Block chain-based credible storing and tracing system for converged media news evidence

Abstract

The invention relates to the technical field of data processing, and discloses a block chain-based converged media news evidence credible storage and traceability system, which comprises an evidence collection and association module, an evidence standardization module, a digital fingerprint generation module, a block chain storage module, an evidence traceability verification module and an evidence state updating module, carrying out association binding on the fusion media evidence metadata to obtain a multi-modal evidence set; adding a timestamp and a data source identifier to the data in the multi-mode evidence set to obtain a standardized evidence data packet; performing hash operation on the standardized evidence data packet to obtain a first digital fingerprint; mapping with a timestamp to a block chain network to obtain a permanent evidence storage record; when evidence tracing is carried out, consistency comparison is carried out on the evidence data to be verified, and a tracing verification conclusion is obtained; according to the traceability verification conclusion, carrying out evidence state updating on the to-be-verified evidence data; according to the method, the credible storing and tracing efficiency of the converged media news evidence can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular to a fusion media news evidence credible storage and traceability system based on a blockchain. BACKGROUND

[0002] The prior art has significant deficiencies in the fusion media news evidence collection and association link. The semantic and time dimension association processing of the fusion media evidence metadata constituting the news event is not carried out systematically. Only single modal evidence such as text, image and video is simply summarized. The semantic feature vector and time context information of each modal evidence are not extracted. The evidence association relationship based on semantic correlation and time continuity cannot be constructed. This leads to logical discontinuity and weak correlation among multi-modal evidences, and it is difficult to form an evidence set that fully reflects the context of the news event. At the same time, the semantic coherence and time logic rationality between evidences are not verified, so that there is invalid information with semantic conflict and time disorder in the collected evidence. The quality of the basic data provided for subsequent evidence standardization and storage is low, which directly affects the integrity and credibility of evidence storage.

[0003] The prior art has significant deficiencies in the fusion media news evidence storage security and traceability accuracy link. When generating evidence identification, layered hash operation and hash tree structure are not used to construct a unique digital fingerprint. Only a single identification is generated by simple hash processing. The uniqueness and tamper resistance of the evidence identification cannot be ensured. The evidence is easy to tamper with and difficult to detect during transmission and storage. When storing evidence, the blockchain network configuration is not dynamically selected in combination with the evidence security level and access control strategy. The stored data is also not layered encrypted, which leads to insufficient evidence storage security and cannot meet the evidence storage requirements of different security needs. In the traceability verification stage, the multi-modal composite feature vector of the evidence to be verified is not extracted. Only a single feature or fixed reference value is used for consistency comparison. The weight difference of the evidence feature and the dynamic adjustment of the verification standard are not considered. This leads to low traceability accuracy, easy misjudgment and omission, and the evidence state after verification is not marked and versioned. The stored evidence state change trajectory cannot be traced, and the demand for credible storage and accurate traceability of fusion media news evidence cannot be met. SUMMARY

[0004] The present application provides a fusion media news evidence credible storage and traceability system based on a blockchain to solve the problems raised in the background art.

[0005] To achieve the above purpose, the fusion media news evidence credible storage and traceability system based on a blockchain provided by the present application is characterized in that the system comprises an information extraction module, a commodity verification module, a verification failure module, a verification success module, a commodity settlement module and a settlement success module, wherein: The evidence collection and association module is configured to collect multi-media evidence metadata constituting a news event, associate and bind the multi-media evidence metadata, and obtain a multi-modal evidence set of the multi-media evidence metadata. The evidence standardization module is configured to attach a timestamp and a data source identifier in a unified format to data in the multi-modal evidence set, and obtain a standardized evidence data package of the multi-modal evidence set. The digital fingerprint generation module is configured to perform a hash operation on the standardized evidence data package, and obtain a first digital fingerprint of the standardized evidence data package. The blockchain evidence storage module is configured to map the first digital fingerprint and the timestamp to a blockchain network, to generate a permanent evidence storage record of the standardized evidence data package. The evidence traceability verification module is configured to, when performing evidence traceability, perform consistency comparison on to-be-verified evidence data based on the first digital fingerprint in the permanent evidence storage record, and obtain a traceability verification conclusion of the to-be-verified evidence data. The evidence state updating module is configured to update an evidence state of the to-be-verified evidence data according to the traceability verification conclusion.

[0006] In a preferred embodiment, when the evidence collection and association module is collecting multi-media evidence metadata constituting a news event, associating and binding the multi-media evidence metadata, and obtaining a multi-modal evidence set of the multi-media evidence metadata, the evidence collection and association module is specifically configured to: extract semantic feature vectors and temporal context information in the multi-media evidence metadata; construct a multi-modal evidence association graph of the multi-media evidence metadata, with modal evidence elements in the semantic feature vectors as nodes, and semantic correlation and temporal continuity as edges; verify semantic coherence between nodes and temporal logic rationality of edges in the multi-modal evidence association graph, to obtain a verified association graph of the multi-media evidence metadata; integrate semantic features of text, image and video based on the verified association graph, to generate a multi-modal evidence set of the multi-media evidence metadata.

[0007] In a preferred embodiment, when the evidence standardization module is attaching a timestamp and a data source identifier in a unified format to data in the multi-modal evidence set, and obtaining a standardized evidence data package of the multi-modal evidence set, the evidence standardization module is specifically configured to: perform spatiotemporal feature analysis on the multi-modal evidence set, to obtain a collection time sequence of the multi-modal evidence set; construct a multi-dimensional credibility evaluation system of the multi-modal evidence set according to semantic content and context relationship of the multi-modal evidence set. Based on the aforementioned multi-dimensional credibility assessment system, the comprehensive credibility weight of the multimodal evidence set is calculated, wherein the formula for calculating the comprehensive credibility weight is:

[0008] In the formula, The comprehensive credibility weight is... As a dynamic adjustment factor, The semantic consistency score is calculated based on the semantic matching degree between the evidence content and the event topic. The contextual coherence score is obtained by analyzing the logical coherence of evidence in the timeline and event flow. Based on the comprehensive credibility weight, dynamic time correction is performed on the collected time series to obtain a unified timestamp sequence of the collected time series; Based on the preset blockchain identity authentication mechanism, a decentralized identifier for the data source is generated; Digital signatures are applied to the unified timestamp sequence and decentralized identifier to obtain standardized metadata for the multimodal evidence set; The standardized metadata is semantically associated and bound with the corresponding multimodal evidence data to generate a standardized evidence data package for the multimodal evidence set.

[0009] In a preferred embodiment, when the digital fingerprint generation module performs a hash operation on the standardized evidence data packet to obtain the first digital fingerprint of the standardized evidence data packet, it is specifically used for: The standardized evidence data packet is divided into blocks to obtain a set of data blocks for the standardized evidence data packet; A one-way hash transformation is performed on the data blocks in the data block set to obtain the data block hash value set of the standardized evidence data packet; The standardized evidence data packet is constructed by using the data block hash values ​​in the data block hash value set as leaf nodes and the combined hash values ​​of the nodes as non-leaf nodes. A deterministic hash transformation is performed on the root node of the hash tree structure to obtain the first digital fingerprint of the standardized evidence data packet.

[0010] In a preferred embodiment, when the digital fingerprint generation module performs a one-way hash transformation on the data blocks in the data block set to obtain the data block hash value set of the standardized evidence data packet, it is specifically used for: The species attributes and growth stage context information in the data block set are integrated into the dryland crop growth semantic features of the standardized evidence data package; Based on the semantic features of dryland crop growth, the features of crop growth cycle and data block content are refined to obtain the semantic enhancement factor of the standardized evidence data package. The semantic enhancement factor and the corresponding data block are subjected to enhanced hashing to obtain the enhanced hash value of the standardized evidence data packet; The enhanced hash value is verified and encapsulated to obtain the data block hash value set of the standardized evidence data packet.

[0011] In a preferred embodiment, when the blockchain evidence storage module maps the first digital fingerprint and the timestamp to the blockchain network to generate a permanent evidence storage record of the standardized evidence data packet, it is specifically used for: Based on semantic association, the first digital fingerprint, the timestamp, the data source identifier, and the evidence content summary are semantically associated and encapsulated to generate the intelligent evidence storage data block of the standardized evidence data package; The intelligent evidence storage data block is subjected to layered encryption processing to obtain the encrypted evidence storage data packet of the standardized evidence data packet; Based on the security level requirements and access control policies in the standardized evidence data package, the optimal blockchain network configuration of the standardized evidence data package is dynamically selected; The encrypted evidence data packet is transmitted to the blockchain network to obtain a permanent evidence record of the standardized evidence data packet.

