Symbol cell library construction and update method and related devices

Abstract

The present disclosure relates to a symbol unit library construction and updating method and related equipment, comprising: obtaining image data of a target entity; positioning at least one target region of the target entity based on the image data to obtain a region identifier corresponding to the target entity; extracting a basic attribute feature of the target region based on the image data; classifying the attribute of the target region according to the basic attribute feature to obtain a basic attribute symbol corresponding to the target region; establishing an association relationship between the region identifier, the basic attribute symbol and the basic attribute feature to obtain a symbol record of the target entity and store the symbol record in a symbol unit library; in response to a query request for the symbol record, outputting information of the symbol record corresponding to the query request; and in response to detecting a symbol record updating trigger condition, updating the symbol record.

Description

Technical Field

[0001] The embodiments disclosed herein relate to the fields of computer technology and artificial intelligence technology, and in particular to a method and related equipment for constructing and updating a symbol unit library. Background Technology

[0002] Currently, data files such as project archives, image annotations, industrial prototype records, or assembly records are typically tagged directly based on images, videos, or text annotations. There is a lack of intermediate data objects that can be reused across multiple projects, versions, and application scenarios. This results in a lack of stable and traceable sources of tag data when processing these data files subsequently. Summary of the Invention

[0003] This disclosure aims to at least partially address one of the technical problems in the aforementioned background or related technologies.

[0004] To address this, this disclosure provides a method and related equipment for constructing and updating a symbol unit library, which can solve the problem of lacking traceable data when managing the recorded data of entity objects in related technologies.

[0005] In a first aspect, this disclosure provides a method for constructing and updating a symbol unit library, comprising: acquiring image data of a target entity; locating at least one target region of the target entity based on the image data to obtain a region identifier corresponding to the target entity; extracting basic attribute features of the target region based on the image data; classifying the attributes of the target region according to the basic attribute features to obtain basic attribute symbols corresponding to the target region; establishing an association relationship between the region identifier, the basic attribute symbols, and the basic attribute features to obtain a symbol record of the target entity and storing the symbol record in a symbol unit library; outputting information of the symbol record corresponding to the query request in response to a query request for the symbol record; and updating the symbol record in response to detecting a symbol record update trigger condition.

[0006] In a second aspect, this disclosure provides a computer device including one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and executed by the one or more processors, and the one or more programs include instructions for performing the method described in the first aspect.

[0007] Thirdly, this disclosure provides a non-volatile computer-readable storage medium containing a computer program that, when executed by one or more processors, causes the one or more processors to perform the method described in the first aspect.

[0008] Fourthly, this disclosure provides a computer program product comprising one or more computer programs that, when executed by one or more processors, implement the method described in the first aspect.

[0009] The symbol unit library construction and updating method of this disclosure involves locating the target region of the target entity based on the target entity's image data, extracting features from the target region based on the target entity's image data, classifying the attributes of the target region based on the extracted features, identifying the basic attributes of the target region, and establishing a correlation between region identifiers, basic attribute identifiers, and basic attribute features to form symbol records for the target entity. Subsequently, when an external query request for a symbol record is received, the corresponding symbol record information can be output. When a symbol record update trigger condition is detected, the symbol record can be updated. This ensures that the labels of the target entity are no longer isolated results but form structured record data with correlations, improving the traceability, verifiability, and reusability of the target entity label data. Attached Figure Description

[0010] The following description and accompanying drawings will better aid in understanding these and other features and advantages of the various embodiments disclosed herein, wherein the same reference numerals in the drawings consistently denote the same parts, wherein: Figure 1 This is a flowchart of an exemplary symbol unit library construction and update method according to an embodiment of this disclosure.

[0011] Figure 2 This is a schematic diagram of the structure of a system for implementing a symbol unit library construction and update method, which is an exemplary embodiment of this disclosure.

[0012] Figure 3 This is a flowchart of an exemplary symbol unit library construction and update method according to an embodiment of this disclosure.

[0013] Figure 4 This is a schematic diagram illustrating an exemplary symbol unit library update process according to an embodiment of this disclosure.

[0014] Figure 5 A schematic diagram of the hardware structure of an exemplary computer device provided in an embodiment of this disclosure is shown. Detailed Implementation

[0015] The technical solutions of this disclosure will be further described in detail below through embodiments and in conjunction with the accompanying drawings. In this specification, the same or similar reference numerals indicate the same or similar components. The following description of the embodiments of this disclosure with reference to the accompanying drawings is intended to explain the overall inventive concept of this disclosure and should not be construed as a limitation thereof.

[0016] Figure 1 This is a flowchart illustrating an exemplary method for constructing and updating a symbol unit library according to an embodiment of this disclosure. This method can be executed by a computer, server, or other general-purpose electronic device with a processor and memory, or it can be implemented through a dedicated hardware module or a combination of hardware and software. Figure 1 As shown, the method includes the following steps: In step S102, image data of the target entity is acquired.

[0017] In one or more embodiments of this disclosure, the target entity can be any entity object. For example, a target entity can refer to a visual work, material work, construction work, industrial prototype, assembly, or versioned object that can be acquired through images, videos, or continuous images, and from which visual, material, spatial, structural, or process variation features can be extracted. Image data of the target entity can include, for example, images, continuous images, or videos.

[0018] In step S104, at least one target region of the target entity is located based on the image data to obtain the region identifier corresponding to the target entity.

[0019] Optionally, at least one target region of the target entity can be identified in the image of the target entity based on image recognition technology to obtain a set of coordinates of the target region, thereby realizing the localization of the target region in the image of the target entity. After the target region is located, a region identifier can be assigned to each target region to uniquely identify each region of the target entity.

[0020] In step S106, basic attribute features of the target region are extracted based on the image data.

[0021] Optionally, based on preset coordinate parameters, contour boundaries, or mask information, a local image matrix containing only the target region can be segmented and extracted from the complete image data, removing irrelevant pixel data outside the region. Basic parameters of all pixels within the target region are read, including pixel grayscale values, RGB (Red, Green, Blue) color components, transparency, pixel position coordinates, and other raw data, serving as the underlying data source for feature extraction. Different types of features are calculated and generated according to preset extraction rules. In one example, these different types of features include, but are not limited to, image features, spatial features, structural features, material appearance features, and process change features.

[0022] In step S108, the attributes of the target region are classified according to the features to obtain the basic attribute symbols corresponding to the target region.

