combination based on task requirements.

[0077] Preferably, after forming the scope, hierarchical relationship, and association relationship of the knowledge description, the boundaries of the knowledge content division can be further adjusted based on the above information to improve the completeness of the content contained in the knowledge expression object.

[0078] In this embodiment, the knowledge content partition boundary is used to define the scope of knowledge content contained in a knowledge representation object. Since the knowledge content in the document to be processed is usually arranged according to the author's writing logic, rather than organized according to the needs of subsequent task processing, information related to the same knowledge object may be distributed in multiple locations, or a single chapter may contain information from multiple different knowledge objects simultaneously. Therefore, when determining the knowledge content partition boundary, the original content partitioning method can be adjusted by considering the correlation between the knowledge description scopes.

[0079] Specifically, when different sections describe the same object, and one part depends on another to form a complete knowledge meaning, the scope of knowledge content division can be expanded, organizing related content into a single knowledge representation object. For example, in an equipment instruction document, a chapter might describe both the equipment's structural composition and the maintenance methods for its corresponding components. Since understanding the maintenance methods requires considering the specific component structure, the division boundaries can be adjusted based on knowledge relationships, allowing the equipment composition information and corresponding maintenance information to jointly form a single knowledge representation object. Conversely, when a chapter contains information about multiple different objects, and there are no knowledge relationships between these different pieces of information, the scope of knowledge content division can be narrowed, forming separate knowledge representation objects for each. For example, in a travel policy document, a chapter might introduce both transportation and accommodation expense standards, but these correspond to different reimbursement categories and do not need to be used together in accommodation expense review tasks. Therefore, transportation expense knowledge representation objects and accommodation expense knowledge representation objects can be formed separately based on knowledge relationships.

[0080] By using the above methods, the resulting knowledge representation objects can maintain the integrity of the knowledge content while avoiding the inclusion of information unrelated to the current knowledge scope, thus providing a more accurate basis for determining the combination of knowledge representation objects according to task requirements.

[0081] Preferably, based on the above embodiments, this application also provides a specific implementation method for determining the content supplementation relationship and task support relationship between knowledge representation objects according to task requirement information.

[0082] In this embodiment, after forming multiple knowledge representation objects through step S1, the existing knowledge representation objects are further analyzed based on task requirements to determine which knowledge representation objects need to participate in the current task processing. Since the knowledge content required for different task processing objectives varies, the same batch of knowledge representation objects does not correspond to a fixed combination. Instead, the corresponding combination of knowledge representation objects needs to be determined based on the current task processing objective and knowledge usage requirements.

[0083] Specifically, task requirement information indicates the processing task that needs to be completed using knowledge content and the corresponding knowledge organization requirements. Based on the task requirement information, the corresponding task processing objective and knowledge usage requirements can be determined. The task processing objective indicates the processing result that the current task needs to achieve, such as completing a judgment, analysis, query, or generating a certain type of processing result; the knowledge usage requirements indicate the scope of knowledge content required to complete the task processing objective and the coordination relationships between different knowledge contents.

[0084] After defining the task processing objective, the corresponding knowledge content scope can be determined based on the information required to complete the task processing objective. Then, knowledge representation objects (KPIs) that can provide the relevant knowledge content can be selected from multiple KPIs. For example, in a accommodation expense reimbursement scenario, when the task processing objective is to determine whether an employee's submitted accommodation expense meets the reimbursement requirements, it is necessary to obtain information such as the accommodation standard corresponding to the employee's level, the limit requirements for the business trip region, accommodation expense information, and approval rules. Therefore, it is necessary to select KPIs from multiple KPIs that respectively contain the aforementioned knowledge content.

[0085] Furthermore, based on the scope of knowledge content corresponding to the task processing objective and the knowledge content contained in the knowledge representation objects, the content complementarity relationship between multiple knowledge representation objects is determined. This content complementarity relationship represents the complementary relationship between different knowledge representation objects in terms of knowledge content coverage to meet the current task requirements. Specifically, it can be determined whether the knowledge content required by the current task is completely provided by a single knowledge representation object. If one knowledge representation object can only provide part of the knowledge required by the task, while another knowledge representation object can provide the corresponding missing knowledge content, then a content complementarity relationship is determined to exist between the two.

[0086] For example, in the task of reimbursing accommodation expenses, the accommodation standard knowledge representation object can provide standard information corresponding to employee levels, while the regional limit knowledge representation object can provide restriction information corresponding to different regions. The two correspond to different judgment criteria in the expense assessment process. Although the knowledge content they contain is different, they jointly determine whether the accommodation expense meets the requirements, thus forming a complementary relationship between them.

[0087] Based on the established content complementarity relationships, the task support relationships between knowledge representation objects are further determined according to the task processing objectives. These task support relationships represent the supporting role of knowledge representation objects in achieving the current task processing objective. Specifically, based on the processing procedures corresponding to the task processing objective, it can be determined whether each knowledge representation object participates in the task judgment, condition matching, or result determination process. If a knowledge representation object can provide the judgment basis, analysis basis, or result generation basis required in the task processing process, then a task support relationship can be established between that knowledge representation object and the current task processing objective.

[0088] For example, in the task of reimbursing accommodation expenses, the knowledge representation object of accommodation standards is used to provide the basis for judging the expense standards, the knowledge representation object of regional limits is used to provide the basis for judging regional restrictions, and the knowledge representation object of approval requirements is used to provide the basis for judging the process. The above knowledge representation objects participate in different judgment stages in the reimbursement review process, thus forming a supporting relationship for the task of reviewing accommodation expenses.

[0089] Preferably, after determining the content supplementation relationship and task support relationship, the combination relationship between multiple knowledge representation objects can be further analyzed based on the knowledge integrity requirements corresponding to the task processing objective, in order to determine the combination of knowledge representation objects that can jointly cover the current task requirements. Specifically, based on the knowledge content range required to be covered by the task processing objective, it can be analyzed whether there are any missing content areas in the currently determined knowledge representation objects, and the corresponding supplementation relationship can be determined based on the knowledge content provided by other knowledge representation objects, thereby selecting knowledge representation objects that can jointly cover the knowledge content range.

[0090] For example, in the task of equipment fault analysis, if one knowledge representation object only contains fault phenomenon information and another knowledge representation object contains equipment operating parameter information, then the two provide fault manifestation and fault judgment basis respectively. Neither can complete the fault analysis task when used alone. Therefore, based on the content supplementation relationship between the two and their supporting role in the fault analysis task, they can be jointly identified as a combination of knowledge representation objects participating in the construction of dynamic ontology objects.

