A server-based, auditable data processing system and method for course recognition, exemption, and equivalency processes in higher education.

Abstract

The invention relates to a server-based system and method that processes data from source academic documents (10) and target curriculum resources in a computer environment for the recognition, exemption, and equivalency evaluation of courses from different institutions, programs, or curriculum resources in higher education. The system can be implemented with a user client device (200), application server (210), document processing server (220), artificial intelligence evaluation server or service connection (230), database server (240), file storage unit (250), network interface (260), reporting and output unit (270), and data reduction, masking, and model orientation layer (280). Course code, course name, grade, national credit, ECTS, and course content information are extracted from the source documents, and this information is collected in unique source course records, with a source field tracking system that tracks which document each field came from.Target curriculum data is validated in terms of the target program's language of instruction and curriculum source. For each source course, a limited list of candidate target courses is determined instead of the entire target program, and deterministic rule checks are performed on this list using evidence-based matching evaluation. Transcripts and grade documents that may contain personal data are evaluated in the in-house model environment, while publicly available course content can be evaluated in a local or external model service according to the institution's policy. Decisions finalized by humans are transferred to the final exemption register (140); the exemption and adaptation report and the audit trail are generated from this finalized decision status without rerunning the artificial intelligence model.

Description

1 TARIFF Server-based systems for course recognition, exemption, and equivalency processes in higher education. auditable data processing system and method Technical Area The invention involves combining courses from different institutional, program, or curriculum sources in higher education. academic resources for recognition, exemption, and adaptation assessment within the target program 10 that process data from documents and target curriculum resources in a computer environment It involves a server-based system and method. State of the Art In current practices, course recognition, exemption, and equivalency processes are mostly completed within 15 days of the student's graduation. through the transcript, course content, course syllabus and application forms provided, manually or It is carried out in a semi-manual manner. The relevant committee, department head, student affairs unit, or The advisor faculty member compares the source courses with the courses in the target program, assigning credits. ECTS takes into account factors such as content compatibility, passing grades, and the pool of elective courses. In this process, the format and content of documents vary from institution to institution. Transcripts, course materials... Programs and catalog information are available in various PDF formats, scanned documents, and web pages. Alternatively, this information can be found in student information systems. This includes course code, course name, credits, ECTS, grade, and content. Because the information is not presented in the same order or with the same clarity in every document, the source course information... It becomes difficult to convert it into a regular and auditable record. 25 Furthermore, similarity in course name or course code alone is not sufficient for a reliable equivalence decision. It is not. Course content, learning outcomes, national credit, ECTS, and grade point average are all considered together. This should be evaluated. However, in the current manual process, it is not possible to determine which target courses are candidates. it was examined, which courses were rejected and why, and to which document or content section the decision was made. The basis on which it is based is often not clearly visible. Ensuring that the target curriculum is sourced correctly is another issue. Different versions of the same program... Catalogs prepared in the language of instruction or curriculum versions from different years may be available. The course 35 is based on an incorrect language catalog, outdated curriculum, or incomplete course content. The comparison may be flawed. 2 In some cases, multiple resource courses may correspond to a single target course, or resources Courses can be counted towards elective course pools. In these cases, the total credit / ECTS sufficiency is determined. weighted grade calculation, no reuse of the same course, elective pool capacity, total exemptions ECTS and adaptation semester / year calculations should be carried out consistently. The current manual or semi-annual calculation method should be followed. In manual methods, performing these calculations in different tables or by different people can lead to inconsistencies. This can lead to consequences. Consequently, the fundamental technical problem in the current technique is the use of different document and curriculum sources. The incoming course information is standardized, its source area is traceable, and it is machine-processable. 