[0023] Optionally, the basic attribute symbols can be basic multimodal symbol units generated from the regional features of the target entity, serving as intermediate data objects between image features and label results.

[0024] Optionally, the basic attributes of the target area can be, for example, its physical attributes, including color, material, space, structure, or process variation attributes. Basic attribute symbols are used to characterize these attributes and can be identifiers or numbers used to uniquely identify them. Taking a visual / material artwork as an example, the physical attributes of the target entity could be overlapping color patches, color layering, or color variations, etc.

[0025] In step S110, the association between the region identifier, the basic attribute symbol, and the basic attribute feature is established to obtain the symbol record of the target entity.

[0026] Optionally, the symbol record of the target entity can be a structured record bound to Symbol_ID (symbol identifier), which may include fields such as Region_ID (region identifier), Symbol_Type (symbol type), Feature_Value (feature value), Modality_Source (modality source), Extraction_Rule (extraction rule), Candidate_Tag_Weight (candidate tag weight), Symbol_Version_No (symbol version number), and Update_Log (update log).

[0027] In step S112, in response to the query request, information of the symbolic record corresponding to the query request is output; Optionally, in response to different query requests, information from different symbolic records can be output. The query request may include key fields indicating the object being invoked, and may further include the purpose of the invocation.

[0028] In step S114, the symbol record is updated in response to the detection of a symbol record update trigger condition.

[0029] The update trigger conditions may include updates to the image data of the target entity or updates to the text data corresponding to the target entity.

[0030] Specifically, updating a symbol record may include updating at least one key field in the symbol record, and may also include merging, splitting, or changing the attributes of key fields in the symbol record.

[0031] The method for constructing and updating a symbol unit library for a target entity according to embodiments of this disclosure involves locating the target region of the target entity based on its image data, extracting features from the target region based on the image data, classifying the attributes of the target region based on the extracted features, identifying the basic attributes of the target region, establishing a relationship between region identifiers, basic attribute identifiers, and basic attribute features to form symbol records for the target entity, and subsequently outputting the corresponding symbol record information when an external query request for the symbol record is received, and updating the symbol record when a symbol record update trigger condition is detected. This ensures that the labels of the target entity are no longer isolated results, but form structured record data with relationships, improving the traceability, verifiability, and reusability of the target entity label data.

[0032] The symbol unit library construction and updating method of one or more embodiments of this disclosure may further include: After classifying the attributes of the target region according to the basic attribute features to obtain the basic attribute symbols corresponding to the target region, at least two different dimensions of first label and second label are generated according to the feature parameters corresponding to the basic attribute symbols. Optionally, the first label and the second label represent label data from two different dimensions. In different embodiments, these could correspond to the dimensions of visual / material expression, exploration of concepts, structural layout, or functional verification.

[0033] Optionally, the feature parameters corresponding to the basic attribute symbols may include at least one of the following: basic attribute parameters, identifier extraction rule parameters, and identifier confidence weight parameters. The basic attribute parameters may include: the symbol type of the basic attribute symbol and the feature value of the feature corresponding to the basic attribute symbol. The identifier extraction rule parameters may include: modality source (modality source refers to the acquisition channel, perception form, input medium, and information dimension type of the original information data in intelligent sensing, data recognition, and feature extraction tasks; it is the different sensing / data modal sources from which the machine acquires external information) and extraction rules. The identifier confidence weight parameters may include: candidate label weights and identifier confidence levels.

[0034] The first and second labels can be used to represent different effects and functions of the same basic attribute, or, for works of art, they can correspond to different creative elements of a basic attribute.

[0035] Based on the region identifier, the basic attribute symbol, the first tag, the second tag, and at least one type of preset structured judgment information that matches the first tag and the second tag, a relationship tag is generated to represent the association between the first tag and the second tag.

[0036] The relationship between the first and second tags can include, for example, data source relationships, spatial relationships, version relationships, or conflict relationships. Examples of data source relationships include `shared_basis`. Examples of spatial relationships include `adjacency_support`. Examples of version relationships include `version_shift`. Examples of conflict relationships include `conflict`.

[0037] Optionally, the preset structured judgment information may include pre-set tag association information, such as region sharing information, symbol sharing information, spatial relationship information, version difference information, and field consistency detection information. Specifically, region sharing information characterizes the spatial overlap, range inclusion, and region coincidence relationships between the regions corresponding to the first tag and the regions corresponding to the second tag. Symbol sharing information characterizes the homology, consistency, and association matching relationships between the basic attribute symbols corresponding to the first tag and the basic attribute symbols corresponding to the second tag. Spatial relationship information quantifies and defines the global spatial location association and topological arrangement relationships between the target entity regions corresponding to the first tag and the target entity regions corresponding to the second tag. Version difference information characterizes the dynamic differences between the first and second tags due to differences in acquisition time sequence, imaging modality, iteration batch, and update version. Field consistency detection information verifies the standardization consistency and content matching of all structured parameter fields corresponding to the first and second tags.

[0038] The relationship label between the first and second labels is used to characterize the aforementioned association between them.

[0039] The region identifier, the basic attribute symbol, the first label, the second label, and the relationship label are associated to generate and output the record data of the target entity.

[0040] Optionally, the region identifier, basic attribute symbols, first label, second label, and the relationship label between the first label and the second label of the target entity can be bound and stored to generate structured record data of the target entity. Optionally, the region identifier, region features, first label, and second label can be stored in a symbol unit library.

[0041] Optionally, the method for constructing and updating the symbol unit library in this embodiment of the present disclosure may further include: A relationship verification identifier is generated based on at least one of the following: the confidence level of the target region identifier, the confidence level of the basic attribute symbol identifier, the confidence level of the first label, the confidence level of the second label, the confidence level of the relationship label, field integrity, and text consistency indicators. The relationship verification identifier is used to indicate whether the relationship label needs to be verified, whether the data is incomplete, or whether there is a rule conflict.

[0042] The relation labels, relation confidence levels, and relation verification identifiers are stored in the symbolic unit library.