[0012] In a preferred embodiment, when the blockchain evidence storage module performs layered encryption processing on the smart evidence storage data block to obtain the encrypted evidence storage data packet of the standardized evidence data packet, it is specifically used for: Semantic feature parsing is performed on the intelligent evidence storage data block to obtain the metadata feature set of the standardized evidence data packet. Based on the metadata feature set, construct a dynamic encryption strategy for the standardized evidence data packet; Based on the dynamic encryption strategy, the intelligent evidence storage data block is subjected to multi-level cryptographic transformation to obtain the encrypted intelligent evidence storage data block of the standardized evidence data packet. The encrypted intelligent evidence storage data block is associated and stored with preset crop growth characteristic metadata to obtain the encrypted evidence storage data packet of the standardized evidence data packet.

[0013] In a preferred embodiment, when the evidence tracing and verification module performs a consistency comparison of the evidence data to be verified based on the first digital fingerprint in the permanent evidence record during evidence tracing, and obtains the tracing and verification conclusion of the evidence data to be verified, it is specifically used for: Multimodal feature extraction is performed on the evidence data to be verified to obtain a composite feature vector of the evidence data to be verified; Based on the weight coefficients of each feature for crop types, the weighted similarity between the composite feature vector and the first digital fingerprint is calculated. Based on the differences in data stability at different stages of the crop growth cycle, the verification benchmark value of the evidence data to be verified is dynamically adjusted. By comparing and analyzing the weighted similarity with the verification benchmark value, the source verification conclusion of the evidence data to be verified is obtained.

[0014] In a preferred embodiment, the formula for calculating the weighted similarity is: ; In the formula, The weighted similarity, The total number of feature dimensions of the composite feature vector. The weighting coefficients are... This represents the composite features of the first digital fingerprint. This represents the composite features of the evidence data to be verified. This is the function for calculating feature similarity.

[0015] In a preferred embodiment, when the evidence status update module updates the evidence status of the evidence data to be verified based on the source tracing verification conclusion, it is specifically used for: The source tracing verification conclusion is analyzed to obtain the verification consistency judgment result and credibility rating of the source tracing verification conclusion; Based on the credibility rating and the characteristics of the current growth stage of dryland crops, the evidence lifecycle status of the evidence data to be verified is determined; Based on the evidence lifecycle status, the evidence data to be verified is classified and marked with different statuses to obtain the updated evidence status record of the evidence data to be verified. The updated evidence status record is version-associated with the permanent evidence record to obtain the evidence status update trajectory of the evidence data to be verified.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention lays a high-quality data foundation for the credible preservation of evidence in converged media news through refined evidence collection, association, and standardization processing. The evidence collection and association module extracts semantic feature vectors and temporal context information from the metadata of converged media evidence, constructs a multimodal evidence association graph, verifies semantic coherence and temporal logic, and integrates multimodal semantic features to generate a multimodal evidence set, ensuring close association and clear logic among the evidence. The evidence standardization module calculates a comprehensive credibility weight based on a multi-dimensional credibility assessment system, dynamically corrects the collection time series to generate a unified timestamp, and combines decentralized identifiers to generate standardized evidence data packages, giving the evidence unified and traceable identification information, providing standardized and reliable data support for subsequent evidence preservation and tracing.

[0017] 2. This invention significantly improves the security and traceability efficiency of evidence storage in converged media news by leveraging secure digital fingerprint generation, blockchain-based evidence storage, and precise source tracing optimization. The digital fingerprint generation module processes standardized evidence data packets in blocks and constructs a hash tree. A unique first digital fingerprint is generated through root node hash transformation, ensuring the uniqueness and integrity of the evidence. The blockchain-based evidence storage module encapsulates and encrypts the digital fingerprint with a semantic association of timestamps, dynamically selecting the optimal blockchain network configuration for permanent storage, ensuring the evidence is tamper-proof and permanently traceable. The evidence source tracing and verification module extracts multimodal features of the evidence to be verified, calculates the weighted similarity to the stored fingerprint, and dynamically adjusts the verification benchmark to accurately arrive at the source tracing conclusion. The evidence status update module marks the evidence status based on the conclusion's hierarchical classification and associates it with the stored records, forming a complete status trajectory. This process achieves security, precision, and efficiency in evidence storage and source tracing throughout the entire process. Attached Figure Description

[0018] Figure 1 The system architecture diagram of a blockchain-based trusted evidence storage and traceability system for converged media news is provided in an embodiment of the present invention. The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments belong to some, but not all, embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “said” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0021] Depending on the context, the word "if" or "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0022] Furthermore, the timing of the steps in the following method embodiments is merely an example and not a strict limitation.

[0023] In practice, the server-side equipment deployed in the blockchain-based trusted evidence storage and traceability system for converged media news may consist of one or more devices. This system can be implemented as a business instance, a virtual machine, or hardware devices. For example, it can be implemented as a business instance deployed on one or more devices in a cloud node. Simply put, it can be understood as software deployed on a cloud node, providing trusted evidence storage and traceability services based on blockchain technology to various user terminals. Alternatively, it can be implemented as a virtual machine deployed on one or more devices in a cloud node, with application software installed to manage various user terminals. Or, it can also be implemented as a server-side system composed of numerous identical or different types of hardware devices, with one or more devices configured to provide trusted evidence storage and traceability services based on blockchain technology to various user terminals.

[0024] In terms of implementation, the blockchain-based trusted evidence storage and traceability system for converged media news and the user terminal are mutually compatible. Specifically, if the blockchain-based trusted evidence storage and traceability system for converged media news is implemented as an application installed on a cloud service platform, then the user terminal acts as a client establishing a communication connection with that application; or if the system is implemented as a website, then the user terminal acts as a webpage; or if the system is implemented as a cloud service platform, then the user terminal acts as a mini-program within an instant messaging application.

[0025] like Figure 1 The diagram shown is a system architecture diagram of a blockchain-based system for credible storage and traceability of converged media news evidence, provided in an embodiment of the present invention.

[0026] The blockchain-based trusted evidence storage and traceability system 100 for converged media news described in this invention can be set up in a cloud server. In terms of implementation, it can be used as one or more service devices, or as an application installed in the cloud (e.g., a mobile service operator's server, server cluster, etc.), or it can be developed into a website. Depending on the functions implemented, the blockchain-based trusted evidence storage and traceability system 100 for converged media news can include an evidence collection and association module 101, an evidence standardization module 102, a digital fingerprint generation module 103, a blockchain evidence storage module 104, an evidence traceability and verification module 105, and an evidence status update module 106. The modules described in this invention can also be called units, referring to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, stored in the memory of the electronic device.

[0027] In this embodiment of the invention, in the blockchain-based trusted storage and traceability system for converged media news evidence, each of the above modules can be implemented independently and can call other modules. Here, "calling" can be understood as one module connecting to multiple modules of another type and providing corresponding services to those connected modules. In the blockchain-based trusted storage and traceability system for converged media news evidence provided by this embodiment of the invention, the applicable scope of the system architecture can be adjusted by adding modules and directly calling them without modifying the program code, achieving cluster-based horizontal expansion to quickly and flexibly expand the system. In practical applications, the above modules can be set in the same device or different devices, or they can be set in virtual devices, such as service instances in a cloud server.

[0028] The following describes, with reference to specific embodiments, each component and its specific workflow of a blockchain-based integrated media news evidence trust storage and traceability system: The evidence collection and association module 101 is used to collect multimedia evidence metadata that constitutes a news event, associate and bind the multimedia evidence metadata, and obtain a multimodal evidence set of the multimedia evidence metadata. In this embodiment of the invention, when the evidence collection and association module collects the multimedia evidence metadata constituting a news event, associates and binds the multimedia evidence metadata, and obtains a multimodal evidence set of the multimedia evidence metadata, it is specifically used for: Extract the semantic feature vector and temporal context information from the metadata of the converged media evidence; Using the modal evidence elements in the semantic feature vector as nodes and semantic relevance and temporal continuity as edges, a multimodal evidence association graph of the converged media evidence metadata is constructed. Verify the semantic coherence between nodes and the temporal logic of edges in the multimodal evidence association graph to obtain the verified association graph of the converged media evidence metadata; Based on the verified association graph, the semantic features of text, images, and videos are integrated to generate a multimodal evidence set of the converged media evidence metadata.