[0091] In this embodiment, the combination of knowledge representation objects is not determined directly based on the similarity between knowledge representation objects, document position relationship, or text association. Instead, it is determined based on the task processing goal to see if different knowledge representation objects can jointly provide the knowledge content required to complete the task. The combination of knowledge representation objects is determined based on the content supplementation relationship and task support relationship, so that the target object unit formed subsequently can adapt to the specific task requirements.

[0092] Preferably, based on the above embodiments, this application also provides a specific implementation method for forming target object units by combining knowledge representation objects.

[0093] In this embodiment, after determining the content supplementation relationship and task support relationship among multiple knowledge representation objects according to task requirements, the combination of knowledge representation objects participating in the construction of the dynamic ontology object is further determined based on the above relationships, and a target object unit is formed based on the combination of knowledge representation objects. It should be noted that the target object unit is not a simple collection of multiple knowledge representation objects, but rather a knowledge organization result formed by associating and organizing knowledge representation objects that can jointly meet task requirements under the constraints of a specific task processing goal.

[0094] Specifically, after determining the combination of knowledge representation objects, the combination scope is first determined based on the knowledge content coverage corresponding to the combination of knowledge representation objects. The combination scope is used to represent the range of knowledge content that the current combination of knowledge representation objects needs to cover, and it is used to limit the types of knowledge content that need to be included when forming the target object unit.

[0095] Specifically, based on the knowledge completeness requirements corresponding to the current task processing objective, the knowledge content required to complete the task can be analyzed, and the scope that can collectively cover the knowledge content can be determined as the combined scope. For example, in the task of reimbursing accommodation expenses, if the task processing objective is to determine whether the accommodation expenses meet the reimbursement requirements, then completing this task requires at least knowledge content such as the standards corresponding to the personnel level, the limits corresponding to the region, and the approval requirements. Therefore, the scope of the above-mentioned content can be determined as the combined scope corresponding to this task.

[0096] Furthermore, the composition boundary corresponding to the combination of knowledge representation objects is determined based on the aforementioned combination range. The composition boundary is used to limit the scope of knowledge representation objects participating in the formation of the target object unit, that is, to determine which knowledge representation objects need to be included in the current target object unit, and which knowledge representation objects, although related, do not belong to the knowledge scope required by the current task and therefore do not need to participate in the combination.

[0097] Specifically, the members of a knowledge representation object combination can be determined based on the relationship between each knowledge representation object and the current task processing objective, as well as whether the knowledge content provided by each knowledge representation object falls within the scope of the combination. For example, in the accommodation expense reimbursement task, the accommodation standard knowledge representation object, the regional limit knowledge representation object, and the approval requirement knowledge representation object can all provide the knowledge content required to complete the reimbursement review, and therefore are included within the composition boundary. However, although the transportation expense standard knowledge representation object also belongs to the travel policy document, it is not included in the current target object unit because it does not participate in the accommodation expense review process.

[0098] After determining the scope and boundaries of the combination, the knowledge representation objects can be associated and organized according to these boundaries to form target object units. Specifically, based on the established content supplementation relationships between knowledge representation objects, different knowledge contents can be linked and organized so that the knowledge contents contained in different knowledge representation objects can jointly cover the current task requirements. At the same time, based on the task support relationships between knowledge representation objects, the organization method of different knowledge representation objects in the target object unit can be determined so that the formed target object unit can reflect the knowledge relationships required for the current task processing.

[0099] For example, in the scenario of accommodation expense reimbursement, the knowledge representation objects corresponding to employee levels, regional accommodation limits, and approval requirements can be linked and organized to form a target object unit to support the accommodation expense review task. Among them, the accommodation standard knowledge representation object provides the basis for judging personnel qualifications, the regional limit knowledge representation object provides the basis for judging expense restrictions, and the approval requirement knowledge representation object provides the basis for judging the process. Together, the three constitute the complete knowledge content for reimbursement review.

[0100] Preferably, after forming the target object unit, the organization method of the knowledge content in the target object unit can be adjusted according to the content association relationship between the knowledge expression objects that form the target object unit.

[0101] Specifically, since different knowledge representation objects originate from different content locations within the document to be processed, their original arrangement order may differ from the order in which they are used during the current task processing. Therefore, it is possible to further analyze the content relationships between knowledge representation objects, determine the connection methods between different knowledge contents, and adjust the organization of knowledge content within the target object unit based on these connection methods.

[0102] For example, in the task of reimbursing accommodation expenses, the knowledge content in the target object unit can be organized according to the process of "determining personnel conditions - matching expense standards - judging reimbursement conditions - determining approval requirements". This will ensure that when a dynamic ontology object is subsequently generated based on the target object unit, the knowledge fields can maintain an association relationship that matches the task processing process.

[0103] The target object unit formed in the above manner is not a mechanical splicing of multiple knowledge representation objects, but rather a task-adaptive knowledge organization result formed under the constraints of task processing objectives, based on the coverage of knowledge content, the composition range of knowledge representation objects, and the relationships between knowledge representation objects. This target object unit not only retains the knowledge content corresponding to each knowledge representation object, but also reflects the organizational relationship in which different knowledge contents jointly serve the current task processing objective, providing a foundation for the subsequent generation of dynamic ontology objects.

[0104] Preferably, based on the above embodiments, this application also provides a specific implementation method for adjusting dynamic ontology objects according to changes in task requirements or knowledge content.

[0105] In this embodiment, since the task processing objectives may change during actual application, or some knowledge content in the document to be processed may change, the target object unit and the corresponding dynamic ontology object originally formed according to a certain task requirement may no longer meet the new task processing requirements. Therefore, this application does not directly replace or delete the already generated knowledge fields, but re-analyzes the content supplementation relationship and task support relationship between knowledge expression objects based on the changed task requirement information or the newly added or changed knowledge expression objects, in order to determine whether the original target object unit needs to be adjusted.

[0106] Specifically, when acquiring changed task requirements, the scope of knowledge content to be covered by the current task can be redefined based on the task processing objectives and knowledge usage requirements corresponding to the changed task requirements. It can also be determined whether the knowledge representation objects contained in the existing target object units can still meet the new task processing objectives. For example, in a accommodation expense reimbursement scenario, the original task requirement information corresponds to the task processing objective of "determining whether accommodation expenses meet reimbursement requirements." Therefore, the resulting target object units include knowledge representation objects such as accommodation standards, regional limits, reimbursement conditions, and approval requirements. When the new task requirement information changes to "querying employee accommodation standards in a specified region," the new task processing objective only needs to acquire the regional standard information and the personnel level standard information. In this case, the approval requirement knowledge representation object in the original target object unit no longer falls within the scope of knowledge required by the current task. At this point, the combination of knowledge representation objects can be redefined based on the new task processing objective, and a new target object unit can be formed.