[0043] The symbol unit library construction and updating method of one or more embodiments of this disclosure may further include: Detect whether text data exists in the target area. If it exists, acquire the text data, wherein the text data includes at least one of the following: bill of materials, assembly list, associated text, process notes, or version difference field. Enhanced attribute features of the target region are extracted based on the text data; Based on the enhanced attribute features, generate enhanced attribute symbols corresponding to the target region; Establish the association between the region identifier, the basic attribute symbol, and the basic attribute feature to obtain the symbol record of the target entity, including: Establish the association relationship between the region identifier, the basic attribute symbol, the basic attribute feature, the enhanced attribute symbol, and the enhanced attribute feature to obtain the symbol record of the target entity.

[0044] For example, the basic Symbol_ID (an example of the basic attribute identifier mentioned above) and the enhanced Symbol_ID (an example of the enhanced attribute identifier mentioned above) are written into the multimodal symbol cell library (hereinafter also referred to as the symbol cell library) to generate a Symbol_Record.

[0045] Specifically, the system generates a corresponding Symbol_Record for each basic Symbol_ID, enhanced Symbol_ID, or Symbol_ID_Set (symbol identifier set). Symbol_Record may include at least one of the following fields: Symbol_Record_ID (symbol record identifier), Work_ID (object identifier), Region_ID (region identifier), Basic_Symbol_ID (basic attribute identifier), Enhanced_Symbol_ID (enhanced attribute identifier), Symbol_ID_Set, Symbol_Type, Feature_Value, Enh_Feature_Set (enhanced attribute feature set), Modality_Source, Extraction_Rule, Candidate_Tag_Weight, Applicable_Scene (applicable scenario), Symbol_Confidence (symbol confidence), Conflict_Frequency (conflict frequency), Symbol_Version_No, Status (status), Previous_Symbol_ID (historical symbol identifier), and Update_Log_ID (update log identifier).

[0046] After writing the Symbol_Record into the multimodal symbol unit library, the system creates at least one index. The created indexes may include: a primary index with Symbol_ID or Symbol_Record_ID as the key; secondary indexes with Work_ID, Region_ID, Symbol_Type, Modality_Source, Feature_Value, Applicable_Scene, or Symbol_Version_No as the key; and a graph index used to represent the relationships between Region_ID, Symbol_ID, First_Tag, Second_Tag, or Relation_Tag.

[0047] The symbol unit library construction and updating method of this disclosure stores the image features, bill of materials, assembly list, text notes and version differences of the target entity in a unified symbol unit library, so that subsequent label mapping, label relationship determination and version hash verification processing can have a stable and traceable field source.

[0048] In one or more embodiments of this disclosure, in response to a query request for the symbol record, information of the symbol record corresponding to the query request is output, which may specifically include: Obtain a query request, wherein the query request includes a call object indicated by a key field and a call purpose, wherein the call object includes at least one of the symbolic records, and the call purpose includes at least one of generating record data of the target entity, verifying the tag relationship of the target entity, and verifying the hash chain of the record data of the target entity.

[0049] If the purpose of the call is to generate record data of the target entity, then a list of symbol identifiers corresponding to the object identifier or the region identifier and a symbol record summary are output, wherein the object identifier is used to identify the target entity; If the purpose of the call is to determine the label relationship of the target entity, then at least one of the following will be output: symbol identifier list, symbol record, symbol type, feature value, modality source, candidate label weight, and symbol confidence. If the purpose of the call is to verify the hash chain of the record data of the target entity, then at least one of the following will be output: symbol update record, symbol version number, update log identifier, or symbol record digest.

[0050] Specifically, the system outputs different forms of data depending on the calling object and the purpose of the call. When the purpose of the call is to generate record data for the target entity, the system can output a Symbol_ID_List (symbol identifier list) and its Symbol_Record summary corresponding to the Work_ID (object identifier) ​​or Region_ID, which are used to generate the First_Tag, Second_Tag, and Relation_Tag (relationship tag). When the purpose of the call is to determine the tag relationship of the target entity, the system can output Symbol_ID_List, Symbol_Record, Symbol_Type, Feature_Value, Modality_Source, Candidate_Tag_Weight, or Symbol_Confidence, which are used to determine whether the tags share Symbol_ID, whether there is a basis for relationship determination, and whether it is necessary to generate Relation_Confidence (relationship confidence) or Relation_Check_Flag. When the purpose of the call is to verify the hash chain of the record data of the target entity, the system can output a symbol update record (Symbol_Update_Record), Symbol_Version_No (symbol version number), Update_Log_ID (update log identifier), or Symbol_Record digest, which can be included in Hash_Key_Field_Set (hash key field set).

[0051] The Symbol_Update_Record is a structured record used to record symbol unit library update events. It can take the form of a table, key-value pair, JSON (JavaScript Object Notation), graph edge record, or a combination thereof. Symbol_Update_Record can include: Symbol_Update_Record_ID (symbol unit update record number), Work_ID, Region_ID, Symbol_ID, Previous_Symbol_ID (the original symbol unit number before the update or after merging, splitting, or deactivation), and Update_Action (update action), which can specifically include add, merge, split, deactivate, weight_adjust, scene_adjust, version_update, or log_write. The `Trigger_Type` can specifically include: `feature_change`, `source_added`, `relation_check_flag`, `relation_confidence_change`, `field_difference`, `verification_result`, or `structured_confirmation`. The `Trigger_Source` includes the invoked source, bill of materials, assembly list, associated text, process notes, or version difference fields.

[0052] Before_Value (field value or summary before update), After_Value (field value or summary after update), Candidate_Tag_Weight_Change (candidate tag weight change), Applicable_Scene_Change (applicable scenario change), Conflict_Frequency_Change (conflict frequency change), Old_Symbol_Version_No (symbol version number before update), New_Symbol_Version_No (symbol version number after update), Related_Tag_Pair (related tag pair), Relation_Tag (related relationship tag), Relation_Check_Flag (relationship verification hint field), Relation_Confidence, Verification_Result (version verification result), Field_Difference_Record_ID (field difference record number), Update_Time (update time), Update_Result (update result status). The system outputs Symbol_Update_Record, which can be used to supplement structured record data, as input for tag relationship determination, or as a candidate field for Hash_Key_Field_Set to support version hash chain verification.