[0029] Specifically, when extracting semantic feature vectors and temporal context information from the metadata of converged media evidence, the text, images, videos, and other content contained in the metadata are first parsed. For text, keywords, themes, and other information are extracted and converted into vectors that can represent semantics. For images, visual information such as objects and scenes is analyzed and converted into corresponding semantic vectors. For videos, semantics are extracted by combining the video content and audio information and vectors are formed. These vectors together constitute the semantic feature vector. At the same time, the publication time of the text, the shooting time of the image, the recording time of the video, and the time nodes of each content in the development of the event are recorded. This information is integrated into temporal context information.

[0030] Furthermore, when constructing a multimodal evidence association graph of converged media evidence metadata, using modal evidence elements in the semantic feature vector as nodes and semantic relevance and temporal continuity as edges, each modal evidence element corresponding to text, image, and video in the semantic feature vector is treated as an independent node in the graph. The semantic relevance between the modal evidence elements corresponding to any two nodes is analyzed; if the content described by the two nodes is related or involves the same event element, the semantic relevance is high. Simultaneously, their temporal sequence or simultaneity is analyzed; if they have a continuous or sequential relationship in the time flow, the temporal continuity is strong. Based on the strength of semantic relevance and temporal continuity, edges are established between corresponding nodes; the stronger the association, the more prominent the edge's attributes, thus constructing the multimodal evidence association graph.

[0031] Furthermore, when verifying the semantic coherence between nodes and the temporal logic of edges in the multimodal evidence association graph to obtain the verified association graph of the converged media evidence metadata, the process involves examining any two nodes connected by an edge in the graph and analyzing whether their corresponding modal evidence elements are semantically coherent and whether there are semantic conflicts or breaks. If conflicts exist, the corresponding edges are adjusted or removed. Simultaneously, the temporal continuity represented by each edge is checked to ensure it conforms to the temporal logic of event development, and whether there are any instances of reversed or contradictory time sequences. Edges that do not conform to logic are corrected or deleted. After verifying and adjusting all nodes and edges, the verified association graph is obtained.

[0032] Furthermore, when generating a multimodal evidence set of converged media evidence metadata by integrating the semantic features of text, images, and videos based on the verified association graph, the semantic features of text, images, and videos corresponding to each node are extracted according to the connection relationships of nodes in the verified association graph. Text semantic features, image semantic features, and video semantic features belonging to the same or related event stages are merged to ensure that the merged features fully reflect the information of that stage. All merged semantic features are organized according to the chronological or logical order of event development to form a set containing multifaceted and interconnected information from text, images, and videos; this set constitutes the multimodal evidence set of converged media evidence metadata.

[0033] In summary, by extracting semantic feature vectors from the metadata of converged media evidence, text keywords, image visual information, and video audio-visual content are transformed into associative semantic data. At the same time, the temporal context information of various types of evidence is recorded to ensure that subsequent association processing has clear semantic and temporal basis and avoids the break in association due to missing information.

[0034] In summary, by constructing a multimodal evidence association map with semantic relevance and temporal continuity as the core, scattered text, image, and video evidence are linked together according to event logic, so that clear relationships are formed between the evidence, the development process of news events is fully presented, and the problem of logical confusion caused by traditional simple summary of evidence is solved.

[0035] In summary, by verifying the semantic coherence of nodes and the temporal logic of edges, invalid evidence with semantic conflicts and temporal discrepancies is eliminated, and logical deviations in the association graph are corrected, ensuring that the final verified association graph is accurate and reliable, thus providing a guarantee for the subsequent generation of a high-quality multimodal evidence set.

[0036] In summary, by integrating semantic features of text, images, and videos based on the verified association graph, the multimodal evidence set not only contains the core information of various types of evidence but also maintains semantic and temporal consistency, thus comprehensively and accurately reflecting the whole picture of news events and providing high-quality data support for subsequent standardized processing and evidence preservation.

[0037] The evidence standardization module 102 is used to attach a unified format timestamp and data source identifier to the data in the multimodal evidence set to obtain a standardized evidence data package of the multimodal evidence set. In this embodiment of the invention, when the evidence standardization module appends a unified format timestamp and data source identifier to the data in the multimodal evidence set to obtain a standardized evidence data package for the multimodal evidence set, it is specifically used for: Spatiotemporal feature analysis is performed on the multimodal evidence set to obtain the acquisition time series of the multimodal evidence set; Based on the semantic content and contextual relationships of the multimodal evidence set, a multi-dimensional credibility evaluation system for the multimodal evidence set is constructed. Based on the aforementioned multi-dimensional credibility assessment system, the comprehensive credibility weight of the multimodal evidence set is calculated, wherein the formula for calculating the comprehensive credibility weight is:

[0038] In the formula, The comprehensive credibility weight is... As a dynamic adjustment factor, The semantic consistency score is calculated based on the semantic matching degree between the evidence content and the event topic. The contextual coherence score is obtained by analyzing the logical coherence of evidence in the timeline and event flow. Based on the comprehensive credibility weight, dynamic time correction is performed on the collected time series to obtain a unified timestamp sequence of the collected time series; Based on the preset blockchain identity authentication mechanism, a decentralized identifier for the data source is generated; Digital signatures are applied to the unified timestamp sequence and decentralized identifier to obtain standardized metadata for the multimodal evidence set; The standardized metadata is semantically associated and bound with the corresponding multimodal evidence data to generate a standardized evidence data package for the multimodal evidence set.

[0039] Specifically, when performing spatiotemporal feature analysis on the multimodal evidence set to obtain its collection time series, the original collection time information of various types of evidence data, such as text, images, and videos, is extracted one by one. For text, the specific time of its publication or recording needs to be confirmed; for images, the shooting time needs to be extracted; and for videos, the start and end times of recording need to be obtained. Simultaneously, combining the spatial correlation of various types of evidence data within the news event—such as images and videos taken at the same location, and text generated within the same event scenario—this time information is arranged according to the chronological order of event development, ensuring that each point in time corresponds to specific evidence data, thus forming the collection time series of the multimodal evidence set.

[0040] Furthermore, when constructing a multi-dimensional credibility assessment system for multi-modal evidence sets based on their semantic content and contextual relationships, the semantic content of various types of evidence data is first analyzed to determine the completeness and accuracy of textual descriptions, the clarity and authenticity of images, and the coherence and tamper-proof nature of video content. Then, considering the contextual relationships between evidence data, consistency is checked in describing the same event elements, such as whether the event time mentioned in the text matches the image capture time, and whether the event process recorded in the video matches the event process recorded in the text. Based on the results of semantic content analysis and contextual relationship verification, multiple dimensions of credibility assessment are determined, including content completeness, accuracy of description, data consistency, and image clarity. Specific assessment criteria are set for each dimension, collectively forming a multi-dimensional credibility assessment system.

[0041] Furthermore, when calculating the overall credibility weight of the multimodal evidence set based on the multidimensional credibility assessment system, each type of evidence data in the multimodal evidence set is scored individually according to the assessment criteria of each dimension in the multidimensional credibility assessment system. For example, in the content completeness dimension, evidence that fully records key information of the event receives a high score, while evidence lacking key information receives a low score; in the data consistency dimension, evidence that does not contradict other evidence receives a high score, while evidence that does contradict receives a low score. Based on the importance of each assessment dimension in the overall credibility judgment, a fixed weight percentage is assigned to each dimension. The score of each type of evidence data in each dimension is multiplied by the corresponding dimension's weight, and then the weighted scores of all dimensions are summed to obtain the credibility score of each type of evidence data. Based on this score, the weight of each type of evidence data in the comprehensive analysis is determined, i.e., the overall credibility weight.

[0042] Furthermore, when obtaining a unified timestamp sequence of the collected time series by dynamically correcting the collected time series based on comprehensive credibility weights, it is first determined that evidence data with high comprehensive credibility weights have higher reliability of their original collection time information and can be used as the benchmark for time correction. For evidence data in the collected time series that deviate from the benchmark time, the degree of deviation is adjusted according to its own comprehensive credibility weight. Evidence data with lower weights, if its collection time deviates significantly from the benchmark time, needs to be corrected more significantly according to the trend of the benchmark time; evidence data with higher weights, if there is a small deviation, only needs to be fine-tuned to retain the original time characteristics. During the correction process, it is ensured that the time sequence of all evidence data conforms to the logic of event development. After the correction is completed, a unified format timestamp is assigned to each piece of evidence data, and these timestamps are arranged in sequence to form a unified timestamp sequence.

