Resource quality control system based on learner trust evaluation in open environment

Abstract

The invention discloses a resource quality control system based on learner trust evaluation in an open environment, which is used for calculating the credibility of a learner based on the historical contribution of the learner, granting resource control rights of different permissions to the learner and exciting the learner to continuously provide high-quality resources and valuable contribution. The system calculates historical contribution records of the learner through a trust factor algorithm under the guidance of a bidirectional interactive feedback model, and establishes a dynamic trust scoring mechanism of the learner. And according to the credibility score of the learner, the system automatically allocates corresponding resource management authority and resource sharing range for the learner, such as resource browsing range, uploading authority, editing authority, auditing authority, sharing authority and the like. According to the system, a data storage module, a resource packaging module, a learner trust evaluation module, a resource cooperative control module, an open access module and other modules are matched with one another to form a dynamic feedback circulation system, so that the resource quality of the whole system is continuously optimized.

Description

Technical Field

[0001] The present invention belongs to the field of Creative Commons licensing agreements, and in particular relates to a resource quality control system based on learner trust evaluation in an open environment. Background Art

[0002] From Web2.0 to Web3.0, the open education environment has gradually become an important network space for communication, sharing, and interactive information and services. Learners can freely edit and modify content in this space, but the reuse, modification, remixing, and contextualization of educational resources are not supervised or restricted, so the quality of the modified content cannot be guaranteed. A sound open environment resource management mechanism is urgently needed to simplify the resource and service selection process and build a learning environment of mutual trust and mutual benefit.

[0003] Currently, it is a common practice in the field of education to let domain experts judge the quality of learning resources; however, learners in an open environment will create a large number of learning resources, making the method of domain expert judgment expensive and time-consuming. In order to find a solution to this problem, more and more researchers have begun to focus on the exchange of teaching materials and the co-creation of open educational resources. Effective collaboration is considered to be a key factor in the long-term success of open educational resources in terms of quality and sustainability. Torres et al. (2017) explored the production of open educational resources in the collaborative process and found that social networks and collaborative processes promote each other, thereby improving the quality of open educational resources. Baldiris et al. (2017) proposed the CO-CREARIA model to support the co-creation of inclusive and accessible open educational resources and found that co-evaluation is an important strategy for building high-quality open educational resources. Goodsett et al. (2016) found in their study that collaboration between teachers and librarians is a successful model to support the creation of open educational resources on campus. Versuti et al. (2013) developed "lost narratives" to improve content production and learning. Torres et al. (2013)'s research shows that social networks are a means of sharing open educational resources in higher education to promote the production of open educational resources.

[0004] At the same time, in the computer field, existing research has used the concept of trusted computing to solve the increasingly serious Internet security problems. At present, the research and application of trusted computing in e-commerce platforms or virtual trading communities are relatively in-depth and mature. The typical TMSB, EigenTrust, PowerTrust and PeerTrust trust models directly, indirectly or comprehensively measure local or global trust values ​​based on the interaction process and transaction relationship of entities or nodes to ensure the reliability of services. However, this method relies too much on the subjective judgment of the scoring learners and is difficult to accurately measure the dynamic learner reputation.

[0005] The results of the above studies show that collaboration has the potential to promote high-quality open educational resources. However, most studies use social networks to complete smaller-scale production tasks, during which personal management and evaluation are required. When it comes to scaling up, monitoring the quality of collaborative editing becomes a problem, and more research based on different evaluation dimensions and targeting different segments is needed to enrich and expand the quality control of resources. Summary of the invention

[0006] In order to solve a series of problems in the open education environment, such as the uneven resource quality, insufficient enthusiasm of high-quality contributors, disconnection between learner ability assessment and authority allocation, low learning effect and resource utilization efficiency, etc., caused by the difficulty of resource quality control, insufficient learner contribution motivation, unreasonable authority allocation, inaccurate learner trust assessment, insufficient support for collaborative learning, and weak intellectual property protection, the present invention provides a resource quality control system based on learner trust evaluation in an open environment. Through an innovative trust evaluation mechanism and a dynamic authority control system, a self-optimizing and sustainable open learning ecosystem is constructed, which can not only improve the overall quality and utilization efficiency of open education resources, but also cultivate learners' sense of responsibility and innovation ability.

