CN-121352802-B - Campus secondhand commodity transaction credible traceability evaluation method and system

CN121352802BCN 121352802 BCN121352802 BCN 121352802BCN-121352802-B

Abstract

The invention provides a campus secondhand commodity transaction credibility traceability assessment method and system, and relates to the technical field of electric digital data processing, wherein the accuracy and the foresight of credibility assessment are fundamentally improved by constructing a closed-loop linkage control system from multi-dimensional feature depth quantization to individual-group relevance comparison and then to self-adaptive strategy dynamic decision; the method has the beneficial effects that 1) the initial trust evaluation capability of a new user without historical data is enhanced, 2) the identification accuracy of the hidden fraudulent activity with the characteristic of 'mode abnormality' which is shielded by using a high credit identity is improved, 3) the dynamic self-adaption of a risk intervention strategy is realized, the management and control force can be intelligently adjusted according to the overall risk situation of the platform, and the better balance between the safety of the platform and the user experience optimization is achieved.

Inventors

Zhang Tuoliang
ZHANG WENLIANG
YANG JINQIU
ZHANG ZHILIANG

Assignees

北京漂洋过海科技有限责任公司

Dates

Publication Date: 20260508
Application Date: 20250929

Claims (10)

1. The credible traceability evaluation method for campus secondhand commodity transaction is characterized by comprising the following specific steps: S1, receiving and verifying trusted entity credentials associated with a user identity and based on zero knowledge proof, wherein the trusted entity credentials are used for proving that the user belongs to a preset authoritative entity; Based on the distributed timestamp record of the trusted entity certificate, calculating the certificate maturity representing the history length and the certificate continuity representing the activity frequency stability of the trusted entity certificate; Nonlinear fusion is carried out on the credential validity state, the credential maturity and the credential continuity to obtain comprehensive credential credibility; s2, acquiring multi-mode commodity information to be evaluated, which is issued by the user; S3, calculating individual information consistency indexes for quantifying the self-consistency degree of the internal logic of the multi-mode commodity information based on the data relevance in the multi-mode commodity information; S4, determining a group behavior baseline associated with the authoritative entity based on the trusted entity certificate, and calculating a certificate content deviation degree representing the deviation degree of user behaviors and the group behaviors of the user behaviors by combining the individual information consistency indexes, wherein the calculating of the certificate content deviation degree specifically comprises the following steps: Constructing a current user behavior feature vector containing individual information consistency indexes and comprehensive credential credibility; determining the group behavior base line as a group behavior feature mean value vector and a group behavior feature covariance matrix corresponding to a group to which a user belongs; Calculating the distance between the current user behavior feature vector and the group behavior base line by applying a Markov distance algorithm, and taking the distance as the credential content deviation degree; and S5, generating the multi-mode commodity information and the credibility evaluation result of the user based on the voucher content deviation.
2. The method for evaluating credibility traceability of campus second-hand commodity transaction according to claim 1, wherein verifying the credible entity certificate specifically comprises the steps of executing a zero-knowledge proof verification algorithm on the credible entity certificate by utilizing a preset public key associated with the authoritative entity, and confirming the certificate validity state of the credible entity certificate on the premise of not acquiring specific identity information of a user.
3. The method for evaluating credibility of campus second-hand commodity transaction according to claim 2, wherein the calculating the individual information consistency index specifically comprises the following steps: Calculating the graphic concept alignment degree of the semantic matching degree of the representation image and the text; calculating a price-value deviation coefficient representing the deviation degree of commodity price and predicted fair value; And calculating content generation process entropy of behavior abnormality in the quantization information creation process.
4. The method for evaluating credible traceability of second-hand commodity transaction in campus according to claim 3, wherein said content generation process entropy comprises jointly quantifying source attribute of said image information and editing process stability of said text information based on behavior log when recording user created multi-mode commodity information; And taking the image-text concept alignment degree, the price-value deviation coefficient and the content generation process entropy as inputs, and carrying out nonlinear fusion based on a fuzzy logic reasoning system to obtain the individual information consistency index.
5. The method for evaluating credibility of second-hand commodity transaction traceability in campus of claim 4, wherein the fuzzy logic reasoning-based system comprises a fuzzy rule base, and rules for generating the fuzzy rule base are automatically optimized by performing machine learning training on fraud marked by historic and normal sample data; and obtaining the entropy of the content generation process by applying shannon entropy calculation logic to the editing and modifying behavior log in the process of analyzing the metadata of the uploaded image and inputting the text of the user.
6. The method for evaluating the credibility of the second-hand commodity transaction in campus according to claim 5, wherein when the credibility of the comprehensive certificate is higher, the characterization system judges that the comprehensive performance of the authenticity, the historical deposit and the activity stability of the user certificate is better, and the priori credibility of the information source is higher; when the individual information consistency index is larger, the characterization system judges that the internal contradiction and abnormality existing in the commodity information issued by the user in each dimension are more serious, the lower the credibility of the content is, and the higher the fraud risk is.
7. The method for traceability evaluation of second-hand commodity transaction in campus of claim 6, wherein said group behavior baseline comprises statistical distribution parameters of historical information consistency indicators associated with said authoritative entity and commodity category.
8. The method for traceability evaluation of second-hand commodity transaction in campus according to claim 7, wherein the step of generating the credibility evaluation result comprises the steps of taking the deviation degree of the content of the certificate as a key input characteristic, updating the dynamic trust score of a user, and triggering a preset self-adaptive inspection strategy based on comprehensive evaluation of the dynamic trust score and the deviation degree of the content of the certificate; Defining an adaptive inspection strategy triggering preset, which specifically comprises the following steps: And outputting an optimal examination strategy with the maximum expected utility value by the reinforcement learning model according to the state input and executing the optimal examination strategy.
9. The campus second-hand commodity transaction credible traceability assessment method according to claim 8, wherein the method comprises the following steps: The state input of the reinforcement learning model further comprises a platform risk water level representing the current overall security situation of the platform; after executing the optimal examination strategy, calculating a reward value according to feedback of a subsequent result executed by the strategy; updating the reinforcement learning model with the reward value to optimize subsequent policy selection; When the deviation degree of the voucher content is smaller, the behavior mode of the current user is characterized to be more consistent with the historical average behavior mode of the group to which the current user belongs, the behavior is more normal, and the fraud risk is lower; the optimal censoring strategy is a specific action in the discrete set of actions output by the reinforcement learning model.
10. The campus secondhand commodity transaction credibility traceability evaluation system is characterized by being used for executing the campus secondhand commodity transaction credibility evaluation method according to any one of claims 1-9, and comprising the following steps: The credential verification and data acquisition module is used for receiving and verifying a trusted entity credential which is associated with the identity of the user and is based on zero knowledge proof, wherein the trusted entity credential is used for proving that the user belongs to a preset authoritative entity; Acquiring multi-mode commodity information to be evaluated, which is issued by the user; The multidimensional feature depth analysis module is used for calculating and quantifying individual information consistency indexes of the internal logic self-consistency degree of the multi-mode commodity information based on the data relevance in the multi-mode commodity information; The individual-group deviation degree quantifying module is used for determining a group behavior baseline associated with the authoritative entity based on the trusted entity certificate, and calculating the deviation degree of the certificate content representing the deviation degree of the user behavior and the group behavior to which the user behavior belongs by combining the individual information consistency index; and the reinforcement learning self-adaptive decision module is used for generating the multi-mode commodity information and the credibility evaluation result of the user based on the credential content deviation.