[0043] Furthermore, when generating a decentralized identifier for the data source according to the preset blockchain identity authentication mechanism, the basic information of the data source is first obtained, including the identity information of the data provider, the unique identifier of the data acquisition device, and relevant information of the data generation platform. This basic information is then input into the preset blockchain identity authentication mechanism, which encrypts the information and transforms it into a unique character sequence using a specific encoding method. This character sequence is not dependent on a central institution for management and can be uniquely identified within the blockchain network; it serves as the decentralized identifier of the data source, used to identify the origin of the data and ensure the traceability of the data source.

[0044] Furthermore, when obtaining standardized metadata for the multimodal evidence set by digitally signing the unified timestamp sequence and decentralized identifiers, a dedicated digital signature key is used to encrypt each timestamp and decentralized identifier in the unified timestamp sequence, generating a corresponding digital signature. This signature ensures that the unified timestamp sequence and decentralized identifiers are not tampered with during subsequent transmission and storage. The encrypted unified timestamp sequence, decentralized identifiers, and corresponding digital signatures are then integrated to supplement basic descriptive content such as category information and data format information of the multimodal evidence set, forming a structured data set, which constitutes the standardized metadata of the multimodal evidence set.

[0045] Furthermore, when generating a standardized evidence data package for a multimodal evidence set by semantically associating and binding standardized metadata with corresponding multimodal evidence data, the information in the standardized metadata, such as timestamps, data source identifiers, and evidence categories, is first analyzed to clarify the specific multimodal evidence data corresponding to each metadata entry, such as text data corresponding to a certain timestamp or image data corresponding to a certain decentralized identifier. Through semantic association technology, a mapping relationship between metadata and evidence data is established to ensure that each piece of evidence data can be accurately associated with metadata describing its time, source, category, and other information. The associated standardized metadata and multimodal evidence data are then packaged and integrated to form a complete data package, which is the standardized evidence data package for the multimodal evidence set.

[0046] Specifically, the dynamic adjustment factor is determined based on the type of evidence and the event scenario. If it is core text and the event is sudden, the value is biased to highlight the impact of semantic consistency score; if it is auxiliary image and the event is continuous, the value is biased to highlight the impact of contextual coherence score.

[0047] Furthermore, the semantic consistency score is derived by comparing the core keywords and thematic ideas of the evidence content with the core elements of the event theme. The score is based on the matching ratio and degree of fit, with a higher matching degree resulting in a higher score.

[0048] Furthermore, the contextual coherence score is derived by examining the matching degree between the evidence record time and the event timeline nodes, and the consistency between the evidence content and the logic before and after the event flow. A high score is obtained if the time matches and the logic is coherent.

[0049] Furthermore, the formula integrates semantic consistency score and contextual coherence score by dynamically adjusting factors to obtain a comprehensive credibility weight, which reflects both the relevance of the evidence content to the event topic and its logical rationality in the development of the event, objectively measuring the overall credibility of the evidence.

[0050] In summary, by generating collection time series through spatiotemporal feature analysis, the temporal order of multimodal evidence is clearly identified, providing a foundation for subsequent time correction and avoiding the impact of temporal confusion on the validity of the evidence. A multi-dimensional credibility assessment system is constructed by combining semantic content and contextual relationships, and then the weights are accurately calculated using a comprehensive credibility weight formula, objectively reflecting the credibility of the evidence and providing a reliable basis for time correction.

[0051] In summary, the system dynamically corrects the collected time series based on comprehensive credibility weights, generating a unified timestamp sequence to resolve time discrepancies between different pieces of evidence and ensure consistent time stamp standards. Decentralized identifiers are generated based on blockchain identity authentication mechanisms to clarify the source of evidence, and digital signatures ensure that standardized metadata is not tampered with, thereby enhancing data security.

[0052] In summary, by binding standardized metadata with the semantic association of multimodal evidence data, standardized evidence data packages are generated, providing standardized and high-quality data support for subsequent blockchain evidence storage and traceability.

[0053] The digital fingerprint generation module 103 is used to perform a hash operation on the standardized evidence data packet to obtain the first digital fingerprint of the standardized evidence data packet. In this embodiment of the invention, when the digital fingerprint generation module performs a hash operation on the standardized evidence data packet to obtain the first digital fingerprint of the standardized evidence data packet, it is specifically used for: The standardized evidence data packet is divided into blocks to obtain a set of data blocks for the standardized evidence data packet; A one-way hash transformation is performed on the data blocks in the data block set to obtain the data block hash value set of the standardized evidence data packet; The standardized evidence data packet is constructed by using the data block hash values ​​in the data block hash value set as leaf nodes and the combined hash values ​​of the nodes as non-leaf nodes. A deterministic hash transformation is performed on the root node of the hash tree structure to obtain the first digital fingerprint of the standardized evidence data packet.

[0054] When the digital fingerprint generation module performs a one-way hash transformation on the data blocks in the data block set to obtain the data block hash value set of the standardized evidence data packet, it is specifically used for: The species attributes and growth stage context information in the data block set are integrated into the dryland crop growth semantic features of the standardized evidence data package; Based on the semantic features of dryland crop growth, the features of crop growth cycle and data block content are refined to obtain the semantic enhancement factor of the standardized evidence data package. The semantic enhancement factor and the corresponding data block are subjected to enhanced hashing to obtain the enhanced hash value of the standardized evidence data packet; The enhanced hash value is verified and encapsulated to obtain the data block hash value set of the standardized evidence data packet.

[0055] Specifically, when dividing a standardized evidence data packet into blocks to obtain a set of data blocks, a fixed block size is first determined. This size must balance data processing efficiency and the accuracy of subsequent hash calculations. According to the set block size, starting from the beginning of the standardized evidence data packet, data segments of corresponding lengths are sequentially extracted, each segment forming an independent data block. If the remaining portion of the data packet is less than a block size, the remaining portion is treated as a separate data block, ensuring that the entire standardized evidence data packet is completely divided without data overlap. All the resulting data blocks together constitute the set of data blocks for the standardized evidence data packet.

[0056] Furthermore, when performing one-way hash transformation on the data blocks in the data block set to obtain the data block hash value set of the standardized evidence data packet, a specific one-way hash processing method is selected. This method can convert data blocks of arbitrary length into fixed-length character sequences, and it is impossible to deduce the original data block from the character sequence. Each data block in the data block set is input into this one-way hash processing flow, and each data block is processed byte by byte. A corresponding fixed-length character sequence is generated according to a specific character conversion rule. Each character sequence is the hash value of the corresponding data block. The hash values ​​of all data blocks are arranged in the order of the original data blocks in the data block set, forming the data block hash value set of the standardized evidence data packet.

[0057] Furthermore, when constructing the hash tree structure of the standardized evidence data packet using the data block hash values ​​as leaf nodes and the combined hash values ​​of nodes as non-leaf nodes, each hash value in the data block hash value set is first used as the bottom-level leaf node of the hash tree structure, ensuring that each leaf node corresponds one-to-one with the corresponding data block hash value. Then, starting from the bottom-level leaf node, the hash values ​​of two adjacent leaf nodes are combined, and the combined character sequence is subjected to a one-way hash transformation to generate a new hash value. This new hash value serves as the parent node of these two leaf nodes, i.e., a non-leaf node. If the number of leaf nodes is odd, the last leaf node is combined with itself and subjected to a one-way hash transformation to generate a parent node. This process is repeated layer by layer upwards, with each layer combining the hash values ​​of adjacent nodes to generate the parent node of the next layer, until a unique top-level node is finally formed. This top-level node is the root node of the hash tree structure, thus completing the construction of the hash tree structure of the standardized evidence data packet.

[0058] Furthermore, when performing a deterministic hash transformation on the root node of the hash tree structure to obtain the first digital fingerprint of the standardized evidence data packet, the same one-way hashing method as used in constructing the hash tree nodes is employed to ensure the consistency and determinism of the hash transformation. The hash value corresponding to the root node of the hash tree structure is used as input, and this hash value is processed according to fixed computational steps. Throughout the processing, the computational rules remain unchanged and are unaffected by external factors, ensuring that the same root node hash value always generates the same output result. The fixed-length character sequence generated after processing is the first digital fingerprint of the standardized evidence data packet, which uniquely identifies the corresponding standardized evidence data packet.

[0059] Specifically, when integrating the species attributes and growth stage context information from the data block set into the semantic features of dryland crop growth in the standardized evidence data package, the species attribute information of each data record is first extracted from the data block set, including the specific variety, category, and biological characteristics of the crop. Simultaneously, the growth stage context information of the corresponding crop is extracted, covering the specific stage in the growth cycle, the environmental conditions of that stage, and a description of the growth status. The species attribute information and growth stage context information corresponding to the same data block are then correlated and matched to ensure that they describe the characteristics of the same dryland crop at the same time. The correlated information is then organized according to the crop growth logic to form a semantic description that fully reflects the growth status of the dryland crop; this content constitutes the semantic features of dryland crop growth in the standardized evidence data package.

