Search

CN-122020732-A - Sign-in data integrity guaranteeing method and system for conference management

CN122020732ACN 122020732 ACN122020732 ACN 122020732ACN-122020732-A

Abstract

The invention relates to the technical field of conference data security and discloses a method and a system for guaranteeing the integrity of sign-in data for conference management, wherein the method comprises the steps of obtaining preregistration information of participants, and obtaining a local feature library of each sign-in terminal by adopting an edge intelligent feature extraction and distributed cache synchronization method; the method comprises the steps of obtaining tamper-proof check-in data packets by adopting a field quick matching and space-time anchoring method, obtaining blockchain records and final confirmation by adopting a layering consensus and blockchain storage method, obtaining data verification results and final disposal decisions by adopting a deep learning enhancement verification and intelligent arbitration method, obtaining globally consistent check-in data by adopting a vector clock and intelligent conflict resolution method, and obtaining privacy protection verification results by adopting a zero knowledge proof and homomorphic encryption method. The invention can realize the integrity guarantee and privacy protection of the check-in data from acquisition to verification of the full life cycle under the multi-meeting-place distributed check-in scene.

Inventors

  • GAO WEIGUO
  • HuangFu Sen

Assignees

  • 杭州昊雨信息技术有限公司

Dates

Publication Date
20260512
Application Date
20260129

Claims (10)