[0107] Furthermore, when acquiring new knowledge representation objects, the content supplementation and task support relationships between the new knowledge representation object and existing knowledge representation objects can be re-analyzed based on the knowledge content it contains. Specifically, it can be determined whether the knowledge content contained in the new knowledge representation object belongs to the knowledge content range corresponding to the current task processing target, and whether the new knowledge representation object can supplement the missing knowledge content in the existing target object unit. If the new knowledge representation object can supplement the knowledge content required by the current task and can support the task processing process, then the new knowledge representation object can be included in the knowledge representation object combination and a new target object unit can be formed.

[0108] For example, in a travel expense review scenario, the original target object unit contains knowledge representation objects such as accommodation standards, regional limits, and reimbursement conditions. When a new knowledge representation object about travel approval rules for special regions is added, if the current task requires reviewing accommodation expenses in special regions, the new knowledge representation object can supplement the approval judgment content in the original target object unit. Therefore, based on its content supplementation relationship and task support relationship with existing knowledge representation objects, it can be incorporated into a new knowledge representation object combination.

[0109] Preferably, after redefining the knowledge representation object combination, the original target object units can be adjusted based on the newly determined content supplementation relationships and task support relationships. Specifically, the affected target object units and their corresponding change ranges can be determined based on the changed knowledge representation objects. The change range is used to represent the range of knowledge content that needs to be adjusted in the target object unit due to the change in the knowledge representation objects. Through the above adjustment process, the dynamic ontology object can evolve with changes in capability requirements and knowledge content, rather than maintaining a fixed structure after initial construction.

[0110] For example, when the accommodation standard in a knowledge representation object changes, we can determine whether the change only affects the accommodation standard field, or whether it will further affect the judgment of regional limits, reimbursement conditions, and approval requirements, based on the content supplementation and task support relationships between this knowledge representation object and other knowledge representation objects. If the change only involves the accommodation standard content, then only the corresponding knowledge content needs to be adjusted; if the change causes a change in the original judgment relationships, then the target object unit needs to be re-formed.

[0111] Specifically, after identifying the knowledge representation object that has changed, the scope of impact can be further determined based on the content supplementation relationship and task support relationship corresponding to that knowledge representation object. The content supplementation relationship is used to determine whether the changed content will cause the knowledge content provided by other knowledge representation objects to be missing or require re-matching; the task support relationship is used to determine whether the changed content will affect the judgment basis or processing result in the current task processing process.

[0112] For example, when a company adjusts its accommodation reimbursement standards, the change first affects the knowledge representation objects of the accommodation standards. If the new standard changes cause the original reimbursement judgment rules to become invalid, it is necessary to further adjust the knowledge representation objects of the reimbursement conditions and the knowledge representation objects of the approval requirements associated with the accommodation standards, so as to re-form the target object units that are adapted to the new reimbursement rules.

[0113] After adjusting the target object units, the corresponding dynamic ontology objects can be updated based on the newly formed target object units. Specifically, the changed knowledge content can be determined first based on the differences between the target object units before and after the adjustment. Then, the knowledge fields that need to be adjusted can be determined based on the changed knowledge content, and the original dynamic ontology objects can be updated. For example, when a new knowledge representation object for special area approval requirements is added, an approval requirement field can be added to the original dynamic ontology object; when accommodation standards change, the field content corresponding to the accommodation standard field can be updated; when a knowledge representation object no longer participates in the current task processing, its corresponding knowledge field can be deleted, and its association with the target object unit can be severed.

[0114] Through the above methods, this application embodiment achieves dynamic adjustment of the dynamic ontology object based on changes in task requirements and knowledge content. Unlike data update methods that directly modify field content, this application first re-analyzes the combination relationships between knowledge representation objects, then adjusts the target object unit, and finally updates the dynamic ontology object, thereby ensuring that the dynamic ontology object always remains consistent with the current task processing target and the knowledge content organization relationship.

[0115] Preferably, based on the above embodiments, this application also provides a specific implementation method for forming knowledge fields according to the knowledge content scope and field organization requirements corresponding to the target object unit.

[0116] In this embodiment, after forming the target object unit, the knowledge content within the target object unit is further structured according to the scope of knowledge content contained in the target object unit and the field organization requirements. It should be noted that the knowledge fields in this embodiment are not directly extracted from the text content of the document to be processed, but rather, based on the target object unit already determined according to the task requirements, the knowledge content in the target object unit related to the current task processing is converted into a structured expression.

[0117] Specifically, the knowledge content to be expressed as fields is first determined based on the knowledge content scope corresponding to the target object unit. The knowledge content scope represents the different types of knowledge content contained within the target object unit, and the role of each knowledge content in the current task processing. Specifically, the information type to be converted into fields can be determined based on the knowledge content corresponding to each knowledge expression object forming the target object unit. For example, in the target object unit corresponding to the accommodation expense reimbursement task, which is formed by accommodation standard knowledge expression objects, regional limit knowledge expression objects, reimbursement condition knowledge expression objects, and approval requirement knowledge expression objects, the corresponding knowledge content scope can be determined to include personnel level information, regional restriction information, expense standard information, and approval requirement information.

[0118] Furthermore, the field division method corresponding to the knowledge content is determined according to the field organization requirements. The field organization requirements represent the organization rules used when converting knowledge content in the target object unit into knowledge fields, including the knowledge content type corresponding to the field, the association method between fields, and the field expression form. Specifically, corresponding fields can be determined based on the role of different knowledge content in the target object unit. For example, for knowledge content used to represent personnel condition judgments, a corresponding personnel level field can be set; for knowledge content used to represent cost restriction conditions, a corresponding cost limit field can be set; and for knowledge content used to represent processing flow requirements, a corresponding approval requirement field can be set.

[0119] By using the above method, we can avoid generating fields directly based on keywords in the text. Instead, we can divide the fields according to the task-related knowledge content that has been determined in the target object unit, so that the resulting knowledge fields can reflect the knowledge structure required by the current task.