[0060] Furthermore, when refining the semantic enhancement factors of standardized evidence data packages based on the semantic features of dryland crop growth to include crop growth cycle features and data block content characteristics, the following steps are taken: First, analyze the crop growth cycle-related information contained in the semantic features of dryland crop growth, extracting typical characteristics of crops at different growth cycle stages, such as growth rate during germination and morphological characteristics during maturity. These features constitute the crop growth cycle features. Simultaneously, analyze the data block content characteristics, extracting core data directly related to dryland crop growth and removing redundant and irrelevant information to obtain refined data block content characteristics. Then, integrate the refined crop growth cycle features with the refined data block content characteristics, combining this with professional knowledge of dryland crop growth to select key feature combinations that distinguish crop growth states and reflect the core value of the data. This combination constitutes the semantic enhancement factor of the standardized evidence data package.

[0061] Furthermore, when obtaining the enhanced hash value of the standardized evidence data packet by performing enhanced hashing on the semantic enhancement factor and the corresponding data block, the semantic enhancement factor and the corresponding data block are first integrated. The semantic enhancement factor is embedded as additional information into the header or tail of the corresponding data block, forming an integrated data block containing core data and key semantic features. A specific hashing method is adopted, which, based on conventional hashing, pays extra attention to the semantic enhancement factor part in the integrated data block, focusing on this part of the information to ensure that the semantic features are fully integrated into the hash result. The entire data block is processed byte by byte according to fixed operation steps. Through character conversion, numerical operation, and other operations, a fixed-length character sequence is generated. This character sequence is the enhanced hash value of the standardized evidence data packet.

[0062] Furthermore, when verifying the integrity of the enhanced hash values ​​to obtain the standardized evidence data packet's data block hash value set, firstly, corresponding integrity verification information is generated for each enhanced hash value. This information is obtained by performing specific verification operations on the enhanced hash value and can be used to subsequently verify whether the hash value has been tampered with. Each enhanced hash value is associated and bound to its corresponding integrity verification information to ensure that each hash value can find corresponding verification evidence. Following the original order of the data blocks in the data block set, the enhanced hash values ​​associated with integrity verification information are arranged sequentially to form an ordered hash value set, which is the standardized evidence data packet's data block hash value set.

[0063] In summary, dividing standardized evidence data packets into blocks creates a data block set, avoiding omissions or errors caused by excessive data volume in single data processing, and laying a stable foundation for subsequent hash operations. A one-way hash transformation is performed on the data blocks to generate a set of data block hash values. Each data block corresponds to a unique hash value, and the original data cannot be derived from the hash value, thus initially ensuring the immutability of the data.

[0064] In summary, by constructing a hash tree with the data block hash value as the leaf node and the combined hash as the non-leaf node, data can be verified layer by layer through the tree structure. Tampering with any data block will cause a change in the tree structure, thereby improving the sensitivity of data tampering detection.

[0065] In summary, a deterministic hash transformation is performed on the root node of the hash tree to obtain the first digital fingerprint, ensuring that each standardized evidence data packet corresponds to a unique and stable digital fingerprint, providing accurate identity identification for subsequent blockchain evidence storage and traceability.

[0066] In summary, extracting species attributes and growth stage context information from data blocks forms semantic features of dryland crop growth, integrating hash processing with crop-specific information, and improving the correlation between fingerprints and crop-related evidence.

[0067] In summary, by optimizing crop growth cycle features and data block content characteristics based on the semantic features of dryland crop growth, a semantic enhancement factor is obtained, which highlights key information, reduces redundant interference, and provides a precise basis for subsequent hash processing.

[0068] In summary, combining semantic enhancement factors with data blocks for enhanced hashing generates reinforced hash values ​​that not only retain the characteristics of the data itself but also incorporate semantic information, further improving the data's resistance to tampering and reducing the risk of forgery.

[0069] In summary, the enhanced hash value is encapsulated with verification integrity to ensure that each hash value is verifiable and tamper-proof. The resulting data block hash value set has high reliability, laying a high-quality foundation for building a hash tree and generating the first digital fingerprint.

[0070] The blockchain evidence storage module 104 is used to map the first digital fingerprint and the timestamp to the blockchain network to generate a permanent evidence storage record of the standardized evidence data package. In this embodiment of the invention, when the blockchain evidence storage module maps the first digital fingerprint and the timestamp to the blockchain network to generate a permanent evidence storage record of the standardized evidence data packet, it is specifically used for: Based on semantic association, the first digital fingerprint, the timestamp, the data source identifier, and the evidence content summary are semantically associated and encapsulated to generate the intelligent evidence storage data block of the standardized evidence data package; The intelligent evidence storage data block is subjected to layered encryption processing to obtain the encrypted evidence storage data packet of the standardized evidence data packet; Based on the security level requirements and access control policies in the standardized evidence data package, the optimal blockchain network configuration of the standardized evidence data package is dynamically selected; The encrypted evidence data packet is transmitted to the blockchain network to obtain a permanent evidence record of the standardized evidence data packet.

[0071] When the blockchain evidence storage module performs layered encryption processing on the smart evidence storage data block to obtain the encrypted evidence storage data packet of the standardized evidence data packet, it is specifically used for: Semantic feature parsing is performed on the intelligent evidence storage data block to obtain the metadata feature set of the standardized evidence data packet. Based on the metadata feature set, construct a dynamic encryption strategy for the standardized evidence data packet; Based on the dynamic encryption strategy, the intelligent evidence storage data block is subjected to multi-level cryptographic transformation to obtain the encrypted intelligent evidence storage data block of the standardized evidence data packet. The encrypted intelligent evidence storage data block is associated and stored with preset crop growth characteristic metadata to obtain the encrypted evidence storage data packet of the standardized evidence data packet.

[0072] Specifically, when generating a standardized evidence data package intelligent evidence storage data block by semantically associating and encapsulating the first digital fingerprint, timestamp, data source identifier, and evidence content summary according to semantic relationships, the unique correspondence between the first digital fingerprint and the standardized evidence data package, the correspondence between the timestamp and the evidence collection time, the correspondence between the data source identifier and the data source, and the correspondence between the evidence content summary and the core information of the evidence are first clarified. Based on these semantic relationships, the above four types of information are organized in a fixed format to ensure that the position and description of each type of information are consistent. At the same time, semantic association identifiers are added during the encapsulation process to clearly mark the association logic between various types of information, so that the encapsulated data package can clearly reflect the semantic relationship between each element, ultimately forming a standardized evidence data package intelligent evidence storage data block.

[0073] Furthermore, when obtaining the encrypted evidence data package of the standardized evidence data package by performing layered encryption processing on the intelligent evidence data blocks, the intelligent evidence data blocks are first divided into different layers according to the importance of the information, such as the core identifier layer, the time information layer, and the content summary layer. Corresponding encryption methods are used for different layers: the core identifier layer uses high-strength encryption to ensure that the unique identifier is not leaked or tampered with; the time information layer uses medium-strength encryption to ensure the accuracy and security of the time data; and the content summary layer uses basic encryption to protect the information while also considering the efficiency of subsequent verification. After completing the encryption of each layer, the encrypted data from each layer is integrated according to the original hierarchical structure to form a complete encrypted data package, which is the encrypted evidence data package of the standardized evidence data package.

[0074] Furthermore, when dynamically selecting the optimal blockchain network configuration for standardized evidence data packages based on the security level requirements and access control policies in the standardized evidence data packages, the pre-set security level requirements in the standardized evidence data packages are first analyzed to clarify the specific standards for confidentiality, integrity, and availability of the data. For example, high security levels require data to be immutable and have strict access permissions, while low security levels can appropriately relax access restrictions. Simultaneously, access control policies are analyzed to determine which entities can access the data, the scope of access permissions, and the access methods. Combining the security level requirements and access control policies, the characteristics of different blockchain networks are compared, such as the strictness of access control in private chains, the degree of decentralization in public chains, and the collaborative security of consortium chains. A blockchain network that meets the current data security requirements and is compatible with the access control policies is selected, and the node configuration, consensus mechanism, and other parameters of this network are determined to form the optimal blockchain network configuration for the standardized evidence data packages.