[0007] To achieve the above purpose, the technical solution adopted by the present invention is as follows: A resource quality control system based on learner trust evaluation in an open environment includes a data storage module, a resource encapsulation module, a learner trust evaluation module, a resource collaborative control module and an open access module; wherein, The data storage module is responsible for the storage, management and access of all system data; The resource encapsulation module is responsible for processing and organizing various types of learning resources and providing standardized resources for the resource collaborative control module; The learner trust evaluation module is used to construct a credibility evaluation index system for two-way interactive feedback, provide trust and authority information to the resource collaborative control module, and receive user behavior information during resource use fed back by the resource collaborative control module; The resource collaborative control module is used to manage learners' access to resources and operation rights, as well as to receive user requests from the open access module, send resource processing requests to the resource encapsulation module, and interact with the data storage module to read and write various types of data; The open access module is responsible for managing the interaction between the system and the external environment, and for managing user authentication and collaborative activities.

[0008] Furthermore, the credibility evaluation index system includes two parts: resource credibility evaluation and learner credibility evaluation, where: The resource trustworthiness is evaluated from two aspects: (1) By providing a resource trust evaluation function in the community, the resource trust can be directly evaluated subjectively in terms of content accuracy, objectivity, completeness, annotation standardization, and update timeliness; (2) Indirectly evaluate resource trust by recording and analyzing the interaction logs between learners and resources; Learners' trust is evaluated from two aspects: (1) The average trust level of resources created by learners; (2) Records of interactions between learners; Through the resource and learner trust update mechanism, the final trust of all resources and learners is output.

[0009] Furthermore, resource trust The calculation formula is as follows: (1) in, It represents the weight of the score, which is an increasing function with the number of direct trust evaluations as the independent variable; represents the direct trust obtained by direct evaluation, and IRT represents the indirect trust obtained by indirect evaluation; (2) Among them, n represents the number of learners participating in the direct trust evaluation of resources, and j is the index of the number of learners; represents the jth learner, represents the learner trust of the jth learner who participated in resource evaluation; |DIF| represents the number of evaluation indicators, and i represents the number index of evaluation indicators; represents the score of the jth learner on the ith evaluation indicator; itemscore represents the full score of each evaluation indicator item; represents the time impact factor, , t represents the time when the trust calculation occurs, Indicates The time when the comment was published. The time difference is calculated in months. (3) in, is the total number of interactions between learners and resources in the open education environment, is the index of the number of interactions, For the The learner's own trust in the interaction, For the The weight value of the interaction type to which the interaction belongs. is the regulating factor, is the penalty factor, : (4) Indirect trust in the calculation Normalization is performed and the indirect trust is mapped into discrete values ​​in the interval [0, 1] by adopting a segmented mapping method.

[0010] Furthermore, the learner trust evaluation problem is formalized into a probabilistic model, and the input of the probabilistic model is defined, including: learner set , a collection of resources created by learners The interaction record matrix I between the learner and the learner is a matrix of size The matrix of is the number of learners participating in the interaction. Each element in the interaction record matrix I Representing different learners The interaction strength between them defines the resource credibility factor function and the interaction factor function : (5) (6) in, Representing learners credibility, Representing learners The average credibility of all resources created; Indicates another learner credibility, Representing different learners The interaction strength between Representing learners The total number of positive interactions with other learners; Based on resource credibility factor function and the interaction factor function , construct the objective function : (7) in, and is the corresponding resource credibility factor function and the interaction factor function The weight parameter, is the normalization factor, are model parameters; Use gradient descent to learn model parameters Until the model converges, the marginal distribution probability of each learner is calculated by the Loopy BeliefPropagation algorithm , and obtain the final learner credibility evaluation result.

[0011] Furthermore, the trust updating mechanism of the resource and learner is as follows: The trust level is updated by triggers. By setting the updated conditional rules, once any conditional rule is met, the trust level update program is triggered to recalculate the trust level. The conditional rules are set as follows: First condition rule 1: When the number of interactions received by the resource exceeds the threshold When , the trust of the resource is updated; Second Condition Rule 2: When the resource trust changes When , the trust of learners who have interacted with the resource is not updated; Third Condition Rule 3: When the resource trust changes When , the trust of learners who have interacted with the resource is updated, but no cross calculation is performed; Fourth Condition Rule 4: When the resource trust changes When , the trust of learners who have interacted with the resource is updated, and cross calculation is performed at the same time; Fifth Condition Rule 5: When the number of interactions received by the learner exceeds the threshold , then update the learner's trust; Sixth Condition Rule 6: When the learner's trust changes When learning, there is no need to update the trust of resources that have interacted with learners; Seventh Condition Rule 7: When the learner's trust changes When , the trust of resources that have interacted with the learner is updated, but no cross calculation is performed; Rule 8: When learners’ trust changes When the learner interacts with the resource, the trust of the resource is updated and cross-calculated at the same time; in, Indicates the minimum number of resource interactions. represents the minimum number of interactions for learners, Indicates the difference before and after a single resource is updated. represents the difference between a single learner’s trust before and after update, Indicates the minimum value of resource trust change, Indicates the maximum value of resource trust change, represents the minimum change of learner's trust, Indicates the maximum change in learner trust.