10 The reasoning behind the decision that it could not be transformed into a structure was the candidate courses used, human The problem is that the intervention and the final report cannot be verified against the same consistent data situation. Technical investigations revealed 15 with registration number US20100255454A1. The application summary is: "A system and method for managing educational courses, a client..." retrieving training course data from the machine and creating a database based on this data. This includes extracting the equivalence data for the educational course. Educational course equivalence data. The training course equivalency data is transmitted to the client machine. This data belongs to the first training institution. the identification data of the first course and / or the 20th course of the second educational institution It includes the definition of the equivalence data for the educational courses, the first and second educational courses. directly equivalent, indirectly equivalent and / or mutually equivalent training courses This can be demonstrated. A user can show that the second training course is equivalent to the first training course. can be accepted as such, transferring the first course from the second educational institution to the second educational course. They can request that it be accepted as an equivalent training course, ask for additional information, or 25 Requesting opinions from other evaluators regarding the equivalence of the first and second courses. It is able to do so. The answer is: "..." As can be seen, the invention relates to a system and method for managing educational courses. and in addition, there is a 30 that can provide a solution to the disadvantages mentioned above. It does not mention the organizational structure. The application with the number TR2023 / 007509 was revealed as a result of technical investigations. Summary: "This invention allows students to transfer to a different university or a different department within the same university." to solve the horizontal transfer problems they experience during the transition and to make the adaptation process easier 35 Automatic horizontal transfer, an application that saves time for its realization. 3 It is related to the adaptation calculation system and its feature is bidirectional operation between numerous modules. communication is controlled over a network and at least one user is included in the system The server logs users into the system by having them enter the required information in the necessary fields. login, profile page where users view their own information, transfer requirements, Adaptation calculation, information module providing information about institutions and departments, required 5 The horizontal transfer module, where scanned documents are uploaded to the system, allows users to view the documents they have received. The course credit calculation module, which allows users to count courses, is used in the horizontal transfer module of the system. By scanning the documents they uploaded, they mark the necessary information on the documents and the student information The system retrieves and processes data from the database to determine if the student is eligible for a transfer. It contains a processor with an artificial intelligence algorithm that evaluates the state of being.” 10 As can be seen, application number 2023 / 007509 generally concerns horizontal transfer and adaptation. User login, profile page, information module, horizontal transfer module, adaptation process via a processor with a calculation module and an artificial intelligence algorithm that scans documents This relates to the evaluation of horizontal transfer / adaptation. 15 The differences between this invention and the one in application number 2023 / 007509 are as follows:  Tracking course data from various document and curriculum sources through source area analysis. combining into single course recordings,  Target curriculum data in terms of teaching language and curriculum source / version 20 verification,  For each resource course, instead of the entire target program, a limited number of candidate target courses. creating a list,  The proven matching output is subjected to deterministic rule checks and directly not considered a final decision, 25  Establishing a final decision situation after human review,  Finalized decisions should be kept as a single account status in the final exemption register.  Preventing the same source / target course from being counted again,  Group credit / ECTS calculation and credit-weighted grade conversion in multiple-to-one equivalencies to be done, 30  The report and audit trail are recorded in the database without rerunning the artificial intelligence model. Generated based on the definitive decision situation,  Creating decision consent without storing raw personal data. 4 Therefore, the invention is not limited to "automatic horizontal transition / adaptation calculation". The invention auditable, traceable source area, rule-controlled, and based on the final decision situation. This is a technical data processing system and is distinct from application number 2023 / 007509. In conclusion, due to the negative aspects described above and the current solutions, topic 5 Due to its shortcomings, it has become necessary to make improvements in the relevant technical field. Purpose of the Invention The invention represents a new breakthrough in this field, unlike the structures used in the current technology. The aim is to create a structure with different technical specifications that bring these elements together. The primary purpose of the invention is to introduce different institutions, programs, or curricula in higher education. Recognition, exemption, and adaptation of courses from related sources within the target program. 