[0075] Furthermore, when transmitting the encrypted evidence data packet to the blockchain network to obtain a permanent record of the standardized evidence data packet, a secure connection is first established with the selected blockchain network according to its configuration to ensure the security of data transmission. The encrypted evidence data packet is uploaded to the blockchain network through its node interface. Nodes in the network verify the packet according to a pre-defined consensus mechanism, checking its encryption integrity and semantic relevance. Upon successful verification, the encrypted evidence data packet is recorded in a blockchain block, forming a chain structure with other existing evidence data. Due to the immutable nature of the blockchain, this packet is permanently stored in the network, forming a permanent record of the standardized evidence data packet, which can be retrieved and verified at any time through the blockchain network's query function.

[0076] Specifically, when performing semantic feature parsing on intelligent evidence storage data blocks to obtain the metadata feature set of standardized evidence data packages, the core information contained in the intelligent evidence storage data blocks, such as the first digital fingerprint, timestamp, data source identifier, and evidence content summary, is first extracted. Then, the semantic relationships between these information are analyzed, such as the unique correspondence between the first digital fingerprint and the evidence content, and the association logic between the timestamp and the evidence collection scenario. Next, the attribute features of each type of information are identified, such as the uniqueness of the first digital fingerprint, the timeliness of the timestamp, the traceability of the data source identifier, and the summarization of the evidence content summary. These semantic relationship information and attribute features are then organized in a unified format to form a structured dataset, which is the metadata feature set of the standardized evidence data package.

[0077] Furthermore, when constructing a dynamic encryption strategy for standardized evidence data packets based on metadata feature sets, the security requirements of various features in the metadata feature set are first analyzed. For example, unique first digital fingerprints require the highest level of encryption protection to prevent tampering; time-sensitive timestamps require medium-level encryption protection to ensure time accuracy; data source identifiers with traceability functions require encryption to protect source privacy; and summary evidence content digests require basic encryption to balance security and verifiability. Then, considering the access scenarios for different features, the encryption triggering conditions and update mechanisms are determined. For instance, when the access subject corresponding to the data source identifier changes, the encryption method for that part is automatically updated. These security requirement analysis results and access scenario adaptation schemes are integrated to form a strategy that adopts differentiated encryption methods for different metadata features. This strategy is the dynamic encryption strategy for standardized evidence data packets.

[0078] Furthermore, when obtaining the encrypted intelligent evidence storage data block of the standardized evidence data package by performing multi-level cryptographic transformations on the intelligent evidence storage data block based on the dynamic encryption strategy, the intelligent evidence storage data block is processed in layers according to the encryption requirements for different metadata features in the dynamic encryption strategy. The first digital fingerprint, timestamp, data source identifier, and evidence content digest are divided into independent data layers. For each data layer, a corresponding cryptographic transformation method is used. The layer containing the first digital fingerprint is encrypted using an asymmetric encryption algorithm to generate a key pair; the private key is used for subsequent decryption, and the public key is used for data transmission verification. The layer containing the timestamp is encrypted using a symmetric encryption algorithm with a dedicated key to ensure that the time data is not tampered with. The layer containing the data source identifier uses hash encryption combined with the key, which protects the source information and facilitates traceability verification. The layer containing the evidence content digest uses simple hash encryption to ensure the integrity of the digest information. After completing the encryption of each layer, all encrypted data layers are integrated in the original semantic order to form the encrypted intelligent evidence storage data block, which is the encrypted intelligent evidence storage data block of the standardized evidence data package.

[0079] Furthermore, when associating the encrypted intelligent evidence storage data block with preset crop growth characteristic metadata to obtain the encrypted evidence storage data package of the standardized evidence data package, the preset crop growth characteristic metadata is first obtained. This metadata includes crop growth information related to the evidence data, such as crop variety, growth stage, and growth environment parameters. The evidence content summary in the encrypted intelligent evidence storage data block is analyzed to find descriptive information related to crop growth characteristics, such as "wheat lodging image during grain filling stage" mentioned in the evidence content summary. This descriptive information is matched with the "wheat" variety and "grain filling stage" growth stage in the preset crop growth characteristic metadata to establish a semantic association between the two. An association identifier for the crop growth characteristic metadata is added to the encrypted intelligent evidence storage data block using association technology to ensure that the two can locate each other. Then, the associated encrypted intelligent evidence storage data block and the preset crop growth characteristic metadata are packaged in a unified storage format to form a complete data package, which is the encrypted evidence storage data package of the standardized evidence data package.

[0080] In summary, by integrating core elements such as the first digital fingerprint and timestamp according to semantic relationships to generate intelligent evidence storage data blocks, the evidence storage data becomes logically coherent, facilitating rapid correlation and verification during subsequent tracing and avoiding information fragmentation. Layered encryption of the intelligent evidence storage data blocks yields encrypted evidence storage data packets, specifically protecting information of varying importance, effectively resisting the risks of tampering and leakage, and enhancing the security of the evidence storage data.

[0081] In summary, selecting the optimal blockchain network configuration based on security level requirements and access control policies ensures that the evidence storage environment matches data security needs, balancing security and applicability. Encrypted evidence storage data packets are then transmitted to the blockchain network to generate permanent evidence records. Leveraging the immutability of the blockchain, long-term traceability and tamper-proof nature are ensured, guaranteeing the credibility of the evidence storage.

[0082] In summary, parsing the semantic features of intelligent evidence storage data blocks yields a metadata feature set, clarifying the core attributes of the data and providing a precise basis for subsequent encryption strategy formulation, thus avoiding deviations in encryption direction. Based on the metadata feature set, a dynamic encryption strategy can be constructed, allowing for the matching and adaptation of encryption methods to different characteristic data, improving the targeting and flexibility of encryption and meeting diverse security needs.

[0083] In summary, multi-layered protection further enhances the data's resistance to tampering and leakage, ensuring the security of core information. Encrypted intelligent evidence storage data blocks are associated with preset crop growth characteristic metadata, giving the evidence data both security attributes and crop-related information, providing more comprehensive data support for subsequent traceability and verification.

[0084] The evidence tracing and verification module 105 is used to perform a consistency comparison of the evidence data to be verified based on the first digital fingerprint in the permanent evidence record when tracing evidence, and to obtain the tracing and verification conclusion of the evidence data to be verified. In this embodiment of the invention, when the evidence tracing and verification module performs a consistency comparison of the evidence data to be verified based on the first digital fingerprint in the permanent evidence record during evidence tracing, and obtains the tracing and verification conclusion of the evidence data to be verified, it is specifically used for: Multimodal feature extraction is performed on the evidence data to be verified to obtain a composite feature vector of the evidence data to be verified; Based on the weight coefficients of each feature for crop types, the weighted similarity between the composite feature vector and the first digital fingerprint is calculated. Based on the differences in data stability at different stages of the crop growth cycle, the verification benchmark value of the evidence data to be verified is dynamically adjusted. By comparing and analyzing the weighted similarity with the verification benchmark value, the source verification conclusion of the evidence data to be verified is obtained.

[0085] The formula for calculating the weighted similarity is: ; In the formula, The weighted similarity, The total number of feature dimensions of the composite feature vector. The weighting coefficients are... This represents the composite features of the first digital fingerprint. This represents the composite features of the evidence data to be verified. This is the function for calculating feature similarity.

[0086] Specifically, when extracting multimodal features from the evidence data to be verified to obtain a composite feature vector, the type of evidence data is first analyzed. If it contains text data, keywords and semantic expressions describing crop characteristics and event information are extracted from the text; if it contains image data, visual features such as crop morphology, color, and growing environment are extracted from the images; if it contains video data, dynamic scene features and relevant descriptive information from the audio are extracted. The extracted features from different modalities such as text, images, and videos are then converted into quantifiable feature data. Following a unified feature dimension standard, these multimodal feature data are integrated into a complete vector structure, which is the composite feature vector of the evidence data to be verified.

[0087] Furthermore, when calculating the weighted similarity between the composite feature vector and the first digital fingerprint based on the weight coefficients of each feature for crop types, the importance of different features in distinguishing crop types is first clarified. For example, morphological features and variety-specific features of crops have a greater impact on crop type identification and are assigned higher weight coefficients accordingly; while environmental auxiliary features have a smaller effect on distinguishing crop types and are assigned lower weight coefficients. The similarity of each feature data in the composite feature vector is calculated separately with the corresponding feature data in the first digital fingerprint to obtain the similarity value of a single feature. Then, the similarity value of each single feature is multiplied by its corresponding weight coefficient to obtain the weighted feature similarity. Finally, all weighted feature similarities are summed to obtain the weighted similarity between the composite feature vector and the first digital fingerprint.