[0012] The beneficial effects of the present invention are: (1) The present invention automatically allocates different levels of resource management permissions based on the learners' trust scores, which not only ensures the quality of resources but also provides more participation opportunities for high-trust learners.

[0013] (2) The present invention changes the permission granting mechanism by calculating the learner's trust, effectively motivating learners to continuously provide high-quality resources and valuable contributions, creating a self-optimizing learning ecosystem and continuously improving the overall quality of resources in an open environment.

[0014] (3) The present invention defines the factors that affect learners' trust as the learners' authority and professional knowledge ability through the three properties of the trusted party and the fine processing possibility model. On this theoretical basis, a credibility evaluation index system is constructed, which dynamically and explicitly links the credibility evaluation results with the content shared by the knowledge contributors to assist learners in judging the quality of resources.

[0015] (4) The present invention can fully and visually display the learner's trust score, resource quality distribution, authority allocation, and the operating status of the entire learning ecosystem, making it easier to monitor the quality of resources in an open education environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a block diagram of a resource quality control system based on learner trust evaluation in an open environment of the present invention; Figure 2 This is the principle diagram of the evaluation index system of the two-way interactive feedback model in the learner trust evaluation module; Figure 3 Encapsulate the core content of the resource encapsulation module; Figure 4 It is the core schematic diagram of the resource collaborative control model; Figure 5 It is a schematic diagram of learner collaborative version record details in the resource collaborative control module; Figure 6 This is a schematic diagram of the historical revision record of learner collaborative versions in the resource collaborative control module. DETAILED DESCRIPTION

[0017] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] like Figure 1 As shown, a resource quality control system based on learner trust evaluation in an open environment of the present invention includes: a data storage module, a learner trust evaluation module, a resource encapsulation module, a resource collaborative control module and an open access module. Among them: The data storage module is responsible for the storage, management, and access of all system data. It adopts a distributed database architecture to ensure efficient data processing capabilities and good scalability. The main functions include: (1) learner information storage, including basic information, historical contribution records, trust scores, etc.; (2) resource data management, storing various types of learning resources and their metadata; (3) behavior data recording, tracking learners' operations and interactions; (4) evaluation data storage, including resource evaluation and learner evaluation; (5) permission configuration information management; (6) system log data recording. This module also implements security functions such as data encryption, backup and recovery, and data consistency maintenance, and provides a flexible API interface to support the data requirements of other modules.