[0053] In one or more embodiments of this disclosure, extracting enhanced attribute features based on the text data and generating enhanced attribute symbols corresponding to the target region based on the enhanced attribute features may specifically include: If a material list corresponding to the target area is detected, a material attribute field is generated based on the material list; If an assembly list corresponding to the target area is detected, an assembly status field is generated based on the assembly list; If associated text corresponding to the target region is detected, extract the text event field based on the associated text; If a process note corresponding to the target area is detected, the process field is extracted based on the process note. All fields generated based on the text data will form an enhanced attribute feature set; The enhanced attribute feature set is matched with preset enhanced symbol rules, candidate enhanced symbol identifier templates, or symbol records in the symbol unit library to obtain matching results, and enhanced attribute symbols are generated based on the matching results.

[0054] Optionally, when there is a bill of materials, assembly list, associated text, or process notes corresponding to the target area of ​​the target entity, these text data are parsed in a structured manner and an enhanced Symbol_ID or enhanced field is generated.

[0055] Specifically, the system retrieves the bill of materials, assembly list, associated text, or process notes associated with Work_ID or Region_ID, and converts them into Enh_Input_Record. Enh_Input_Record may include at least one of the following fields: Source_ID, Work_ID, Region_ID, Time_Point, Input_Type, Input_Content, and Source_Reliability.

[0056] For the bill of materials, the system can parse at least one of the following: material name, material category, color, transparency, stiffness, flexibility, thickness, surface texture, connecting materials, or material replacement information, to generate a Material_Property_Field.

[0057] For the assembly list, the system can parse at least one of the following fields: component name, component number, installation status, connection method, fastener status, assembly location, support relationship, or assembly sequence, to generate the Assembly_Status_Field (assembly status field).

[0058] For associated text or process notes, the system can extract Text_Event_Field or Process_Field using at least one of the following methods: keyword matching, field extraction, preset thesaurus matching, rule template matching, or text event recognition. The text event field or process field can include event or attribute information such as "add support," "base widening," "material replacement," "transparent color sheet," "tape connection," "further modification," and "version revision."

[0059] The system associates the aforementioned Material_Property_Field, Assembly_Status_Field, Text_Event_Field, Process_Field, or version change fields with the corresponding Region_ID, base Symbol_ID, or Time_Point to form an Enh_Feature_Set. The system then matches the Enh_Feature_Set with the preset enhanced symbol rule Enh_Symbol_Rule, candidate enhanced Symbol_ID templates, or existing Symbol_Records in the symbol unit library. Based on at least one of the following: field hit count, source reliability, field completeness, matching degree with the base Symbol_ID, or historical review pass rate, it generates an enhanced Symbol_ID or enhanced field.

[0060] For example, when a "transparent color" field exists in the bill of materials or text notes, and this field is associated with the overlapping area of ​​Region_ID=R03, the system can generate an enhanced Symbol_ID=material_transparency (material transparency); when a "bolts installed" field exists in the assembly list and is associated with the bolt connection area R04, the system can generate an enhanced Symbol_ID=fastener_installed (fasteners installed) or fastener_alignment (fastener assembly alignment). When the process notes contain the fields "Added support" or "Base widening" and are associated with the corresponding support or base area, the system can generate an enhanced Symbol_ID=support_added (Added support) or base_widening (Base widening).

[0061] In one or more embodiments of this disclosure, the difference between the basic Symbol_ID and the enhanced Symbol_ID is as follows: the basic Symbol_ID is mainly generated from regional features in an image, video, or continuous image, reflecting the identifiable color, shape, space, structure, material appearance, or process change features of the region corresponding to the Region_ID; the enhanced Symbol_ID is mainly generated by supplementing the basic Symbol_ID with material lists, assembly lists, associated text, process notes, or version difference fields, and is used to corroborate, supplement, correct, or enhance the basic Symbol_ID. The two can together form a Symbol_ID_Set for subsequent tag mapping, relationship determination, and version hash chain verification.

[0062] In one or more embodiments of this disclosure, the symbol record is stored in a symbol unit library, which stores symbol records of multiple entity objects. In response to detecting a symbol record update trigger condition, the symbol record is updated, which may specifically include: If the matching degree between the basic attribute feature set or enhanced attribute feature set corresponding to the region identifier and the stored symbol record does not meet the preset matching degree, or if the text data corresponding to the region identifier is updated, then the symbol record corresponding to the region identifier is regenerated. For example, if the matching degree between the Feature_Set or Enh_Feature_Set corresponding to the current Region_ID and the existing Symbol_Record in the symbol unit library does not meet the preset matching degree, or if a new material source, assembly source, text event, or version difference field is detected, the system generates a new Symbol_ID, establishes a new Symbol_Record, marks its Status as active, and writes it to the Symbol_Library_Index (symbol unit library index).

[0063] If at least one of the feature value, symbol type, modality source, or candidate label mapping result corresponding to at least two symbol identifiers is detected to satisfy a preset similarity, one of the at least two symbol identifiers is determined to be the primary symbol identifier, and the symbol record information corresponding to the at least two symbol identifiers is merged. For example, when the Feature_Value, Symbol_Type, Modality_Source, or candidate label mapping result corresponding to two or more Symbol_IDs satisfies the preset similarity, or when multiple structured confirmations indicate that they represent the same symbol unit, the system selects one of the Symbol_IDs as the primary Symbol_ID, writes the other Symbol_IDs into the Previous_Symbol_ID (historical symbol identifier) ​​or Alias_Symbol_ID (alternate symbol identifier) ​​fields, and merges their Feature_Value, Candidate_Tag_Weight, Applicable_Scene, and Update_Log.

[0064] If at least one of the following is detected as a symbol identifier that satisfies a preset differentiation index: feature value, modality source, candidate label mapping result, symbol type, applicable scenario, or relationship determination result, the detected symbol identifier will be split into at least two different sub-symbol identifiers. For example, when the Feature_Value, Modality_Source, or candidate label mapping result under the same Symbol_ID shows obvious differentiation, or when the same Symbol_ID corresponds to multiple different Symbol_Types, applicable scenarios, or relationship determination results, the system can split the Symbol_ID into two or more sub-Symbol_IDs and record the Parent_Symbol_ID (parent symbol identifier) ​​or Previous_Symbol_ID (historical symbol identifier) ​​in each sub-Symbol_Record to preserve the evolution source.

[0065] If a symbol is detected to have a negative feedback result that meets a first preset metric when it is called or queried, the status of the symbol that meets the first preset metric is set to disabled. For example, when a Symbol_ID triggers conflict, rule_conflict multiple times, has incomplete fields, or is no longer applicable to the current scenario after structured confirmation, the system updates the Status of the Symbol_ID to inactive, deprecated, or suspended. The disabled action does not delete the historical Symbol_Record, but retains its historical version and Update_Log for subsequent traceability.