[0120] Preferably, when forming knowledge fields based on the scope of knowledge content and field organization requirements, the correspondence between fields can also be determined based on the organizational relationships between different knowledge contents. For example, in the scenario of accommodation expense reimbursement, there is a correspondence between the personnel level field and the accommodation standard field because the accommodation standard needs to be determined based on the personnel level; there is a correspondence between the region field and the expense limit field because the expense limit varies according to the business trip region. By retaining the above correspondence between fields, the generated dynamic ontology object can support subsequent knowledge invocation and processing based on the relationships between fields.

[0121] In this embodiment, the field division basis is determined based on the scope of knowledge content in the target object unit, and the field expression method is determined according to the field organization requirements, thereby converting the knowledge content formed by task-oriented organization into knowledge fields with clear structural relationships.

[0122] Preferably, based on the above embodiments, this application also provides a specific implementation method for establishing the association between knowledge fields and target object units.

[0123] In this embodiment, after forming knowledge fields, the association between knowledge fields and target object units is further established to retain the correspondence between knowledge fields and target object units. This ensures that the generated dynamic ontology object not only contains the fieldized knowledge content but also reflects the relationship between fields and the knowledge organization results formed by task-driven processes. Specifically, the knowledge representation object from which a knowledge field originates can be determined based on the field content corresponding to that knowledge field, and the compositional position of that knowledge representation object within the target object unit can be further determined, thereby establishing the association between knowledge fields and target object units.

[0124] For example, in the task of reimbursing accommodation expenses, the target object unit consists of an accommodation standard knowledge representation object, a regional limit knowledge representation object, and an approval requirement knowledge representation object. When an accommodation standard field is formed, its source can be determined based on the knowledge content corresponding to the field, and a relationship can be established between the accommodation standard field and the corresponding target object unit. When an approval requirement field is formed, its source can be determined based on the source of the field content, and a corresponding relationship can be established.

[0125] Furthermore, by establishing the aforementioned relationships, the traceability of the knowledge content source can be maintained for dynamic ontology objects during invocation or adjustment. For example, when the content in the accommodation standard knowledge representation object changes, the affected knowledge fields can be determined based on the relationship between the knowledge fields and the target object unit, and the corresponding dynamic ontology object can be adjusted accordingly.

[0126] It should be noted that the association relationship in this embodiment is not a simple record of the correspondence between field names and field values, but rather a representation of the organizational relationship between knowledge fields and target object units formed by task requirements. This association relationship preserves the evolutionary process from knowledge representation objects to target object units and then to knowledge fields. For example, in the accommodation expense reimbursement scenario, the generated dynamic ontology object not only includes field information such as "accommodation standard field," "regional limit field," and "approval requirement field," but also, through the association relationship, can determine that these fields all originate from the target object unit formed for the accommodation expense review task, thereby maintaining the correspondence between the dynamic ontology object and the task processing target.

[0127] Through the above methods, this application embodiment establishes the association between knowledge fields and target object units, enabling dynamic ontology objects to not only realize the field-based expression of knowledge content, but also retain task association information in the knowledge content organization process, providing a foundation for subsequent knowledge object updates, adjustments and applications.

[0128] Preferably, based on the above embodiments, this application also provides a specific implementation method for constructing dynamic ontology objects according to multiple task requirement information.

[0129] In this embodiment, after parsing the document to be processed and forming multiple knowledge representation objects through step S1, the corresponding combination of knowledge representation objects can be determined for multiple different task requirements, and the corresponding target object unit can be formed based on the different combinations of knowledge representation objects, thereby generating a dynamic ontology object that adapts to different task processing targets.

[0130] It should be noted that since different task requirements correspond to different task processing objectives and knowledge usage needs, even when faced with the same document to be processed and the same set of knowledge representation objects, it is not necessary to form the same knowledge object structure. This application enables the knowledge representation objects to be recombinated according to the actual task needs by analyzing the knowledge content requirements corresponding to different task requirements separately. Specifically, after obtaining multiple task requirement information, the task processing objectives and knowledge usage needs corresponding to each task requirement information can be analyzed separately to determine the scope of knowledge content to be covered by each task processing objective, and the corresponding knowledge representation objects to participate in the combination can be determined from multiple knowledge representation objects based on the knowledge content scope.

[0131] For example, in a corporate travel management scenario, the documents to be processed may include multiple document contents such as travel management regulations, accommodation standard descriptions, and approval rule descriptions. Step S1 can generate multiple knowledge representation objects, such as accommodation standard knowledge representation objects, regional limit knowledge representation objects, approval requirement knowledge representation objects, and transportation expense knowledge representation objects. When the task processing objective corresponding to the first task requirement information is "determine whether employee accommodation expenses meet reimbursement requirements," based on the knowledge usage requirements corresponding to this task processing objective, it is necessary to obtain knowledge content such as personnel conditions, accommodation standards, regional restrictions, and approval requirements. Therefore, the accommodation standard knowledge representation object, regional limit knowledge representation object, and approval requirement knowledge representation object can be selected to form a corresponding knowledge representation object combination, further forming a target object unit to support the accommodation expense review task. When the task processing objective corresponding to the second task requirement information is "query accommodation expense standards for a specified region," this task only needs to obtain the accommodation standard information corresponding to the region. Therefore, the regional limit knowledge representation object and accommodation standard knowledge representation object can be selected to form a new knowledge representation object combination, forming a target object unit to support the standard query task. This shows that the same batch of knowledge representation objects does not correspond to a single fixed combination, but is reorganized according to the task processing objectives and knowledge usage requirements corresponding to different task needs.

[0132] Furthermore, when determining the combination of knowledge representation objects based on different task requirements, the scope of the combination can be adjusted according to the knowledge content requirements corresponding to each task. Specifically, based on the scope of knowledge content to be covered for different task processing objectives, it can be determined whether the existing knowledge representation objects can meet the corresponding task requirements, and the knowledge representation objects that need to participate in the combination can be determined based on the content supplementation relationship and task support relationship between the knowledge representation objects. For example, for the task of reviewing accommodation fees, it is necessary to consider personnel level, regional standards, and approval requirements simultaneously. Therefore, the combination scope needs to cover the knowledge content related to personnel condition judgment, fee standard judgment, and process judgment. However, for the task of querying accommodation standards, it is only necessary to determine the standard information corresponding to the region and personnel level. Therefore, the combination scope can be narrowed, reducing the participation of knowledge representation objects related to the approval process.