15 sources for evaluation were taken from academic documents and target curriculum resources. The goal is to develop a server-based system and method for processing data in a computer environment. The invention includes a user client device, an application server, a document processing server, and artificial intelligence. evaluation server or service connection, database server, file storage unit, network interface, data classification / model routing layer and reporting / output unit with 20 This can be done via the user's client device. The system that is the subject of the invention allows for horizontal or vertical transitions, internal program changes, and dual Major / minor, summer school, exchange program, special student / visiting student status, curriculum Different course recognition options such as course substitution and international credit recognition after course change 25 These scenarios can be configured via the same technical pipeline. The invention specifically refers to transcripts and course materials obtained from the student's source institution or the institution where they are taking courses. curriculum, course content, weekly lesson plan, learning outcomes, national credit, ECTS credit, and Information on achievement scores is based on the target higher education institution's program curriculum, course catalogs, 30 elective course pools, institutional exemption rules, and report formats. It allows for evaluation. The system can be configured specifically for a particular university, or for different higher education institutions. 35 It can be used. In an application, the system can determine the exemption criteria, courses, and other criteria of a specific target institution. It can be configured to work with catalogs and report formats. In another application, however... A different university's credit system, elective course structure, and decision-making rules are incorporated into the system. definable. The technical focus of the invention is to combine academic data from different document and curriculum sources. to make it verifiable, this data is in the course recordings that can be traced as a decision source. to combine, to limit AI-powered suggestions with deterministic controls, human to consolidate the decisions finalized by the parties into a single final decision case and to provide a report and audit. The goal is to generate the trace from the same exact data state. 10 In the invention, lesson data is extracted from source documents. This data is unified through source area tracking. The target curriculum data is collected in course recordings in terms of teaching language and curriculum source. It is verified. Proven matching and deterministic rule on a limited list of candidate courses. Controls are carried out. Directly identifying personal data is masked or reduced. 15 Data suitable for model evaluation is generated. This data is used in accordance with institutional policy and system. Depending on its configuration, it can be an external model service, a private cloud model, or an on-premises model server. or it can be processed on a local model. Using a local / in-house model is preferable. It can also be used as an application. Decisions finalized by humans are final. It is transferred to the exemption register. The report and audit trail are completed 20 days before the AI ​​model is run again. This is produced from a definitive decision situation. The technical core of the invention is the following elements mounted on a hardware data processing infrastructure: It relies on working together:  Extracting course data from heterogeneous sources: Transcript, course syllabus, course content, 25 course code, course name, grade, national from weekly plans or similar academic documents Extraction of credit, ECTS, and content information.  Creating unified course recordings through source area tracking: From different documents Incoming course information is combined into a single source course recording, and which area each section... The source of the document or data is stored. 30  Validation of target curriculum data: Target program courses, language of instruction, and Checked for curriculum resources, it should be noted that it is not in the wrong language or is inappropriate. Matching with curriculum data is prevented, or a warning is generated for human review.  Proven matching of a limited list of candidate courses: Source course, all target courses. Courses are not freely available in terms of code, name, content, prior approved equivalency, and program. 35 The evaluation takes place on a limited list of candidates, generated by signals such as ranking. 6  Selecting the model execution environment based on data type: such as transcripts and grade reports. Documents that may contain personal data are made public while they are in an internal or local model environment. Course content is delivered locally, in a private cloud, or via an external model, depending on institutional policy. It can be processed in the service.  Limiting AI-assisted evaluation with deterministic controls: 5 Proven matching assessment with fixed or deterministic parameters It can be implemented. Content overlap, credit / ECTS sufficiency, course status, decision-score. It is monitored by rules such as consistency and similar regulations.  Human confirmation: The system generates a suggestion, the final decision is accepted by the authorized user. It is finalized through rejection or modification. 10  Ensuring data integrity with the final exemption register: Finalized decisions, used Resource courses, target courses, elective capacity, multiple-to-one equivalencies, total Exempt ECTS credits and upper limit results are kept in a single account.  