[0066] The system adjusts the candidate tag weights and corresponding applicable scenarios of symbols based on the positive or negative feedback results obtained when the symbols are invoked or queried. For example, when a Symbol_ID forms a stable mapping with the same candidate tag multiple times, and Relation_Confidence meets the preset requirements, the system can increase its Candidate_Tag_Weight. When Relation_Check_Flag is pending_check, data_incomplete, or rule_conflict, or when Relation_Confidence does not meet the preset requirements, the system can decrease the corresponding Candidate_Tag_Weight or mark it as pending verification. The weight adjustment results are written to Symbol_Record and Update_Log. In one example, after locating the overlapping area R03 of transparent color patches, the system generates a basic attribute identifier of S01_color_overlap. If the bill of materials contains transparent color patches, the system parses the material attribute fields and generates an enhanced Symbol_ID: S02_material_transparency. When subsequent determinations show that S01 and color_composition and light_color_change have formed a shared_basis relationship multiple times, and Relation_Confidence meets the preset requirements, the system increases the Candidate_Tag_Weight of S01 for the above candidate tag combination and writes the adjusted Candidate_Tag_Weight to Update_Log.

[0067] When the same Symbol_ID is repeatedly called across different work types, material types, assembly types, or project cycles, the system adjusts Applicable_Scene or Scene_Weight based on the call frequency, relationship determination results, or version verification results. Scene adjustment is used to limit or expand the applicability of the Symbol_ID in visual works, material works, industrial prototypes, assemblies, or children's works.

[0068] If a change is detected in a key field corresponding to a symbolic record, a new symbolic record version identifier is generated. For example, when the Feature_Value, Enh_Feature_Set, Candidate_Tag_Weight, Applicable_Scene, Status, or other key fields of a Symbol_Record change, the system generates a new Symbol_Version_No and records the previous version number as Previous_Symbol_Version_No. The new Symbol_Version_No is bound to the updated Symbol_Record for subsequent version tracking and hash chain verification. In one example, version V1 of the target entity has no slant, while version V2 of the target entity adds a slant. The system generates Symbol_ID_Lists for both versions V1 and V2, and generates S20_version_difference through version differentiation. If Relation_Tag=version_shift (version change) is further generated based on version difference, S20 can be written into the symbol unit library, and its applicable scenarios and version sources as a process_change (process change) type symbol unit can be recorded.

[0069] The following example illustrates the process of updating the symbol unit library based on update trigger conditions.

[0070] Taking the industrial bracket prototype as an example, the diagonal bracing area R03 generates S10_diagonal_reinforcement (diagonal reinforcement structure), and the bolt connection area R04 generates S11_fastener_alignment (fastener alignment) in conjunction with the assembly list.

[0071] If Relation_Tag=conflict, or Relation_Check_Flag is pending_check, data_incomplete, or rule_conflict when generating relation tags, after calling the structured relation result when generating the record data of the target entity, instead of directly making a pass / fail judgment, it updates at least one of Conflict_Frequency, Candidate_Tag_Weight, applicable scenario weight, or Symbol_Version_No in S11 according to the verification prompt source, and records the value of Relation_Check_Flag, trigger source, and associated Tag_Pair in Update_Log.

[0072] Among them, `Relation_Tag=conflict` can indicate inconsistencies between image recognition results, assembly lists, material lists, associated text, or historical version records; when `Relation_Check_Flag` is `pending_check`, `data_incomplete`, or `rule_conflict`, it can indicate that the relationship result needs further verification, key fields are missing leading to insufficient basis for relationship determination, or inconsistencies exist between multiple relationship rules, respectively. All of the above fields can serve as structured triggers for dynamic updates of the symbol unit library.

[0073] The symbol unit library construction and updating method of one or more embodiments of this disclosure may further include: After updating the symbol record, a symbol update log is generated. The symbol update log includes at least one of the following: update time, update action, trigger type, trigger source, original value, new value, associated region identifier, associated symbol identifier, associated tag pair, association relationship verification flag, relationship confidence, or verification result. For example, the system generates an Update_Log each time it performs an update, merge, split, deactivate, weight adjustment, scene adjustment, or Symbol_Version_No update. The Update_Log records at least one or more fields from Update_Time, Update_Action, Trigger_Type, Trigger_Source, Before_Value, After_Value, Related_Region_ID, Related_Symbol_ID, Related_Tag_Pair, Relation_Check_Flag, Relation_Confidence, or Verification_Result.

[0074] In one or more embodiments of this disclosure, the image data of the target entity may include at least one of images, continuous images, or videos, and the basic attribute features may include at least one of image features, spatial features, structural features, material appearance features, and process change features. The basic attributes of the target region are classified according to the basic attribute features to obtain the basic attribute symbols corresponding to the target region, which may specifically include: The features are preprocessed to obtain feature vectors.

[0075] Optionally, the preprocessing of features may include at least one of the following: feature cleaning, feature denoising, feature normalization, feature dimension unification, feature invalid value removal, feature registration and alignment, feature interpolation completion, feature weight initialization, multimodal feature fusion preprocessing, and feature redundancy removal.

[0076] The feature vector is matched with a preset symbol matching rule, a preset candidate symbol identifier, or a symbol record in a symbol unit library to obtain a matching result.

[0077] The basic attribute symbols corresponding to the target region are determined based on the matching results.

[0078] The preset symbol matching rules can be a set of standardized calculation and judgment rules that are pre-configured and stored, and may include pre-established mapping relationships between pre-processed quantized features, target region attributes, and standardized attribute symbols. During matching, one or more basic attribute symbols that match the feature vector can be obtained based on these mapping relationships.

[0079] The preset candidate symbol identifiers can be a set of all possible standardized symbol identifiers pre-enumerated based on the inherent attributes and dynamically changing attributes of the target entity. During matching, one or more basic attribute symbols can be matched from the symbol identifier set based on the feature vector of the target region.

[0080] The symbol unit library is a pre-built standardized symbol resource database containing all compliant symbol records. Based on the characteristics of the target region, one or more basic attribute symbols can be matched within the symbol unit library according to the symbol records.