[0133] After determining the adjusted combination range, the corresponding composition boundaries can be further determined. These composition boundaries define the scope of specific knowledge representation objects that participate in forming the target object unit under different task requirements. Specifically, it can be determined whether each knowledge representation object is a component of the current target object unit based on its supporting role in the current task processing objective. Knowledge representation objects that can supplement the knowledge content required by the current task and participate in the task processing process can be included in the composition boundaries; knowledge representation objects that, although related, do not participate in the current task processing process are not included in the composition boundaries.

[0134] For example, in the task of reviewing accommodation expenses, the knowledge representation object required by the approval criteria must fall within the composition boundary because it participates in the review process for accommodation expenses exceeding the standard. However, while the knowledge representation object for transportation expense standards also belongs to travel management documents, it does not participate in the accommodation expense review process and therefore does not fall within the composition boundary of this target object unit. Furthermore, based on the composition boundaries corresponding to different task requirements, knowledge representation objects can be associated and organized to form multiple different target object units. Then, based on the knowledge content scope and field organization requirements of each target object unit, a corresponding dynamic ontology object can be generated.

[0135] Through the above methods, the embodiments of this application realize the knowledge object construction process for multiple task requirements, so that the knowledge expression object, as the basic knowledge organization unit, remains stable, while the target object unit and dynamic ontology object can be dynamically formed according to different task processing objectives.

[0136] It should be noted that the multi-task construction process in this embodiment does not involve repeatedly performing field extraction operations on the same document. Instead, it redefines the knowledge organization relationships based on the existing knowledge representation object and the specific task requirements. This approach avoids the problem of fixed knowledge structures being unable to adapt to different application tasks, allowing the generated dynamic ontology object to better suit the specific task processing needs.

[0137] Preferably, to further illustrate the dynamic ontology object construction process based on capability issues in this application, the following example is a corporate travel expense reimbursement scenario.

[0138] In this embodiment, enterprises typically store a large number of documents related to travel expense management, such as "Employee Travel Management System," "Accommodation Expense Reimbursement Standards," "Regional Travel Subsidy Regulations," and "Expense Approval Process Instructions." These documents contain knowledge content such as accommodation standards corresponding to employee levels, accommodation limits for different regions, approval requirements for special circumstances, reimbursement conditions, and expense review processes. Since these documents are usually organized according to the organizational methods used during the system's formulation and release, rather than according to specific reimbursement tasks, the knowledge content related to a single accommodation expense review task may be scattered across different documents or different chapters of the same document. For example, the "Employee Travel Management System" may stipulate the basic principles of accommodation expense reimbursement, the "Accommodation Expense Reimbursement Standards" may stipulate accommodation amount limits corresponding to different employee levels, the "Regional Travel Subsidy Regulations" may stipulate accommodation limits corresponding to different city levels, and the "Expense Approval Process Instructions" may stipulate the approval process for exceeding the standards. If knowledge objects are generated directly according to the document chapters, the generated knowledge objects only reflect the original organizational form of the system documents and cannot directly support the accommodation expense review task.

[0139] First, obtain the aforementioned travel management-related documents as pending documents and analyze their content expression relationships, structural organization relationships, and knowledge association relationships. Specifically, the scope of different knowledge content can be determined based on the content descriptions in the documents. For example, the description of "accommodation amount corresponding to employee level" in the "Accommodation Expense Reimbursement Standard" is used to express the correspondence between personnel conditions and expense standards; the description of "accommodation limit corresponding to city category" in the "Regional Travel Subsidy Regulations" is used to express the correspondence between regional conditions and expense limits; and the description of "accommodation exceeding the standard requires approval" in the "Expense Approval Process Instructions" is used to express special processing requirements.

[0140] Furthermore, based on the chapter organization, heading relationships, and content citation relationships within the document, the structural organizational relationships between different knowledge contents can be determined. For example, the "Employee Travel Management System," as a general rule document, includes chapters on accommodation management, transportation management, and expense approval, thus establishing the hierarchical relationship between the general rules and specific rules. Similarly, the "Expense Approval Process Description," which uses accommodation standards as the basis for approval decisions, establishes an organizational connection between the approval process content and the accommodation standard content.

[0141] Furthermore, knowledge associations are determined based on the correlation between the objects described and the processing procedures of different knowledge contents. For example, employee grade information and accommodation standard information are both used to determine the scope of accommodation that employees can be reimbursed, and there is a knowledge association between the two; accommodation standard information and approval process information have a processing association because the approval process needs to judge whether the accommodation standard is exceeded.

[0142] Based on the aforementioned content expression relationships, structural organization relationships, and knowledge association relationships, the knowledge content in travel management documents can be reorganized into multiple knowledge expression objects. For example, this could result in: a first knowledge expression object representing the correspondence between employee levels and accommodation standards; a second knowledge expression object representing the correspondence between business trip regions and accommodation limits; a third knowledge expression object representing the conditions for accommodation expense reimbursement; and a fourth knowledge expression object representing the approval process for accommodation expenses exceeding the standard. These knowledge expression objects do not simply correspond to chapters in the original document, but are formed based on the scope of knowledge content and the relationships between different knowledge contents, ensuring that each knowledge expression object can carry relatively complete knowledge content.

[0143] When an employee submits a request for reimbursement of accommodation expenses, the corresponding task requirement information is obtained. In this embodiment, the task requirement information indicates the accommodation expense review task that needs to be completed and the knowledge content required to complete the task. The task processing objective is to determine whether the accommodation expense submitted by the employee meets the company's reimbursement requirements. The knowledge usage requirement is to combine employee identity information, business trip location information, accommodation expense information, and approval rule information to complete the review and judgment.

[0144] Based on the aforementioned task processing objectives and knowledge usage requirements, the scope of knowledge content required to complete the accommodation expense review task can be determined, and relevant knowledge representation objects can be selected from multiple knowledge representation objects. Specifically, the first knowledge representation object corresponding to employee level and accommodation standard is used to provide a basis for judging personnel qualifications; the second knowledge representation object corresponding to business trip region and accommodation limit is used to provide a basis for judging regional restrictions; the third knowledge representation object corresponding to accommodation expense reimbursement conditions is used to provide a basis for judging expense compliance; and the fourth knowledge representation object corresponding to the approval process is used to provide a basis for judging the process in special circumstances.