[0019] The learner trust evaluation module is the core module of the system, which is responsible for dynamically evaluating and updating the learner's trust. The key is to comprehensively consider the learner's authority and professional knowledge ability based on the three characteristics of the trustee and the theory of the elaboration possibility model, and construct an evaluation index system of a two-way interactive feedback model that can evaluate the learner trust and resource trust in the open knowledge community. The index system includes two core components: resource trust and learner trust, which influence each other. Resource trust can be evaluated from two aspects: one is the explicit trust evaluation (direct evaluation) of resources, which allows learners to make subjective evaluations from the aspects of content accuracy, objectivity, completeness, annotation standardization and update timeliness by providing resource trust evaluation functions in the community; the other is to make implicit evaluations (indirect evaluations) of resource trust by recording and analyzing the interaction logs between learners and resources, including the frequency of learners' collaborative editing, subscription, collection, citation, browsing and recommendation. Of course, different types of interactive operations in different open environments are also supported. The factors that affect learners' trust in an open education environment also come from two aspects. On the one hand, it is the average trust of the resources created by learners, and on the other hand, it is the interaction records between learners. Different interaction behaviors represent implicit evaluations between learners. Specifically, it can be expressed as the credibility of creating resources, the number of times invited to collaborate or cancel collaboration, the number of times added as a friend or canceled as a friend, and the number of times revisions are accepted or rejected. The specific calculation method is as follows: Resource Trust The calculation formula is as follows: (1) in, It represents the weight of the score, which is an increasing function with the number of direct trust evaluations as the independent variable; represents the direct trust obtained by direct evaluation, and IRT represents the indirect trust obtained by indirect evaluation; (2) Among them, n represents the number of learners participating in the direct trust evaluation of resources, and j is the index of the number of learners; represents the jth learner, represents the learner trust of the jth learner who participates in resource evaluation; |DIF| represents the number of evaluation indicators, and i represents the number index of evaluation indicators; represents the score of the j-th learner on the i-th evaluation indicator; itemscore represents the full score of each evaluation indicator item; represents the time impact factor, , t represents the time when the trust calculation occurs, Indicates The time when the comment was published. The time difference is calculated in months. (3) in, is the total number of interactions between learners and resources in the open education environment, is the index of the number of interactions, For the The learner's own trust in the interaction, For the The weight value of the interaction type to which the interaction belongs. is the regulating factor, is the penalty factor, : (4) Indirect trust in the calculation Normalization is performed and the indirect trust is mapped into discrete values ​​in the interval [0, 1] by adopting a segmented mapping method.

[0020] The learner trust evaluation problem is formalized into a probabilistic model, and the input of the probabilistic model is defined, including: learner set , a collection of resources created by learners The interaction record matrix I between the learner and the learner is a matrix of size The matrix of is the number of learners participating in the interaction. Each element in the interaction record matrix I Representing different learners The interaction strength between them defines the resource credibility factor function and the interaction factor function : (5) (6) in, Representing learners credibility, Representing learners The average credibility of all resources created; Indicates another learner credibility, Representing different learners The interaction strength between Representing learners The total number of positive interactions with other learners; Based on resource credibility factor function and the interaction factor function , construct the objective function : (7) in, and is the corresponding resource credibility factor function and the interaction factor function The weight parameter, is the normalization factor, are model parameters; Use gradient descent to learn model parameters Until the model converges, the marginal distribution probability of each learner is calculated by the Loopy BeliefPropagation algorithm , and obtain the final learner credibility evaluation result.

[0021] The trust update mechanism of the resource and learner is as follows: The trust level is updated by triggers. By setting the updated conditional rules, once any conditional rule is met, the trust level update program is triggered to recalculate the trust level. The conditional rules are set as follows: First condition rule 1: When the number of interactions received by the resource exceeds the threshold When , the trust of the resource is updated; Second Condition Rule 2: When the resource trust changes When , the trust of learners who have interacted with the resource is not updated; Third Condition Rule 3: When the resource trust changes When , the trust of learners who have interacted with the resource is updated, but no cross calculation is performed; Fourth Condition Rule 4: When the resource trust changes When , the trust of learners who have interacted with the resource is updated, and cross calculation is performed at the same time; Fifth Condition Rule 5: When the number of interactions received by the learner exceeds the threshold , then update the learner's trust; Sixth Condition Rule 6: When the learner's trust changes When the learner interacts with a resource, the trust of the resource does not need to be updated; Seventh Condition Rule 7: When the learner's trust changes When , the trust of resources that have interacted with the learner is updated, but no cross calculation is performed; Rule 8: When learners’ trust changes When the learner interacts with the resource, the trust of the resource is updated and cross-calculated at the same time; in, Indicates the minimum number of resource interactions. represents the minimum number of interactions for learners, Indicates the difference before and after a single resource is updated. represents the difference between a single learner’s trust before and after update, Indicates the minimum value of resource trust change, Indicates the maximum value of resource trust change, represents the minimum change of learner's trust, Indicates the maximum change in learner trust.

[0022] The resource encapsulation module is responsible for processing and organizing various types of learning resources to ensure that they meet the system standards. The main functions include: (1) supporting resource uploads in multiple formats, such as text, images, audio, video, etc.; (2) automatically extracting and generating resource metadata, including title, author, creation time, keywords, etc.; (3) realizing resource format conversion to ensure cross-platform compatibility; (4) integrating version control systems to track resource modification history; (5) realizing intelligent tagging systems to automatically analyze content and generate tags; (6) providing resource preview functions; (7) integrating digital copyright management tools to protect intellectual property rights; (8) realizing preliminary assessment of resource quality and filtering out obviously unqualified content; (9) supporting resource classification and organization for easy retrieval and management.