[0081] Optionally, the matching result may include at least one of the following: rule hit count, feature weight, number of modality sources, similarity score, or historical review pass rate. The rule hit count refers to the total number of valid rules in the preset symbol matching rule set that can successfully match or meet the adaptation conditions of the preprocessed features of the current target region. Feature weight is a pre-configured or adaptively updated quantitative priority and contribution parameter for different dimensions of features such as image features, spatial features, structural features, material appearance features, and process change features. The number of modality sources refers to the total number of categories of valid data modalities currently participating in feature extraction, symbol matching, and attribute determination of the target region. The similarity score is a quantitative score that represents the degree of matching between the real-time feature set of the current target region and the standard feature samples and standard attribute templates corresponding to each preset candidate symbol identifier in the symbol unit library. The historical review pass rate refers to the compliance pass rate of the basic attribute symbols and label determination results generated by historical matching under the same type of target entity, the same type of scene, and the same type of feature matching scenario, after review, secondary verification, and system iterative verification.

[0082] In one or more embodiments of this disclosure, at least two different dimensions of a first label and a second label are generated based on the feature parameters corresponding to the basic attribute symbols, which may specifically include: The first label mapping input dataset is constructed based on at least one of the following: the symbol type, feature value, modality source, extraction rule, candidate label weight, and label confidence level corresponding to the basic attribute symbol.

[0083] Based on the first label mapping input dataset, feature matching is performed through prior constraints, or inference is performed through a pre-trained model to determine the first label and the second label. The first label and the second label are used to represent creative elements of different dimensions corresponding to the basic attribute labels.

[0084] Optionally, the prior constraints can be any of the following: label mapping rules, candidate label mapping tables, and candidate label weights. Label mapping rules can be used to establish a one-to-one or many-to-one matching standardization constraint and transformation logic between quantified feature data, attribute determination results, and standardized candidate labels. The candidate label mapping table can be a pre-built, structured, indexed data table that can solidify the correspondence between legal feature combinations, standard attribute types, and candidate labels. Candidate label weights are independent quantized weight parameters configured for each of the retrieved candidate labels.

[0085] The first label mapping can be input into the pre-trained model, so that the pre-trained model can perform feature inference and attribute classification on the input dataset based on the label mapping rules and feature association logic, and output the prediction results of the first label and the second label.

[0086] In one or more embodiments of this disclosure, the first label and the second label are determined based on the first label mapping input dataset, through feature matching by prior constraints, or through inference by a pre-trained model. Specifically, this may include: Based on the first label mapping input dataset, feature matching is performed using prior constraints, or inference is performed using a pre-trained model to determine a first candidate label set and a second candidate label set. Both the first and second candidate label sets may include multiple candidate labels.

[0087] The first label and the second label are determined from the first candidate label set and the second candidate label set, respectively, based on at least one of the following: rule hit count, feature weight, number of modal sources, priority of enhanced input sources, historical review pass rate, and candidate label score.

[0088] Optionally, based on the feature parameters corresponding to the basic attribute symbols, at least two different dimensions of first and second labels are generated, which may specifically include: A second label mapping input dataset is constructed based on the basic attribute symbols and the feature parameters corresponding to the enhanced attribute symbols, wherein the feature parameters corresponding to the enhanced attribute symbols include at least one of the following: Enhance input source, material property field, assembly status field, text event field, version difference field, candidate label field, and confidence field.

[0089] Based on the second label mapping input dataset, feature matching is performed through prior constraints, or inference is performed through a pre-trained model to determine the first label and the second label.

[0090] Establishing associations among the region identifier, the basic attribute symbol, the first label, the second label, and the relationship label to generate and output the record data of the target entity may specifically include: The region identifier, the basic attribute symbol, the enhanced attribute symbol, the first label, the second label, and the relationship label are associated to generate and output the record data of the target entity.

[0091] The disclosure also provides a computer device including one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and executed by the one or more processors, and the one or more programs include instructions for a symbol unit library construction and update method according to one or more embodiments of the present disclosure.

[0092] Figure 2 This is a schematic diagram of the structure of an exemplary system for implementing a symbol unit library construction and update method according to an embodiment of this disclosure, such as... Figure 2 As shown, the system includes the following components: The data acquisition module 100 is configured to acquire images, videos, or continuous images, or, when the target entity exists, acquire material lists, assembly lists, associated text, user notes, or process notes, and acquire such text data.

[0093] The region localization module 200 is configured to perform region localization based on the acquired image data and generate region identifiers, region coordinates, bounding boxes, masks, or region contours.

[0094] The basic attribute symbol generation module 300 is configured to generate basic Symbol_IDs based on extracting region features from the region corresponding to the Region_ID.

[0095] The enhanced attribute symbol generation module 400 is configured to perform structured parsing of bill of materials, assembly list, associated text, or version fields to generate enhanced Symbol_ID or enhanced fields.

[0096] The symbol record generation module 500 is configured to bind Symbol_ID with fields such as Region_ID, Feature_Value, Modality_Source, Extraction_Rule, and Symbol_Version_No to a Symbol_Record.

[0097] The Symbol Unit Library Module 600 is configured to store, index, query, and invoke Symbol_Record; it can be implemented as a data set, an index table, a graph structure, or a combination thereof.

[0098] The dynamic update module 700 is configured to receive structured feedback input and perform operations such as adding, merging, splitting, disabling, weight adjustment, scenario adjustment, Symbol_Version_No update, or log writing based on update trigger conditions. The structured feedback input may include at least one of the following: relationship confidence, relationship verification flag, version change record, field difference record, and version verification result.

[0099] The output module 800 is called and configured to output Symbol_ID_List, Symbol_Record, Symbol_Update_Record, or Symbol_Library_Index for subsequent calls.

[0100] Figure 3 This is a flowchart of an exemplary symbol unit library construction and update method according to an embodiment of this disclosure, such as... Figure 3 As shown, the method includes the following steps: In step S302, image data of the artwork or object is acquired.

[0101] In step S304, region positioning is performed based on the acquired image data to generate at least one region identifier and region coordinates.

[0102] In step S306, features within the localization area are extracted.

[0103] In step S308, basic attribute identifiers are generated based on the extracted features.

[0104] In step S310, the text data of the work or object is parsed to generate enhanced attribute identifiers.

[0105] In step S312, the basic attribute identifier and the enhanced attribute identifier are written into the symbol unit library.

[0106] In step S314, it is determined whether the update trigger condition is met.