[0145] Furthermore, based on the knowledge content and task processing procedures provided by the aforementioned knowledge representation objects, the content supplementation relationships and task support relationships among the knowledge representation objects are determined. Specifically, the first knowledge representation object provides employee level corresponding standard information, the second knowledge representation object provides regional restriction information, and both supplement the basis for judging accommodation expenses from the personnel and regional dimensions, respectively; the third knowledge representation object provides reimbursement condition judgment rules, and the fourth knowledge representation object provides approval basis for exceeding standards. Therefore, there is a content supplementation relationship among the aforementioned knowledge representation objects. Simultaneously, since the first, second, third, and fourth knowledge representation objects participate in employee identity matching, expense standard judgment, compliance judgment, and approval judgment processes, respectively, these knowledge representation objects collectively form a task support relationship for the accommodation expense review task. Based on the aforementioned content supplementation relationships and task support relationships, the knowledge representation object combination participating in the construction of the dynamic ontology object can be determined, that is, the first, second, third, and fourth knowledge representation objects are determined as the knowledge representation object combination corresponding to the current accommodation expense review task.

[0146] Furthermore, the scope of the combination is determined based on the range of knowledge content required to cover the accommodation expense review task. In this embodiment, the scope of the combination includes the scope of personnel condition judgment, the scope of regional standard judgment, the scope of expense compliance judgment, and the scope of approval process judgment. Simultaneously, the composition boundaries are determined based on the role of each knowledge representation object in the current task. For example, although knowledge representation objects related to transportation expenses also belong to travel management documents, they do not participate in the accommodation expense judgment process since the current task only focuses on accommodation expense review; therefore, they are not within the composition boundary of the current knowledge representation object combination. After determining the scope of the combination and the composition boundaries, the knowledge representation objects are associated and organized according to the content supplementation relationship and task support relationship between them, forming a target object unit corresponding to the accommodation expense review task. This target object unit contains knowledge content such as employee level, accommodation standard, regional limit, reimbursement conditions, and approval requirements, and is organized according to the accommodation expense review process, enabling different knowledge representation objects to jointly support the current task processing.

[0147] Furthermore, based on the scope of knowledge content and field organization requirements corresponding to the target object unit, the target object unit is structured. Specifically, corresponding knowledge fields can be formed based on the knowledge content in the target object unit, including: an employee level field, used to represent the applicant's corresponding level information; an accommodation standard field, used to represent the standard amount corresponding to different levels; a business trip region field, used to represent the region information corresponding to the accommodation location; a region limit field, used to represent the maximum amount allowed for reimbursement in the corresponding region; a reimbursement condition field, used to indicate whether the expenses meet the reimbursement requirements; and an approval requirement field, used to represent the approval process that needs to be executed in special circumstances.

[0148] Simultaneously, based on the source of the knowledge content corresponding to each knowledge field, a relationship is established between the knowledge field and the target object unit. For example, the accommodation standard field originates from the first knowledge representation object, the regional quota field originates from the second knowledge representation object, and the approval requirements field originates from the fourth knowledge representation object. Through the above relationships, it is possible to determine which target object unit's content each field in the dynamic ontology object originates from, enabling accurate location of the affected knowledge content during subsequent updates.

[0149] When a company adjusts its accommodation expense standards, it can obtain the changed knowledge content information. For example, if a company adjusts the accommodation standard corresponding to a certain employee level, it first identifies the changed knowledge representation object as the accommodation standard knowledge representation object, and then determines the scope of the impact based on the content supplementation relationship and task support relationship between this knowledge representation object and other knowledge representation objects. If the adjustment only affects the accommodation standard field, the corresponding field content is updated; if the adjustment causes changes to the original reimbursement judgment rules, such as affecting the approval conditions for exceeding the standard, the knowledge content corresponding to the approval requirements is further adjusted, and a new target object unit is formed.

[0150] Furthermore, for different task requirements, different dynamic ontology objects can be formed based on the same knowledge representation object. For example, when the new task requirement is "query the accommodation standards for employees in a certain region," the task processing objective is only to obtain the corresponding accommodation standard information for the specified region. Therefore, it is not necessary to call the approval requirement knowledge representation object. Instead, the employee level, accommodation standard knowledge representation object, and regional quota knowledge representation object can be selected to form a new target object unit, and a dynamic ontology object for accommodation standard query can be generated. When the new task requirement is "analyze the company's accommodation expense situation," relevant knowledge representation objects for expense statistics can be added according to this task requirement, the combination of knowledge representation objects can be redefined, and a dynamic ontology object for expense analysis task can be formed.

[0151] As can be seen from the above application scenarios, this application does not directly convert travel policy documents into a fixed field structure. Instead, it parses and forms knowledge representation objects, determines the combination relationships between these objects based on different task requirements, and further forms target object units and dynamic ontology objects. In this way, the same batch of document knowledge content can form different dynamic ontology objects according to different task processing objectives, transforming the knowledge organization process from a fixed conversion method oriented towards document structure to a dynamic construction method oriented towards task requirements.

[0152] Preferably, based on the above embodiments, the specific process of performing the accommodation fee review task based on the dynamic ontology object is further described.

[0153] In this embodiment, when an employee submits a reimbursement request for accommodation expenses, the corresponding task requirement information is obtained. This task requirement information specifically reflects the business capability issue of completing a compliance assessment of accommodation expenses. The reimbursement request information may include applicant information, business trip location information, accommodation expense information, and accommodation duration information. For example, an employee submits a reimbursement request for accommodation expenses, which includes the applicant's level being "department head," the business trip location being a "Category A city," the accommodation expense amount being 1500 yuan, and the accommodation duration being 2 days. For this reimbursement request task, the dynamic ontology object formed for the aforementioned accommodation expense review task can be invoked.

[0154] Specifically, firstly, the personnel level information corresponding to the applicant is obtained based on the employee level field in the dynamic ontology object. Then, based on the association between this field and the target object unit, it is determined that this personnel level information originates from the knowledge representation object corresponding to the employee level and accommodation standard. Further, based on the knowledge content corresponding to the employee level field and the knowledge content corresponding to the accommodation standard field, the accommodation standard information corresponding to that employee level is determined. For example, based on the accommodation standard field in the dynamic ontology object, the daily accommodation standard for the department head is determined to be 800 yuan. Simultaneously, the region information corresponding to the current business trip region is obtained based on the business trip region field, and the corresponding accommodation fee limit for that region is determined based on the region limit field. For example, based on the region limit field, the accommodation fee limit for Category A cities is determined to be 900 yuan per day. Further, the actual accommodation fee amount submitted by the employee is obtained based on the accommodation fee field, and the corresponding average daily accommodation fee per person is calculated based on the number of nights stayed. For example, if an employee submits an accommodation fee of 1500 yuan for 2 nights, the corresponding daily accommodation fee is 750 yuan. Since 750 yuan is lower than the accommodation standard of 800 yuan for employee level and the regional limit of 900 yuan, it can be determined that the accommodation fee meets the fee standard requirements.