[0023] The resource collaborative control module manages learners' access and operation rights to resources and implements collaborative editing functions. The main functions include: (1) Learner role analysis. Through the trust value calculated by the learner trust evaluation module, different permissions are assigned to learners, making them collaborators, collaborators, ordinary users, commentators and other roles of resources, so that they can complete resource updates together; (2) Behavior restrictions. Based on the learner role, learners' behavior is restricted. Collaborators have all permissions such as browsing, downloading, editing, and reviewing. Collaborators have browsing, editing, and reviewing permissions. Ordinary users have browsing, downloading, and commentators have browsing permissions. Behavior restrictions can be accurate to a single resource or a collection of resources; (3) Resource sharing. It provides a collaborative editing environment that supports multiple learners to edit at the same time, share resources, and synchronize changes in real time; (4) Realize collaborative editing breakthroughs. solution mechanism; (5) provide resource sharing scope control, allowing resource owners to customize the visible scope; (6) provide a concurrency control module to ensure that when learners access and modify resources at the same time, the system can ensure data consistency and integrity through mechanisms such as lock mechanisms, conflict detection and transaction management; (7) implement an operation log system to record all learners' operations on resources; (8) provide a resource review workflow and support multi-level review; (9) implement a resource comment and feedback system to promote learner communication; (10) provide resource trust evaluation and learner trust evaluation functions, which can calculate trust in real time based on learners' behavioral operations, and then update learners' permissions, assign different roles to learners, and complete a content micro-circulation.

[0024] The open access module is responsible for managing the interaction between the system and the external environment to ensure openness and security. The main functions include: (1) providing learner authentication and authorization services, supporting multiple login methods; (2) implementing API gateways and managing external access interfaces; (3) providing resource open sharing services, supporting different levels of open licenses; (4) implementing cross-platform data exchange and supporting integration with other learning systems; (5) providing data export functions, allowing learners to obtain personal data; (6) implementing security protection measures, such as anti-DDoS attacks, data encryption transmission, etc.; (7) providing learner privacy settings, allowing learners to control the scope of personal information sharing; (8) implementing system usage statistics and analysis functions, and collecting learner feedback; (9) providing developer toolkits to support third-party application development; (10) managing system announcements and learner notification services.

[0025] like Figure 2As shown in the figure, the principle diagram of the evaluation index system of the two-way interactive feedback model in the learner trust evaluation module is given. The evaluation index system focuses on two core factors: resource trust and learner trust, and adopts direct trust evaluation and indirect trust evaluation to comprehensively consider. The direct evaluation of resources allows learners to make subjective evaluations from the aspects of content accuracy, objectivity, completeness, annotation standardization and update timeliness by providing resource trust evaluation functions in the community. Indirect evaluation records and analyzes the interaction logs between learners and resources, including the frequency of learners' collaborative editing, subscription, collection, citation, browsing and recommendation behaviors. At the same time, the learner's trust is calculated from the average trust of the resources created by the learner and the interaction records between learners, such as the trustworthiness of creating resources, the number of times invited to collaborate or cancel collaboration, the number of times added as friends or canceled as friends, and the number of times revisions are accepted or rejected. Finally, the two-way interactive feedback between the two is comprehensively considered, and the comprehensive trust of the learner is obtained to open different permissions for new resources, and at the same time, different levels are given to learners to highlight their contributions and encourage learners to further contribute higher-quality content.

[0026] like Figure 3 As shown, in order to maintain content attributes during the collaboration process, the resource encapsulation module sets the attributes and collaborative content in separate blocks, specifically including an attribute block and a content block, wherein: (1) The attribute block is constructed and maintained by the holder. It contains information such as content ID, holder ID, and the holder's signature on the attribute. When the holder wants to modify the attribute information, it must be verified by the system.

[0027] (2) The content block is initialized by the holder. All collaborators can modify the content block. This block contains version information, modified content, collaborator ID, and their signature on the content. Collaborators can modify the content and update the content version, collaborator ID, and content signature according to their own permissions.