[0107] In step S316, when the update trigger condition is met, the symbol records in the symbol unit library are updated.

[0108] In step S318, at least one of the following is output: a symbol identifier list, a symbol record, and a symbol update log.

[0109] Figure 4 This is a schematic diagram illustrating an exemplary symbol unit library update process according to an embodiment of this disclosure, such as... Figure 4 As shown, updating the symbol unit library involves the following processes.

[0110] In step S402, the region identifier, region feature, first label, and second label are stored in the symbol unit library.

[0111] In step S404, the relation label, relation confidence level, and relation verification flag are stored in the symbol unit library.

[0112] In step S406, the version association record, verification result, and field difference record are stored in the symbol unit library.

[0113] Among them, the version association record, verification result, and field difference record are generated during the hash chain verification process of the structured work evidence data version of the target entity.

[0114] The symbol cell library stores symbol records, symbol cell library indexes, and update logs.

[0115] In step S408, when the update triggering condition of the symbol unit library is met, an update action is performed. The update action includes, but is not limited to, adding, merging, splitting, disabling, weight adjustment, scene change, and symbol version number change.

[0116] In step S410, after performing the update action, update data is output, such as a list of basic attribute symbol identifiers and symbol update records.

[0117] One or more embodiments of this disclosure also provide a non-volatile computer-readable storage medium containing a computer program that, when executed by one or more processors, causes the one or more processors to perform a symbol unit library construction and update method according to one or more embodiments of this disclosure.

[0118] One or more embodiments of this disclosure also provide a computer program product, including one or more computer programs, which, when executed by one or more processors, implement the symbol unit library construction and update method of one or more embodiments of this disclosure.

[0119] It should be noted that the method of this disclosure embodiment can be executed by a single device, such as a computer or server. The method of this embodiment can also be applied to a distributed scenario, where multiple devices cooperate to complete the task. In such a distributed scenario, one of these devices may execute only one or more steps of the method of this disclosure embodiment, and the multiple devices will interact with each other to complete the method described.

[0120] It should be noted that the above description describes some embodiments of this disclosure. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in a different order than that shown in the above embodiments and still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0121] For ease of description, the above computer devices are described in terms of function, divided into various modules. Of course, in implementing this disclosure, the functions of each module can be implemented in one or more software and / or hardware.

[0122] The computer device described in the above embodiments is used to implement the corresponding symbol unit library construction and update method in any of the foregoing embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0123] This disclosure also provides a computer device for implementing the above-described method for constructing and updating a symbol unit library. Figure 5 A schematic diagram of the hardware structure of an exemplary computer device 500 provided in an embodiment of this disclosure is shown.

[0124] like Figure 5 As shown, the computer device 500 may include: a processor 502, a memory 504, a network interface 506, a peripheral interface 508, and a bus 510. The processor 502, memory 504, network interface 506, and peripheral interface 508 are interconnected within the computer device 500 via the bus 510.

[0125] Processor 502 may be a central processing unit (CPU), image processor, neural network processor (NPU), microcontroller (MCU), programmable logic device, digital signal processor (DSP), application-specific integrated circuit (ASIC), or one or more integrated circuits. Processor 502 can be used to perform functions related to the techniques described in this disclosure. In some embodiments, processor 502 may also include multiple processors integrated as a single logic component. For example, such as... Figure 5 As shown, processor 502 may include multiple processors 502a, 502b and 502c.

[0126] Memory 504 can be configured to store data (e.g., instructions, computer code, etc.). Figure 5 As shown, the data stored in memory 504 may include program instructions (e.g., one or more programs for implementing the symbol unit library construction and update method of embodiments of this disclosure) and data to be processed (e.g., the memory may store configuration files of other modules, etc.). Processor 502 may also access the program instructions and data stored in memory 504 and execute the program instructions to operate on the data to be processed. Memory 504 may include volatile storage devices or non-volatile storage devices. In some embodiments, memory 504 may include random access memory (RAM), read-only memory (ROM), optical disk, magnetic disk, hard disk, solid-state drive (SSD), flash memory, memory stick, etc.

[0127] Network interface 506 can be configured to provide communication with other external devices to computer device 500 via a network. This network can be any wired or wireless network capable of transmitting and receiving data. For example, the network can be a wired network, a local wireless network (e.g., Bluetooth, WiFi, Near Field Communication (NFC), etc.), a cellular network, the Internet, or a combination thereof. It is understood that the type of network is not limited to the specific examples described above.

[0128] Peripheral interface 508 can be configured to connect computer device 500 to one or more peripheral devices to enable information input and output. For example, peripheral devices may include input devices such as keyboard, mouse, touchpad, touch screen, microphone, and various sensors, as well as output devices such as monitor, speaker, vibrator, and indicator lights.

[0129] Bus 510 can be configured to transfer information between various components of computer device 500 (such as processor 502, memory 504, network interface 506, and peripheral interface 508), such as internal buses (e.g., processor-memory bus), external buses (USB port, PCI-E bus), etc.

[0130] It should be noted that although the architecture of the computer device 500 described above only shows the processor 502, memory 504, network interface 506, peripheral interface 508, and bus 510, in specific implementations, the architecture of the computer device 500 may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the architecture of the computer device 500 described above may only include the components necessary for implementing the embodiments of this disclosure, and does not necessarily include all the components shown in the figures.

[0131] Based on the same inventive concept, corresponding to any of the above embodiments, this disclosure also provides a non-volatile computer-readable storage medium containing a computer program, which, when executed by one or more processors, causes the one or more processors to perform the symbol unit library construction and update method.

[0132] The computer-readable medium of this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0133] The computer program stored in the storage medium of the above embodiments is used to cause the one or more processors to execute the symbol unit library construction and update method as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0134] Based on the same inventive concept, corresponding to the symbol unit library construction and updating method of any of the above embodiments, this disclosure also provides a computer program product, which includes one or more computer programs. In some embodiments, the one or more computer programs are executable by one or more processors to cause the one or more processors to perform the symbol unit library construction and updating method. Corresponding to the execution entity for each step in each embodiment of the symbol unit library construction and updating method, the processor executing the corresponding step may belong to the corresponding execution entity.

[0135] The computer program product of the above embodiments is used to cause the processor to execute the symbol unit library construction and update method as described in any of the above embodiments, and has the beneficial effects of the corresponding method embodiments, which will not be repeated here.