[0155] In the above judgment process, the dynamic ontology object does not simply provide a few independent fields, but rather, through the association between fields and target object units, different fields can collectively correspond to the knowledge organization results corresponding to the accommodation expense review task. For example, when the accommodation standard field changes, the association between this field and the target object unit can be used to determine that the change originates from the accommodation standard knowledge representation object, and further determine whether the change affects the judgment of regional limits, reimbursement conditions, and approval requirements.

[0156] If a company adjusts the accommodation standard for department heads, raising it from 800 yuan to 1000 yuan per day, the affected knowledge field can be determined as the accommodation standard field based on the changed knowledge content, and the corresponding dynamic ontology object can be updated. If this adjustment also causes changes to the approval rules for exceeding the standard, the relevant knowledge fields for approval requirements can be further adjusted based on the task support relationships between knowledge representation objects.

[0157] Furthermore, when an enterprise adds new travel management rules for special regions, it can obtain the knowledge representation object corresponding to the new rule and analyze the content supplementation relationship and task support relationship between this knowledge representation object and the existing accommodation standard knowledge representation object and regional limit knowledge representation object. If the new rule is used to supplement the accommodation expense review requirements for special regions and can participate in the accommodation expense review task, the new knowledge representation object can be included in the corresponding knowledge representation object combination to re-form the target object unit and generate an updated dynamic ontology object.

[0158] For example, if a company adds a new management rule that "accommodation fees in high-altitude areas can be increased by 20%", a corresponding special area rule knowledge representation object can be formed. Since this knowledge representation object can supplement the special area judgment content in the original area limit knowledge representation object and participate in the accommodation fee review process, it can be determined that there is a content supplementation relationship and a task support relationship between the two, and the original target object unit can be adjusted.

[0159] Through the above methods, this embodiment demonstrates the application process of dynamic ontology objects in actual task processing. It can be seen that the dynamic ontology objects generated in this application are not fixed-stored document summaries or field sets, but rather knowledge representation results obtained by further structuring target object units based on task requirements. They can maintain the correspondence between knowledge content, task requirements, and field representations.

[0160] As can be seen from the above embodiments, this application does not directly extract fields or convert text content in the document to be processed in a fixed form, but rather forms multiple knowledge expression objects based on parsing the content expression relationships, structural organization relationships, and knowledge association relationships in the document to be processed.

[0161] In this context, the knowledge representation object is used to characterize knowledge units within a document that possess a relatively complete range of knowledge content. This allows continuous content within the document to be transformed into basic knowledge organizational units with clearly defined knowledge boundaries. This processing method avoids the problem of accurately determining the required knowledge content in subsequent task processing due to differences in document chapter division methods, writing order, and content arrangement.

[0162] Furthermore, this application analyzes the content supplementation relationships and task support relationships among multiple knowledge representation objects based on the task processing objectives and knowledge usage requirements corresponding to the task requirement information. The content supplementation relationship determines whether different knowledge representation objects can jointly cover the scope of knowledge content required for the current task, while the task support relationship determines whether different knowledge representation objects can participate in the current task processing. By considering both types of relationships simultaneously, the problem of selecting knowledge content solely based on text similarity, document position relationships, or fixed knowledge associations can be avoided, ensuring that the knowledge representation objects participating in the construction of the dynamic ontology object truly adapt to the current task processing objectives.

[0163] Furthermore, this application determines the combination of knowledge representation objects based on content supplementation relationships and task support relationships, and forms target object units according to the combination scope and composition boundaries. Since the target object unit is a knowledge organization result formed under the constraints of task requirements, it is not a simple summary of the original knowledge content in the document to be processed, nor a mechanical splicing of multiple knowledge representation objects. Instead, it is a reorganization of knowledge representation objects that can jointly provide knowledge support according to the needs of the current task. In this way, the knowledge content in the target object unit can maintain a correspondence with the current task processing objective.

[0164] Furthermore, based on the knowledge content scope and field organization requirements corresponding to the target object unit, this application performs structured processing on the target object unit to form knowledge fields, and establishes the association between the knowledge fields and the target object unit. By establishing the above association, the generated dynamic ontology object not only contains the fieldized knowledge content, but also retains the correspondence between the knowledge fields and the target object unit formed by task requirements. When subsequent task requirements change or knowledge content changes, the scope of affected knowledge content can be quickly determined based on this association, and the target object unit and the corresponding dynamic ontology object can be readjusted.

[0165] For example, in the scenario of accommodation expense reimbursement, when a company adjusts the accommodation allowance standard for a certain region, this application does not directly modify the content of a specific field. Instead, it determines the corresponding knowledge representation object based on the changed knowledge content, and combines the content supplementation relationship and task support relationship between this knowledge representation object and other knowledge representation objects to determine whether the change affects the target object unit corresponding to the current accommodation expense review task. If the scope of impact changes, the target object unit is re-formed, and the corresponding dynamic ontology object is updated.

[0166] Furthermore, this application supports the creation of different dynamic ontology objects for multiple different task requirements. For example, for the same batch of travel management documents, a dynamic ontology object containing accommodation standards, regional limits, reimbursement conditions, and approval requirements can be created for an accommodation standard query task; a dynamic ontology object containing regional and standard information can be created for an accommodation standard query task; and a new dynamic ontology object can be created by combining expense statistics-related knowledge content for an expense analysis task. Thus, the knowledge object construction process in this application is not a fixed transformation centered on documents, but rather driven by task requirements, reorganizing knowledge content according to different task processing objectives, so that knowledge content from the same source can adapt to different application tasks.

[0167] Through the above technical solution, this application has at least the following technical effects: First, this application analyzes the content expression relationship, structural organization relationship, and knowledge association relationship to form a knowledge expression object, enabling the knowledge content in the document to break through the limitations of the original chapter structure and be organized in a form with clear knowledge boundaries, thus providing a foundation for subsequent task-driven knowledge combination.

[0168] Second, this application determines the content supplementation relationship and task support relationship by combining task processing objectives and knowledge usage needs, which transforms the knowledge content selection process for constructing dynamic ontology objects from being driven by fixed rules to being driven by task requirements, thereby improving the matching degree between the knowledge organization results and the actual task processing process.

[0169] Third, by forming target object units and further processing them into fields, this application enables dynamic ontology objects to not only retain knowledge content but also the organizational relationships between knowledge content based on tasks, thereby improving the adjustability and applicability of dynamic ontology objects.