  1. 1. The sign-in data integrity guarantee method for the conference management is characterized by comprising the following steps of: s1, collecting preregistration information of participants, and obtaining a local feature library of each sign-in terminal by adopting an edge intelligent feature extraction and distributed cache synchronization method; S2, acquiring a local feature library of each sign-in terminal, and acquiring a tamper-proof sign-in data packet by adopting a field quick matching and space-time anchoring method; S3, obtaining a tamper-proof sign-in data packet, and obtaining a blockchain record and final confirmation by adopting a hierarchical consensus and blockchain storage method; S4, obtaining a blockchain record and final confirmation, and obtaining a data verification result and a final disposal decision by adopting a deep learning enhancement verification and intelligent arbitration method; S5, acquiring a data verification result and a final disposal decision, and obtaining globally consistent sign-in data by adopting a vector clock and an intelligent conflict resolution method; s6, obtaining globally consistent sign-in data, and obtaining privacy protection verification results by adopting a zero knowledge proof and homomorphic encryption method.
  2. 2. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein S1 comprises: The method comprises the steps of obtaining preregistration information of participants, and obtaining a standard avatar file containing unique reference numbers by adopting data cleaning and format standardization processing; Acquiring a standard avatar archive, acquiring a face image through a mobile terminal application, and performing quality evaluation on the image to obtain a face image data set; Acquiring a face image dataset, processing by adopting a quality-aware self-adaptive feature extraction network, wherein the feature extraction network adopts a double-branch architecture, a main branch is a lightweight convolutional neural network responsible for feature extraction, an auxiliary branch is a quality evaluation network responsible for predicting image quality and generating self-adaptive weights, and carrying out fusion modulation on the quality weights and original feature vectors to obtain face feature reference vectors; and acquiring a face feature reference vector, constructing a feature retrieval index by adopting a local sensitive hash algorithm, and synchronously distributing the feature retrieval index to each sign-in terminal through a distributed cache system to obtain a local feature library of each sign-in terminal.
  3. 3. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein the step S2 comprises: acquiring a local feature library of a sign-in terminal, acquiring face images in real time by adopting a camera of the sign-in terminal, and performing face alignment processing on the images to obtain face images to be identified; Acquiring a face image to be identified, adopting local feature extraction and hash retrieval, and calculating cosine similarity between a query feature vector and a reference vector to obtain a candidate matching identity list; Acquiring a candidate matching identity list, and carrying out identity confirmation decision by adopting a preset similarity judgment threshold value to obtain the confirmed identities of the participants; acquiring the identity of the confirmed participants, collecting equipment fingerprints and space-time information, and signing the sign-in metadata by adopting a national secret SM2 elliptic curve digital signature algorithm to obtain the tamper-proof sign-in data packet.
  4. 4. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein the step S3 comprises: Obtaining a tamper-proof sign-in data packet, and performing format integrity check, digital signature verification, equipment validity verification and timeliness verification by adopting an edge computing node to obtain a temporary confirmation result; the temporary confirmation result is obtained, the data packet which passes verification is temporarily stored by adopting a local buffer mechanism, and batch uplink time is determined by adopting a time triggering and capacity triggering mechanism, so as to obtain a queue to be uplink; Obtaining a queue to be uplink, adopting an improved practical Bayesian-busy fault-tolerant consensus algorithm, achieving consensus among verification nodes through three stages of preparation, preparation and submission, and introducing parallel processing and fast path optimization to obtain a transaction with consensus confirmation; and acquiring the transaction with the consensus confirmation, constructing an merck tree index for the transaction, generating a block header, calculating a block hash, writing the block into a block chain database, and sending a final confirmation message to a check-in terminal through an asynchronous notification mechanism to obtain a block chain record and final confirmation.
  5. 5. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein the step S4 comprises: deploying a hybrid verification architecture comprising a rule engine layer and a deep learning detection layer; The rule engine layer performs time window verification, geofence verification, format integrity verification and device validity verification to obtain a basic rule verification result; The deep learning detection layer adopts an anomaly detection neural network, the network structure is a multi-layer perceptron combined with a long-term and short-term memory network, multi-dimensional features of sign-in data are encoded into a feature vector input network, the network outputs probability distribution of various anomalies, and the probability distribution comprises repeated sign-in, space-time contradiction, anomaly high frequency and behavior mutation types, so that an anomaly mode recognition result is obtained; And integrating verification results of the rule engine and the deep learning detection layer, adding the data packets passing verification into a normal uplink queue, and adding the data packets with abnormal confidence exceeding a preset threshold into a queue to be arbitrated to obtain an under-chain verification result.
  6. 6. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein S4 further comprises: Deploying arbitration intelligent contracts on the block chain, wherein the contracts realize repeated sign-in detection functions, space-time contradiction analysis functions and abnormal frequency statistics functions; When the data to be arbitrated is linked up, the contract automatically calls a verification function, repeatedly checks whether the check-in detection function inquiry history record check time interval is smaller than a preset repeated threshold value, the space-time contradiction analysis function calculates whether the implicit speed judgment exceeds the preset speed upper limit, and the check-in quantity of the abnormal frequency statistics function statistics equipment in a time window judges whether the check-in quantity exceeds the preset frequency threshold value to obtain an abnormal identification result; and checking the sign-in record marked as abnormal by adopting a manual rechecking mechanism, and submitting a correction application or confirming the abnormal according to the check conclusion to obtain a final disposal decision.
  7. 7. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein S5 comprises: Distributing unique node identification for each sign-in edge node, maintaining N-dimensional vector clocks for each node, wherein N is the total number of nodes; When a node processes a check-in event, the corresponding vector clock component is increased, and the updated vector clock is added into a check-in data packet; When a node receives check-in data packets of other nodes, the received vector clock is combined with a local vector clock, and the combination operation is that each component is added with the corresponding component after the maximum value is taken; the Gossip protocol is adopted to transmit sign-in data and vector clock states among nodes, and the freshness of the data is judged through vector clock comparison, so that state synchronization among the nodes is realized.
  8. 8. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein S5 further comprises: Checking whether the reference number in the new data exists in the local sign-in records or not, judging the causal relationship between the records through the partial sequence relationship of the vector clock, and identifying causal sequence conflicts and concurrency conflicts; for causal sequence conflict, a time stamp priority strategy is adopted, and a sign-in record with an earlier time stamp is reserved; For concurrent conflict, an intelligent conflict resolution model is called for analysis, wherein the model adopts a gradient lifting decision tree algorithm, input characteristics comprise time intervals, geographical position distances, historical moving speeds and equipment historical reliability scores, and the rationality probability of the concurrent conflict is output; And reserving two records for the concurrent processing according to the model judgment result, and selecting the record with higher credibility score for the abnormal concurrent processing as effective check-in to obtain globally consistent check-in data.
  9. 9. The method for guaranteeing integrity of check-in data for meeting management according to claim 1, wherein S6 comprises: Privacy class division is carried out on the check-in data fields, and sensitive fields needing to be protected are identified; constructing a check-in validity verification circuit by adopting a zk-SNARKs algorithm, wherein the public input of the circuit comprises a conference number, a conference place number and a check-in time range, the private input comprises an identity token, a check-in time stamp and geographic coordinates, and the circuit realizes identity token constraint, time window constraint, geofence constraint and signature constraint; The proving generator executes a proving algorithm to generate a zero-knowledge proving file, and the verifier verifies the proving by using the verification key to obtain the data proving with privacy protection; And encrypting the check-in data by using a Paillier homomorphic encryption algorithm, directly adding the ciphertext by a statistics node, and decrypting by a key management center to obtain an aggregate statistics result, thereby obtaining a privacy protection verification result.
  10. 10. A system for guaranteeing integrity of check-in data for meeting management, for executing the steps in a method for guaranteeing integrity of check-in data for meeting management as claimed in any one of claims 1 to 9, comprising: The edge intelligent feature extraction module is used for acquiring preregistration information of participants, and a local feature library of each sign-in terminal is obtained by adopting an edge intelligent feature extraction and distributed cache synchronization method; the on-site quick matching module is used for acquiring a local feature library of each sign-in terminal, and obtaining a tamper-proof sign-in data packet by adopting an on-site quick matching and space-time anchoring method; The hierarchical consensus storage module is used for acquiring the tamper-proof sign-in data packet, and obtaining a blockchain record and final confirmation by adopting a hierarchical consensus and blockchain storage method; the intelligent verification arbitration module is used for acquiring the blockchain record and the final confirmation, and adopting a deep learning enhancement verification and intelligent arbitration method to obtain a data verification result and a final disposal decision; The conflict resolution module is used for acquiring a data verification result and a final disposal decision, and a vector clock and an intelligent conflict resolution method are adopted to obtain globally consistent check-in data; And the privacy protection verification module is used for acquiring globally consistent sign-in data and obtaining privacy protection verification results by adopting a zero knowledge proof and homomorphic encryption method.