[0136] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this disclosure (including the claims) is limited to these examples; within the framework of this disclosure, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this disclosure as described above, which are not provided in detail for the sake of brevity.

[0137] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this disclosure, the provided drawings may or may not show well-known power / ground connections to integrated circuit (IC) chips and other components. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this disclosure, and this also takes into account the fact that the details of implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this disclosure will be implemented (i.e., these details should be fully understood by those skilled in the art). While specific details (e.g., circuits) have been set forth to describe exemplary embodiments of this disclosure, it will be apparent to those skilled in the art that the embodiments of this disclosure can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0138] Although this disclosure has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may be used with the embodiments discussed.

[0139] This disclosure is intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A method for constructing and updating a symbol unit library, characterized in that, include: Acquire image data of the target entity; Based on the image data, at least one target region of the target entity is located to obtain the region identifier corresponding to the target entity; Based on the image data, extract the basic attribute features of the target region; The attributes of the target region are classified according to the basic attribute features to obtain the basic attribute symbols corresponding to the target region; Establish the association between the region identifier, the basic attribute symbol, and the basic attribute feature to obtain the symbol record of the target entity and store the symbol record in the symbol unit library; In response to a query request for the symbol record, output the information of the symbol record corresponding to the query request; In response to the detection of a symbol record update trigger condition, the symbol record is updated.

2. The method according to claim 1, characterized in that, The method further includes: After classifying the attributes of the target region according to the basic attribute features to obtain the basic attribute symbols corresponding to the target region, at least two different dimensions of first label and second label are generated according to the feature parameters corresponding to the basic attribute symbols. Based on the region identifier, the basic attribute symbol, the first tag, the second tag, and at least one type of preset structured judgment information that matches the first tag and the second tag, a relationship tag is generated to characterize the association between the first tag and the second tag; The region identifier, the basic attribute symbol, the first label, the second label, and the relationship label are associated to generate and output the record data of the target entity.

3. The method according to claim 2, characterized in that, In response to a query request for the symbol record, information about the symbol record corresponding to the query request is output, including: Obtain a query request, wherein the query request includes a call object indicated by a key field and a call purpose, wherein the call object includes at least one piece of information in the symbol record, and the call purpose includes: generating record data of the target entity, verifying the tag relationship of the target entity, or verifying the hash chain of the record data of the target entity; If the purpose of the call is to generate record data of the target entity, then a list of symbol identifiers corresponding to the object identifier or the region identifier and a symbol record summary are output, wherein the object identifier is used to identify the target entity; If the purpose of the call is to determine the label relationship of the target entity, then at least one of the following will be output: symbol identifier list, symbol record, symbol type, feature value, modality source, candidate label weight, and symbol confidence. If the purpose of the call is to verify the hash chain of the record data of the target entity, then at least one of the following will be output: symbol update record, symbol version number, update log identifier, or symbol record digest.

4. The method according to claim 1, characterized in that, The method further includes: Detect whether text data exists in the target area. If it exists, acquire the text data, wherein the text data includes at least one of the following: bill of materials, assembly list, associated text, process notes, or version difference field. Enhanced attribute features of the target region are extracted based on the text data; Based on the enhanced attribute features, generate enhanced attribute symbols corresponding to the target region; Establish the association between the region identifier, the basic attribute symbol, and the basic attribute feature to obtain the symbol record of the target entity, including: Establish the association relationship between the region identifier, the basic attribute symbol, the basic attribute feature, the enhanced attribute symbol, and the enhanced attribute feature to obtain the symbol record of the target entity.

5. The method according to claim 4, characterized in that, Enhanced attribute features are extracted based on the text data, and enhanced attribute symbols corresponding to the target region are generated based on the enhanced attribute features, including: If a material list corresponding to the target area is detected, a material attribute field is generated based on the material list; If an assembly list corresponding to the target area is detected, an assembly status field is generated based on the assembly list; If associated text corresponding to the target region is detected, extract the text event field based on the associated text; If a process note corresponding to the target area is detected, the process field is extracted based on the process note. All fields generated based on the text data will form an enhanced attribute feature set; The enhanced attribute feature set is matched with preset enhanced symbol rules, candidate enhanced symbol identifier templates, or symbol records in the symbol unit library to obtain matching results, and enhanced attribute symbols are generated based on the matching results.

6. The method according to claim 1, characterized in that, The symbol records are stored in a symbol unit library, which stores symbol records for multiple entity objects. In response to detecting a symbol record update trigger condition, the symbol records are updated, including: If the matching degree between the basic attribute feature set or enhanced attribute feature set corresponding to the region identifier and the stored symbol record does not meet the preset matching degree, or if the text data corresponding to the region identifier is updated, then the symbol record corresponding to the region identifier is regenerated. If it is detected that at least two symbol identifiers have feature values, symbol types, modality sources, and candidate label mapping results that satisfy a preset similarity, one of the at least two symbol identifiers is determined as the main symbol identifier, and the symbol record information corresponding to the at least two symbol identifiers is merged. If at least one of the following is detected as a symbol identifier, satisfying the preset differentiation index: feature value, modality source, candidate label mapping result, symbol type, applicable scenario, or relationship determination result, the detected symbol identifier will be split into at least two different sub-symbol identifiers. If a symbol is detected to have a negative feedback result that meets the first preset index when it is called or queried, the status of the symbol that meets the first preset index is set to disabled. Adjust the candidate label weights of the symbol identifiers and their corresponding applicable scenarios based on the positive or negative feedback results obtained when the symbol identifiers are invoked or queried; If a change is detected in the key field corresponding to a signed record, a new symbolic record version identifier is generated.

7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: After the symbol record is updated, a symbol update log is generated. The symbol update log includes at least one of the following: update time, update action, trigger type, trigger source, original value, new value, associated region identifier, associated symbol identifier, associated label pair, association relationship verification flag, relationship confidence or verification result.

8. A computer device comprising one or more processors, a memory; and one or more programs, wherein the one or more programs are stored in the memory and executed by the one or more processors, the one or more programs comprising instructions for performing the method of any one of claims 1 to 7.

9. A non-volatile computer-readable storage medium comprising a computer program, which, when executed by one or more processors, causes the one or more processors to perform the method of any one of claims 1 to 7.

10. A computer program product comprising one or more computer programs that, when executed by one or more processors, implement the method as described in any one of claims 1 to 7.

Images