[0170] Fourth, by establishing the association between knowledge fields and target object units, this application enables dynamic ontology objects to support adjustments in the event of subsequent task changes and knowledge content changes, thus avoiding the need to re-traverse all document content for knowledge construction.

[0171] like Figure 2As shown, this application also provides a dynamic ontology object construction apparatus based on capability problem inversion, comprising: The knowledge parsing module is used to acquire the document to be processed, parse the content expression relationship, structural organization relationship and knowledge association relationship in the document to be processed, and determine multiple knowledge expression objects; The object construction module is used to obtain task requirement information, determine the content supplementation relationship and task support relationship between multiple knowledge expression objects based on the task processing goal and knowledge usage requirements corresponding to the task requirement information, determine the combination of knowledge expression objects participating in the construction of dynamic ontology objects based on the content supplementation relationship and task support relationship, determine the combination range and composition boundary corresponding to the combination of knowledge expression objects, and organize the combination of knowledge expression objects according to the combination range and composition boundary to form a target object unit; The structure generation module is used to perform structured processing on the target object unit according to the knowledge content scope and field organization requirements corresponding to the target object unit, form knowledge fields, and establish the association between the knowledge fields and the target object unit to generate a dynamic ontology object.

[0172] like Figure 3 As shown, an electronic device according to an embodiment of this application includes: a memory, a processor, and a computer program stored in the memory and executable on the processor. The computer program is configured to implement the steps of the dynamic ontology object construction method based on capability problem back-reasoning as described above in this application.

[0173] like Figure 4 As shown, this is a computer-readable storage medium according to an embodiment of the present application. The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the dynamic ontology object construction method based on capability problem back-reasoning as described above in this application.

[0174] The above Figures 2-4 For an exemplary description, please refer to the above. Figure 1 This will not be elaborated upon here.

Claims

1. A method for constructing dynamic ontology objects based on capability problem inversion, characterized in that, include: Obtain the document to be processed, parse the content expression relationship, structural organization relationship and knowledge association relationship in the document to be processed, and determine multiple knowledge expression objects; Obtain task requirement information, determine the content supplementation relationship and task support relationship between multiple knowledge expression objects based on the task processing objectives and knowledge usage requirements corresponding to the task requirement information, determine the combination of knowledge expression objects participating in the construction of dynamic ontology objects based on the content supplementation relationship and task support relationship, determine the combination range and composition boundary corresponding to the combination of knowledge expression objects, and organize the combination of knowledge expression objects according to the combination range and composition boundary to form target object units; Based on the knowledge content scope and field organization requirements corresponding to the target object unit, the target object unit is structured to form knowledge fields, and the association between the knowledge fields and the target object unit is established to generate a dynamic ontology object.

2. The method according to claim 1, characterized in that, The process involves parsing the content expression relationships, structural organization relationships, and knowledge association relationships in the document to be processed to determine multiple knowledge expression objects, including: The scope of knowledge description in the document to be processed is determined based on the content expression relationship; The hierarchical relationship between different knowledge description scopes is determined based on the structural organization relationship, and the association relationship between different knowledge description scopes is determined based on the knowledge association relationship; The boundaries of knowledge content are adjusted according to the scope, hierarchical relationship, and association relationship of the knowledge description to form multiple knowledge expression objects.

3. The method according to claim 2, characterized in that, The adjustment of knowledge content boundaries based on the knowledge description scope, hierarchical relationship, and association relationship includes: The scope of associated content corresponding to the scope of knowledge description is determined based on the hierarchical relationship; Determine the associated knowledge content corresponding to the scope of associated content based on the knowledge association relationship; The knowledge content division boundaries are adjusted according to the scope of the associated content and the associated knowledge content to form the knowledge expression object.

4. The method according to claim 1, characterized in that, Determining the content supplementation relationship and task support relationship among multiple knowledge representation objects includes: Determine the knowledge content required to complete the corresponding task based on the task processing objectives; Based on the knowledge content, determine the corresponding knowledge representation object from among the multiple knowledge representation objects; Based on the knowledge content deficiencies corresponding to the knowledge representation object and its supporting role in the task processing objective, the content supplementation relationship and task support relationship are determined.

5. The method according to claim 4, characterized in that, The step of determining the content supplementation relationship and task support relationship based on the knowledge content deficiencies corresponding to the knowledge representation object and its supporting role in the task processing objective includes: Based on the knowledge integrity requirements corresponding to the task processing objectives, determine the scope of knowledge content that needs to be covered to complete the task processing objectives; Based on the scope of the knowledge content, analyze the missing content areas and corresponding supplementary content among multiple knowledge expression objects; Based on the missing content areas and the corresponding supplementary content, a combination of knowledge expression objects that jointly cover the scope of the knowledge content is determined.

6. The method according to claim 5, characterized in that, The process of combining the knowledge representation objects to form the target object unit includes: The combination boundary is determined based on the knowledge content coverage of the combination of knowledge representation objects. The composition range corresponding to the combination of knowledge representation objects is determined based on the combination boundary. The knowledge representation objects are associated and organized according to the defined composition range to form the target object unit.

7. The method according to claim 1, characterized in that, After forming the target object unit, the process further includes: Based on the content association relationships between the knowledge representation objects that form the target object unit, determine the content connection relationships between different knowledge representation objects; Adjust the organization of knowledge content corresponding to the target object unit according to the content connection relationship.

8. A dynamic ontology object construction device based on capability problem inverse reasoning, characterized in that, include: The knowledge parsing module is used to acquire the document to be processed, parse the content expression relationship, structural organization relationship and knowledge association relationship in the document to be processed, and determine multiple knowledge expression objects; The object construction module is used to obtain task requirement information, determine the content supplementation relationship and task support relationship between multiple knowledge expression objects based on the task processing goal and knowledge usage requirements corresponding to the task requirement information, determine the combination of knowledge expression objects participating in the construction of dynamic ontology objects based on the content supplementation relationship and task support relationship, determine the combination range and composition boundary corresponding to the combination of knowledge expression objects, and organize the combination of knowledge expression objects according to the combination range and composition boundary to form a target object unit; The structure generation module is used to perform structured processing on the target object unit according to the knowledge content scope and field organization requirements corresponding to the target object unit, form knowledge fields, and establish the association between the knowledge fields and the target object unit to generate a dynamic ontology object.

9. An electronic device, characterized in that, The electronic device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the dynamic ontology object construction method based on capability problem backpropagation as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the dynamic ontology object construction method based on capability problem backpropagation as described in any one of claims 1 to 7.

Images