[0088] Furthermore, when dynamically adjusting the validation benchmark value of the evidence data to be verified based on the differences in data stability at different stages of the crop growth cycle, the various stages of the crop growth cycle, such as germination, growth, maturity, and harvest, are first determined. The stability characteristics of the data at each stage are analyzed. For example, the morphology and texture data of mature crops are relatively stable with small fluctuations, while the data of crops in the growth stage changes more rapidly and has lower stability. For growth stages with high data stability, a higher validation benchmark value is set because stable data is more likely to be consistent with the first digital fingerprint and requires a more stringent matching standard. For growth stages with low data stability, the validation benchmark value is appropriately lowered to accommodate natural data changes and avoid misjudgments caused by normal data fluctuations. Based on the crop growth stage corresponding to the evidence data to be verified, the corresponding validation benchmark value is selected and adjusted to obtain a validation benchmark value suitable for the current stage.

[0089] Furthermore, when comparing the weighted similarity with the verification benchmark to obtain the source verification conclusion of the evidence data to be verified, the comparison rules are first clarified. If the weighted similarity is greater than or equal to the adjusted verification benchmark, it indicates that the evidence data to be verified has a high degree of matching with the data features in the permanent evidence record, and the data has not been tampered with or the deviation is within an acceptable range. If the weighted similarity is less than the verification benchmark, it indicates that there is a significant difference between the evidence data to be verified and the evidence data, and there may be data tampering or inconsistent data sources. According to this comparison rule, the calculated weighted similarity is compared with the adjusted verification benchmark. Based on the comparison result, it is determined whether the evidence data to be verified is consistent with the permanent evidence record, and finally a clear source verification conclusion is formed, such as "the evidence data to be verified is consistent with the evidence data, and the source verification is valid" or "the evidence data to be verified is inconsistent with the evidence data, and the source verification is invalid".

[0090] Specifically, the total number of feature dimensions of the composite feature vector comes from the multimodal feature extraction results of the evidence data to be verified. When extracting multimodal features such as text, images, and videos and integrating them into a composite feature vector, the number of feature dimensions contained in the vector is specified. This number is the total number of feature dimensions of the composite feature vector.

[0091] Furthermore, the weighting coefficients are determined based on the importance of each feature to the crop type. First, the role of different features in distinguishing crop types is analyzed. For example, crop morphological features and variety-specific features have a greater impact on crop type identification, while environmental auxiliary features have a smaller impact. Based on this difference in role, a corresponding weighting coefficient is assigned to each feature to ensure that important features account for a higher proportion in similarity calculation.

[0092] Furthermore, the composite feature representation of the first digital fingerprint comes from the permanent evidence record. When generating the first digital fingerprint from the standardized evidence data package, the multimodal features corresponding to the data package are extracted simultaneously and integrated into a composite feature. This composite feature is the composite feature representation of the first digital fingerprint, which is used for subsequent comparison with the features of the evidence data to be verified.

[0093] Furthermore, the composite feature representation of the evidence data to be verified comes from the multimodal feature extraction process of the evidence data to be verified. By extracting features from the text, image, video and other modalities in the evidence to be verified, they are transformed into quantifiable feature data and integrated into a vector structure. This vector structure is the composite feature representation of the evidence data to be verified.

[0094] Furthermore, the feature similarity calculation function is a fixed method for calculating the similarity of a single feature. Specifically, it compares the feature data of each dimension in the composite feature representation of the first digital fingerprint with the feature data of the corresponding dimension in the composite feature representation of the evidence data to be verified, and obtains the similarity value of a single feature by calculating the degree of difference between the two. The smaller the difference, the higher the similarity value, and the larger the difference, the lower the similarity value.

[0095] Furthermore, the significance of this formula lies in obtaining a weighted similarity through a reasonable calculation method to accurately measure the degree of matching between the evidence data to be verified and the first digital fingerprint. The calculation first multiplies the weight coefficient of each feature dimension with the feature similarity of the corresponding dimension, then sums all products to obtain the weighted total similarity. Next, the square root of the sum of squares of all weight coefficients is calculated as a normalization factor. Finally, the weighted total similarity is divided by the normalization factor to obtain the final weighted similarity. This calculation method considers the differences in importance of each feature to the crop species, highlighting the role of important features through weight coefficients, while avoiding calculation bias caused by the magnitude of the weight coefficients themselves through normalization. This ensures that the weighted similarity objectively and accurately reflects the matching situation between the two, providing a reliable basis for subsequent source tracing and verification conclusions.

[0096] In summary, multimodal features are extracted from the evidence data to generate composite feature vectors, covering multiple dimensions of information such as text and images. This avoids the bias caused by single features in comparisons and lays the foundation for accurate comparisons. Weighted similarity is calculated by combining the weight coefficients of features for crop types, highlighting the influence of key features, reducing interference from irrelevant features, and improving the accuracy of similarity calculations.

[0097] In summary, adjusting the verification benchmark value based on the stability of data at different stages of the crop growth cycle adapts to the natural variation patterns of the data, avoids misjudgments caused by fixed benchmarks, and improves verification flexibility. By comparing and analyzing weighted similarity with dynamic benchmark values, the resulting traceability verification conclusions are more consistent with reality, accurately assess the consistency between the evidence to be verified and the stored evidence, and ensure the credibility of the traceability results.

[0098] In summary, by weighting the similarity results of different features using weighting coefficients, the proportion of features more important for distinguishing crop types is increased in the calculation, reducing the interference of irrelevant features and improving the targeting of similarity calculation. Introducing the total number of feature dimensions of the composite feature vector and normalizing it using the square root of the sum of squares of the weighting coefficients avoids calculation bias caused by the magnitude of the weighting coefficients themselves, ensuring the comparability of similarity results under different feature dimensions and guaranteeing the objectivity of the calculation.

[0099] In summary, the composite feature representation of the first digital fingerprint and the composite feature representation of the evidence data to be verified are used as the calculation objects. The feature similarity calculation function is combined to accurately compare the feature differences between the two, providing accurate data support for subsequent source tracing and verification conclusions, and ensuring the accuracy of source tracing.

[0100] The evidence status update module 106 is used to update the evidence status of the evidence data to be verified based on the source tracing verification conclusion.

[0101] In this embodiment of the invention, when the evidence status update module updates the evidence status of the evidence data to be verified based on the source tracing verification conclusion, it is specifically used for: The source tracing verification conclusion is analyzed to obtain the verification consistency judgment result and credibility rating of the source tracing verification conclusion; Based on the credibility rating and the characteristics of the current growth stage of dryland crops, the evidence lifecycle status of the evidence data to be verified is determined; Based on the evidence lifecycle status, the evidence data to be verified is classified and marked with different statuses to obtain the updated evidence status record of the evidence data to be verified. The updated evidence status record is version-associated with the permanent evidence record to obtain the evidence status update trajectory of the evidence data to be verified.

[0102] Specifically, when analyzing the source tracing verification conclusions to obtain the verification consistency judgment results and credibility ratings, the consistency between the evidence data to be verified and the permanent evidence records is extracted from the source tracing verification conclusions as the verification consistency judgment results; based on the degree of comparison between the weighted similarity and the verification benchmark value and the tightness of feature matching, the rating of "high credibility", "medium credibility" or "low credibility" is determined.

[0103] Furthermore, when determining the evidence lifecycle status of the evidence data to be verified based on the credibility rating and the characteristics of the current growth stage of dryland crops, the current growth stage of the crop and the stability of the data at that stage are clearly defined. Combined with the credibility rating, evidence with "high credibility" and stable data stage is defined as "valid and usable", evidence with "medium credibility" or moderate stability stage is defined as "pending verification", and evidence with "low credibility" or low stability stage is defined as "invalid and discarded".

[0104] Furthermore, when obtaining the updated evidence status record of the evidence data to be verified by hierarchically marking the evidence data based on the evidence lifecycle status, first-level, second-level, and third-level marks are set for "valid and usable", "pending review" and "invalid and discarded" respectively. The marks and the corresponding time and the verification conclusions on which they are based are integrated to form the updated evidence status record.

[0105] Furthermore, when obtaining the evidence status update trajectory of the evidence data to be verified by versioning and associating the updated evidence status record with the permanent evidence record, a unique version identifier containing time and number is assigned to the updated evidence status record, which is bound to the evidence identifier of the permanent evidence record. The associated version records are stored in chronological order to form the evidence status update trajectory.

[0106] In summary, the analysis of the source tracing verification conclusions yielded consistency judgment results and credibility ratings, clearly extracting key verification information and providing an accurate basis for subsequent status determination. Adapting to crop growth characteristics: Combining the credibility rating with the current growth stage characteristics of dryland crops to determine the evidence lifecycle status ensures that the status determination aligns with actual crop growth, enhancing its rationality.