[0028] like Figure 4 As shown in the figure, it is a schematic diagram of the resource collaborative control module, which is explained by taking holders, collaborators and ordinary users as examples. When ordinary users reach the threshold of becoming collaborators through trust evaluation, the system will automatically grant the learner the role of collaborator, and the learner can exercise the power of the collaborator. At the same time, the collaborator can further provide his own power level by contributing to the quality of resources and obtain more permissions. However, the ultimate control over the resources is owned by the holder, who can evaluate the collaborator and has all the permissions on the resources.

[0029] like Figure 5As shown in the figure, a detailed diagram of learner collaborative version record in the resource collaborative control module is given. When learners contribute high-quality content, after the trust is calculated, different resource control permissions are granted to learners based on the trust value. Learners can exercise different rights such as browsing, downloading, editing, and reviewing. The holder compares different submitted content and selects high-quality content for the new version. Then the collaborators can continue to improve the content based on the new version, and finally form different resource versions.

[0030] like Figure 6 As shown, the revision history of learner collaborative version record details in the resource collaborative control module is displayed, including the browsed version, the version to which it belongs, the reason for the change, the contributor, the reviewer, and the review result. This can clearly track and evaluate the evolution of resource content, facilitating resource management and quality control in an open education environment.

[0031] The steps that the system finally implements on the learner side are as follows: Step (1) Registration and initialization phase: The learner first registers an account on the platform and provides basic information such as educational background, professional field, etc. The system assigns an initial trust score and basic permissions to the new learner. The learner can browse public resources and participate in simple interactions such as commenting and liking.

[0032] Step (2) Participation and contribution stage: Learners begin to actively participate in platform activities, such as providing resources, commenting on other people's works, participating in discussions, etc. The system records these behaviors and continuously updates the learner's trust score through the learner trust assessment module. As contributions increase and quality improves, learners' trust gradually increases.

[0033] Step (3) Permission enhancement phase: When the learner's trust reaches a certain threshold, the system automatically enhances their permissions. For example, they are allowed to upload more types of resources, participate in the editing of resources, etc. The learner will receive a system notification to understand the newly acquired permissions and responsibilities.

[0034] Step (4) Collaborative editing participation stage: 1) Invitation to participate: The system pushes resource invitations suitable for collaborative editing to learners based on their trust and professional fields. Learners can choose to accept or reject. 2) Content editing: After accepting the invitation, learners can view the content initialized by the resource holder and start editing and modifying it. The system will record each modification and form different versions. 3) Interaction and feedback: During the editing process, learners can discuss with other collaborators, and can also view the comments and feedback of ordinary learners to continuously improve the content.

[0035] Step (5) Results display and evaluation phase: After collaborative editing is completed, learners can view their contributions. The system will update the learner's trust score based on the quality and quantity of their contributions, as well as the evaluation of other learners. Excellent contributions may lead to a significant increase in trust.

[0036] Step (6) Authority adjustment and role conversion stage: Based on the performance during the collaboration process, the learner's authority may be further upgraded. For example, from a collaborator to a reviewer of certain resources. At the same time, ordinary learners with high trust may receive invitations to become new collaborators.

[0037] Step (7) Continuous learning and contribution stage: Learners can view their trust score history and contribution records to understand their personal growth trajectory. At the same time, the system will label learners with different level avatars based on the accumulated trust points to highlight the learner's identity. The system will also continuously recommend suitable projects and resources based on the learner's expertise and interests to encourage learners to continue learning and contributing.

[0038] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A resource quality control system based on learner trust evaluation in an open environment, characterized in that: It includes data storage module, resource encapsulation module, learner trust assessment module, resource collaborative control module and open access module; among them, The data storage module is responsible for the storage, management and access of all system data; The resource encapsulation module is responsible for processing and organizing various types of learning resources and providing standardized resources for the resource collaborative control module; The learner trust evaluation module is used to calculate the learner trust, provide trust and authority information to the resource collaborative control module, and receive the user's behavior information during resource use fed back by the resource collaborative control module; The resource collaborative control module is used to manage learners' access to resources and operation rights, as well as to receive user requests from the open access module, send resource processing requests to the resource encapsulation module, and interact with the data storage module to read and write various types of data; The open access module is responsible for managing the interaction between the system and the external environment, and for managing user authentication and collaborative activities.