Description

Sign-in data integrity guaranteeing method and system for conference management Technical Field The invention relates to the technical field of conference data security, in particular to a sign-in data integrity guarantee method and system for conference management. Background With the deep advancement of digital transformation, the meeting management system is widely applied to various conferences of government authorities, enterprises and public institutions, sign-in data is taken as an important basis for meeting attendance management and subsequent audit, and the integrity and the credibility of the meeting management system directly influence the standardization and the effectiveness of the meeting management. The traditional paper signing-in mode has the problems of low efficiency, easiness in counterfeiting and difficulty in tracing, and the existing electronic signing-in system has the defects in the aspect of data integrity assurance although the signing-in efficiency is improved. In the prior art, an electronic sign-in system mainly adopts a centralized database to store sign-in records, single-point fault risks and hidden danger of data tampering exist, and an effective tamper-proof mechanism and trusted storage capability are lacked. Meanwhile, under the distributed check-in scene of multiple meeting places, the data synchronization among the check-in terminals has a consistency problem, and when a network is delayed or partitioned, data conflict and loss are easy to generate. In addition, the detection of the abnormal sign-in behavior by the existing system mainly depends on simple rule judgment, and complex cheating modes and hidden abnormal behaviors are difficult to identify. Aiming at the problems, the prior art lacks a sign-in data integrity guarantee scheme capable of simultaneously meeting the requirements of data tamper resistance, distributed consistency, intelligent anomaly detection and privacy protection, cannot provide full life cycle data credibility guarantee for the conference management, and restricts the deep development of digital transformation of the conference management. Disclosure of Invention The invention provides a check-in data integrity guarantee method and system for meeting management, which solve the technical problems that check-in data has falsification risk, consistency is difficult to guarantee, abnormal detection capability is insufficient and privacy protection is lost in a multi-meeting-place distributed check-in scene in the related technology. The invention provides a sign-in data integrity guarantee method for meeting management, which comprises the following steps: s1, collecting preregistration information of participants, and obtaining a local feature library of each sign-in terminal by adopting an edge intelligent feature extraction and distributed cache synchronization method; S2, acquiring a local feature library of each sign-in terminal, and acquiring a tamper-proof sign-in data packet by adopting a field quick matching and space-time anchoring method; S3, obtaining a tamper-proof sign-in data packet, and obtaining a blockchain record and final confirmation by adopting a hierarchical consensus and blockchain storage method; S4, obtaining a blockchain record and final confirmation, and obtaining a data verification result and a final disposal decision by adopting a deep learning enhancement verification and intelligent arbitration method; S5, acquiring a data verification result and a final disposal decision, and obtaining globally consistent sign-in data by adopting a vector clock and an intelligent conflict resolution method; s6, obtaining globally consistent sign-in data, and obtaining privacy protection verification results by adopting a zero knowledge proof and homomorphic encryption method. In a preferred embodiment, the S1 includes: The method comprises the steps of obtaining preregistration information of participants, and obtaining a standard avatar file containing unique reference numbers by adopting data cleaning and format standardization processing; Acquiring a standard avatar archive, acquiring a face image through a mobile terminal application, and performing quality evaluation on the image to obtain a face image data set; Acquiring a face image dataset, processing by adopting a quality-aware self-adaptive feature extraction network, wherein the feature extraction network adopts a double-branch architecture, a main branch is a lightweight convolutional neural network responsible for feature extraction, an auxiliary branch is a quality evaluation network responsible for predicting image quality and generating self-adaptive weights, and carrying out fusion modulation on the quality weights and original feature vectors to obtain face feature reference vectors; and acquiring a face feature reference vector, constructing a feature retrieval index by adopting a local sensitive hash algorithm, and synchronously distributing the feature ret