[0107] In summary, grading and labeling evidence based on its lifecycle status generates updated status records, making the evidence status clearly identifiable and facilitating management and identification. Linking and storing these updated status records with versioned permanent evidence records creates an evidence status update trajectory, allowing for tracing the process of evidence status changes and ensuring the continuity and traceability of evidence management.

[0108] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0109] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0110] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A blockchain-based system for trusted storage and traceability of converged media news evidence, characterized in that: The system includes an evidence collection and association module, an evidence standardization module, a digital fingerprint generation module, a blockchain evidence storage module, an evidence tracing and verification module, and an evidence status update module, wherein: The evidence collection and association module is used to collect multimedia evidence metadata that constitutes a news event, associate and bind the multimedia evidence metadata, and obtain a multimodal evidence set of the multimedia evidence metadata. The evidence standardization module is used to attach a unified format timestamp and data source identifier to the data in the multimodal evidence set to obtain a standardized evidence data package of the multimodal evidence set. The digital fingerprint generation module is used to perform a hash operation on the standardized evidence data packet to obtain the first digital fingerprint of the standardized evidence data packet; The blockchain evidence storage module is used to map the first digital fingerprint and the timestamp to the blockchain network to generate a permanent evidence storage record of the standardized evidence data package. The evidence tracing and verification module is used to perform a consistency comparison of the evidence data to be verified based on the first digital fingerprint in the permanent evidence record when tracing evidence, and to obtain the tracing and verification conclusion of the evidence data to be verified. The evidence status update module is used to update the evidence status of the evidence data to be verified based on the source tracing verification conclusion.

2. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 1, characterized in that, When the evidence collection and association module collects the multimedia evidence metadata constituting a news event, and associates and binds the multimedia evidence metadata to obtain a multimodal evidence set of the multimedia evidence metadata, it is specifically used for: Extract the semantic feature vector and temporal context information from the metadata of the converged media evidence; Using the modal evidence elements in the semantic feature vector as nodes and semantic relevance and temporal continuity as edges, a multimodal evidence association graph of the converged media evidence metadata is constructed. Verify the semantic coherence between nodes and the temporal logic of edges in the multimodal evidence association graph to obtain the verified association graph of the converged media evidence metadata; Based on the verified association graph, the semantic features of text, images, and videos are integrated to generate a multimodal evidence set of the converged media evidence metadata.

3. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 1, characterized in that, When the evidence standardization module appends a unified format timestamp and data source identifier to the data in the multimodal evidence set to obtain a standardized evidence data package for the multimodal evidence set, it is specifically used for: Spatiotemporal feature analysis is performed on the multimodal evidence set to obtain the acquisition time series of the multimodal evidence set; Based on the semantic content and contextual relationships of the multimodal evidence set, a multi-dimensional credibility evaluation system for the multimodal evidence set is constructed. Based on the aforementioned multi-dimensional credibility assessment system, the comprehensive credibility weight of the multimodal evidence set is calculated, wherein the formula for calculating the comprehensive credibility weight is: ; In the formula, The comprehensive credibility weight is... As a dynamic adjustment factor, The semantic consistency score is calculated based on the semantic matching degree between the evidence content and the event topic. The contextual coherence score is obtained by analyzing the logical coherence of evidence in the timeline and event flow. Based on the comprehensive credibility weight, dynamic time correction is performed on the collected time series to obtain a unified timestamp sequence of the collected time series; Based on the preset blockchain identity authentication mechanism, a decentralized identifier for the data source is generated; Digital signatures are applied to the unified timestamp sequence and the decentralized identifier to obtain standardized metadata for the multimodal evidence set; The standardized metadata is semantically associated and bound with the corresponding multimodal evidence data to generate a standardized evidence data package for the multimodal evidence set.

4. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 1, characterized in that, When the digital fingerprint generation module performs a hash operation on the standardized evidence data packet to obtain the first digital fingerprint of the standardized evidence data packet, it is specifically used for: The standardized evidence data packet is divided into blocks to obtain a set of data blocks for the standardized evidence data packet; A one-way hash transformation is performed on the data blocks in the data block set to obtain the data block hash value set of the standardized evidence data packet; The standardized evidence data packet is constructed by using the data block hash values ​​in the data block hash value set as leaf nodes and the combined hash values ​​of the nodes as non-leaf nodes. A deterministic hash transformation is performed on the root node of the hash tree structure to obtain the first digital fingerprint of the standardized evidence data packet.

5. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 4, characterized in that, When the digital fingerprint generation module performs a one-way hash transformation on the data blocks in the data block set to obtain the data block hash value set of the standardized evidence data packet, it is specifically used for: The species attributes and growth stage context information in the data block set are integrated into the dryland crop growth semantic features of the standardized evidence data package; Based on the semantic features of dryland crop growth, the features of crop growth cycle and data block content are refined to obtain the semantic enhancement factor of the standardized evidence data package. The semantic enhancement factor and the corresponding data block are subjected to enhanced hashing to obtain the enhanced hash value of the standardized evidence data packet; The enhanced hash value is verified and encapsulated to obtain the data block hash value set of the standardized evidence data packet.

6. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 1, characterized in that, When the blockchain evidence storage module maps the first digital fingerprint and the timestamp to the blockchain network to generate a permanent evidence storage record for the standardized evidence data packet, it is specifically used for: Based on semantic association, the first digital fingerprint, the timestamp, the data source identifier, and the evidence content summary are semantically associated and encapsulated to generate the intelligent evidence storage data block of the standardized evidence data package; The intelligent evidence storage data block is subjected to layered encryption processing to obtain the encrypted evidence storage data packet of the standardized evidence data packet; Based on the security level requirements and access control policies in the standardized evidence data package, the optimal blockchain network configuration of the standardized evidence data package is dynamically selected; The encrypted evidence data packet is transmitted to the blockchain network to obtain a permanent evidence record of the standardized evidence data packet.

7. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 6, characterized in that, When the blockchain evidence storage module performs layered encryption processing on the smart evidence storage data block to obtain the encrypted evidence storage data packet of the standardized evidence data packet, it is specifically used for: Semantic feature parsing is performed on the intelligent evidence storage data block to obtain the metadata feature set of the standardized evidence data packet. Based on the metadata feature set, construct a dynamic encryption strategy for the standardized evidence data packet; Based on the dynamic encryption strategy, the intelligent evidence storage data block is subjected to multi-level cryptographic transformation to obtain the encrypted intelligent evidence storage data block of the standardized evidence data packet. The encrypted intelligent evidence storage data block is associated and stored with preset crop growth characteristic metadata to obtain the encrypted evidence storage data packet of the standardized evidence data packet.

8. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 1, characterized in that, When performing evidence tracing and verification, the evidence tracing and verification module compares the consistency of the evidence data to be verified based on the first digital fingerprint in the permanent evidence record to obtain the tracing and verification conclusion of the evidence data to be verified. Specifically, it is used for: Multimodal feature extraction is performed on the evidence data to be verified to obtain a composite feature vector of the evidence data to be verified; Based on the weight coefficients of each feature for crop types, the weighted similarity between the composite feature vector and the first digital fingerprint is calculated. Based on the differences in data stability at different stages of the crop growth cycle, the verification benchmark value of the evidence data to be verified is dynamically adjusted. By comparing and analyzing the weighted similarity with the verification benchmark value, the source verification conclusion of the evidence data to be verified is obtained.

9. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 8, characterized in that, The formula for calculating the weighted similarity is: ； In the formula, The weighted similarity, The total number of feature dimensions of the composite feature vector. The weighting coefficients are... This represents the composite features of the first digital fingerprint. This represents the composite features of the evidence data to be verified. This is the function for calculating feature similarity.

10. The blockchain-based trusted evidence storage and traceability system for converged media news as described in claim 1, characterized in that, When the evidence status update module updates the evidence status of the evidence data to be verified based on the source tracing verification conclusion, it is specifically used for: The source tracing verification conclusion is analyzed to obtain the verification consistency judgment result and credibility rating of the source tracing verification conclusion; Based on the credibility rating and the characteristics of the current growth stage of dryland crops, the evidence lifecycle status of the evidence data to be verified is determined; Based on the evidence lifecycle status, the evidence data to be verified is classified and marked with different statuses to obtain the updated evidence status record of the evidence data to be verified. The updated evidence status record is version-associated with the permanent evidence record to obtain the evidence status update trajectory of the evidence data to be verified.

Images