2. According to claim 1, a resource quality control system based on learner trust evaluation in an open environment is characterized in that: The credibility evaluation index system includes two parts: resource trust evaluation and learner trust evaluation. The resource trustworthiness is evaluated from two aspects: (1) By providing a resource trust evaluation function in the community, the resource trust can be directly evaluated subjectively in terms of content accuracy, objectivity, completeness, annotation standardization, and update timeliness; (2) Indirectly evaluate resource trust by recording and analyzing the interaction logs between learners and resources; Learners' trust is evaluated from two aspects: (1) The average trust level of resources created by learners; (2) Records of interactions between learners; Through the resource and learner trust update mechanism, the final trust of all resources and learners is output.

3. A resource quality control system based on learner trust evaluation in an open environment according to claim 2, characterized in that: Resource Trust The calculation formula is as follows: (1) in, It represents the weight of the score, which is an increasing function with the number of direct trust evaluations as the independent variable; DRT represents the direct trust obtained by direct evaluation, and IRT represents the indirect trust obtained by indirect evaluation; (2) Among them, n represents the number of learners participating in the direct trust evaluation of resources, and j is the index of the number of learners; represents the jth learner, represents the learner trust of the jth learner who participates in resource evaluation; |DIF| represents the number of evaluation indicators, and i represents the number index of evaluation indicators; represents the score of the j-th learner on the i-th evaluation indicator; itemscore represents the full score of each evaluation indicator item; represents the time impact factor, , t represents the time when the trust calculation occurs, Indicates The time when the comment was published. The time difference is calculated in months. (3) in, is the total number of interactions between learners and resources in the open education environment, is the index of the number of interactions, For the The learner's own trust in the interaction, For the The weight value of the interaction type to which the interaction belongs. is the regulating factor, is the penalty factor, : (4) Indirect trust in the calculation Normalization is performed and the indirect trust is mapped into discrete values ​​in the interval [0, 1] by adopting a segmented mapping method.

4. A resource quality control system based on learner trust evaluation in an open environment according to claim 2, characterized in that: The learner trust evaluation problem is formalized into a probabilistic model, and the input of the probabilistic model is defined, including: learner set , a collection of resources created by learners The interaction record matrix I between the learner and the learner is a matrix of size The matrix of is the number of learners participating in the interaction. Each element in the interaction record matrix I Representing different learners The interaction strength between them defines the resource credibility factor function and the interaction factor function : (5) (6) in, Representing learners credibility, Representing learners The average credibility of all resources created; Indicates another learner credibility, Representing different learners The interaction strength between Representing learners The total number of positive interactions with other learners; Based on resource credibility factor function and the interaction factor function , construct the objective function : (7) in, and is the corresponding resource credibility factor function and the interaction factor function The weight parameter, is the normalization factor, are model parameters; Use gradient descent to learn model parameters Until the model converges, the marginal distribution probability of each learner is calculated through the Loopy Belief Propagation algorithm , and obtain the final learner credibility evaluation result.

5. According to claim 2, a resource quality control system based on learner trust evaluation in an open environment is characterized in that: The trust update mechanism of the resource and learner is as follows: The trust level is updated by triggers. By setting the updated conditional rules, once any conditional rule is met, the trust level update program is triggered to recalculate the trust level. The conditional rules are set as follows: First condition rule 1: When the number of interactions received by the resource exceeds the threshold When , the trust of the resource is updated; Second Condition Rule 2: When the resource trust changes When , the trust of learners who have interacted with the resource is not updated; Third Condition Rule 3: When the resource trust changes When , the trust of learners who have interacted with the resource is updated, but no cross calculation is performed; Fourth Condition Rule 4: When the resource trust changes When , the trust of learners who have interacted with the resource is updated, and cross calculation is performed at the same time; Fifth Condition Rule 5: When the number of interactions received by the learner exceeds the threshold , then update the learner's trust; Sixth Condition Rule 6: When the learner's trust changes When learning, there is no need to update the trust of resources that have interacted with learners; Seventh Condition Rule 7: When the learner's trust changes When , the trust of resources that have interacted with the learner is updated, but no cross calculation is performed; Rule 8: When learners’ trust changes When the learner interacts with the resource, the trust of the resource is updated and cross-calculated at the same time; in, Indicates the minimum number of resource interactions. represents the minimum number of interactions for learners, Indicates the difference before and after a single resource is updated. represents the difference between a single learner’s trust before and after update, Indicates the minimum value of resource trust change, Indicates the maximum value of resource trust change, represents the minimum change of learner's trust, Indicates the maximum change in learner trust.

Images