Description

Campus secondhand commodity transaction credible traceability evaluation method and system Technical Field The invention relates to the technical field of electric digital data processing, in particular to a campus secondhand commodity transaction credible traceability evaluation method and system. Background With the deepening of digital economy, a User Generated Content (UGC) based point-to-point (Peer-to-Peer) information interaction platform is increasingly popular, and a plurality of fields such as social contact, electronic commerce and local service are covered. In such platforms, the credibility of the information is a cornerstone that maintains ecological health and user liveness. Therefore, a computing method capable of evaluating the reliability of massive, heterogeneous and dynamically generated user information accurately, efficiently and automatically is developed, and the computing method becomes a frontier technical trend of common attention in the fields of data science and network space management, and has important significance in improving the platform operation efficiency and guaranteeing the user rights. Under the specific application scene of campus second-hand commodity transaction, the prior art mainly faces the following challenges when realizing high-efficiency and accurate credibility evaluation: 1. Traditional credibility assessment methods are highly dependent on historical behavior data of users, such as transaction records, historical evaluations and the like. This model makes it difficult to develop an effective initial trust judgment for newly registered users or sporadic traffic users lacking historical data, while the evaluation of such dependency historical aggregate data is inherently lagged, and the risk early warning capability is to be improved for elaborate fraud suddenly implemented with long-term accumulated high credit identities. 2. The prior art deals with review of published content and credit assessment of publishers as two independent flows. For example, content review may employ keyword filtering, image similarity ratio peering, and source evaluation may analyze its historical transaction scores. This split analysis framework makes it difficult to capture deep associative anomalies between "sources" and "content". For example, a deep user who should release a high-quality digital product (a trusted source) suddenly releases a commodity (abnormal content) which is completely irrelevant to the professional field and has abnormal price, and the risk implied by the mismatch of the source and the content is easily ignored in the splitting analysis mode. 3. Existing risk interventions are mostly based on fixed rule threshold systems. For example, chinese patent CN120316182a discloses a data analysis method, which, although implementing integration and analysis of multi-source data to support operational decision, the risk coping logic (e.g. fault detection, dynamic pricing) is essentially still a response based on preset rules. Such rigid systems are not sufficiently adaptable to face complex and varying means of fraud and dynamically varying overall risk situations for the platform. The system can not automatically adjust the tightness of the intervention strategy according to the global risk water level, and dynamic balance is difficult to achieve between ensuring safety and optimizing user experience. The above information disclosed in the above background section is only for enhancement of understanding of the background of the disclosure and therefore it may include information that does not form the prior art that is already known to those of ordinary skill in the art. Disclosure of Invention The invention aims to provide a campus second-hand commodity transaction credible traceability assessment method and system, which are used for solving the problems in the background technology. In order to achieve the above purpose, the present invention provides the following technical solutions: a campus secondhand commodity transaction credibility traceability evaluation method specifically comprises the following steps: S1, receiving and verifying trusted entity credentials associated with a user identity and based on zero knowledge proof, wherein the trusted entity credentials are used for proving that the user belongs to a preset authoritative entity; s2, acquiring multi-mode commodity information to be evaluated, which is issued by the user; S3, calculating individual information consistency indexes for quantifying the self-consistency degree of the internal logic of the multi-mode commodity information based on the data relevance in the multi-mode commodity information; S4, determining a group behavior baseline associated with the authoritative entity based on the trusted entity certificate, and calculating a certificate content deviation degree representing the deviation degree of user behaviors and the group behaviors to which the user behaviors belong by co