Generating reports and audit trails without re-running the model: Exemption / adaptation report and the decision trail, the stored decision state and the final 15 without recalling the AI ​​model. It is created through the exemption register. The technical problem – technical solution – technical impact matrix is ​​as follows: Technical problem, Technical solution, Technical effect Different types of PDFs and web pages. course data from resources to be taken Text with document processing server layer reading, image-based reading and student-based page separation Uniformity across different document types and processable course recording creation Transcript and program information remaining disconnected Application server and on the database server with source area tracking unique course recording For the same course, grade, credit, ECTS and Content information in a single record auditable combination In the wrong language or inappropriate matching with curriculum data Teaching target curriculum data language and curriculum source verification in terms of Not in line with the target program. risk of matching with data reduction The model proposal is variable and untraceable Limited list of candidates, deterministic. controls and decision summary The contradictory model outcomes of humans referred for examination and preservation of the decision trace 7 Technical problem, Technical solution, Technical effect Student containing personal data documents to external provider risk of transfer Data classification / model routing layer and in-house model run option Transcripts and grade documents in a local / in-house environment processing, public lecture different models for their content the selection of environments The report is being remodeled. depends on operation Final decisions and conclusions. from the notebook, AI server report without re-call production Lower transaction costs, more high repeatability and model independent report on change The same source or target course counting more than one Final exemption booklet report totals, elective in capacity and adaptation calculation data integrity Personal data with decision trail storing more than necessary Masking, numerical summary, and raw data. body concealing inspection trail While ensuring auditability the amount of personal data stored reduction Here, we will explain how the main technical pipeline works in different higher education course recognition scenarios. It is explained that it can be adapted. The following examples show the system's document retrieval, resource course... registration, target curriculum validation, rule-checked matching, human validation, without changing the elements of the final exemption booklet and audit trail production, different institutional rules 5 These are the preferred applications that can be configured accordingly.  In an application, the system includes summer school, exchange program, special student / guest student status, internal program change, double major / minor, or curriculum change It can be used in processes such as course substitution. In these cases, the source is academic records. This could be a full transcript from the institution the student came from, or a transcript for a specific course or subject. course syllabus, approval form, learning agreement or transcript belonging to the group it could be.  If a course has not yet been taken in an elective program, the system will calculate the passing grade. Instead of producing a definitive exemption or grade conversion result because it is not available, the course Preliminary eligibility output based on content, credit / ECTS, language of instruction and target course equivalent 15 It can be structured in a way that will create it. After the course is taken and the grade is documented. Then the same technical process follows: definitive recognition, final exemption register, and report generation. It can run the stages.  In another preferential practice, to the extent permitted by institutional regulations, ECTS or Micro-qualification, certificate or short program registrations containing equivalent workload information 20 8 The source can be considered an academic record. Such records can be used instead of a course. Whether it can be counted depends on the acceptance rules of the relevant institution; the system will make a final decision in this case. It can be configured to provide a warning for human inspection before production.  In an application, the relationship between the target course and the elective slot or pool is reliable. If defined in the database in this way, the relevant 5 elective courses are considered directly equivalent. It can be deducted from the pool capacity. If the relationship in question cannot be determined, the system will do so. Instead of acting on assumptions, it serves as a warning for human review. can mark.  In an application, if a student has given a preference order for exemption requests, the exempted ECTS credits will be calculated accordingly. If the limit is exceeded, this order of preference may be taken into account. If the student does not have a preference order, the system will use 10. It can generate recommendations based on success scores and leave the final selection to human review. can leave.  In an application, the system stores transcripts and grade documents that may contain personal data internally. In the model server or the local machine learning model, course content is based on Bologna. Catalog data and publicly available curriculum resources are also subject to institutional policy, as per the same 15. Hybrid model guidance by processing in the local model or in an external AI service. This structure can reduce the scope of data sent to the external provider while preserving course content. It enables the use of different model environments in its matching process.  In an application, once a report is generated, it cannot be modified or is a final signed document. It can be saved as a record. Thus, the same final exemption 20 can be included in the report later. It can be verified that it was produced from his notebook.  In an application, the source university or the target institution may have different student information systems. If it uses them, the system will work with data retrieval connectors specific to these different sources. It can be expanded in this way. It is uploaded when the course syllabus cannot be found on the web resource. Backup readings can be done from the PDF course syllabus. 25 In order to fulfill the purposes described above, the invention may be used in different institutions of higher education. Recognition or exemption of courses from program or curriculum resources in the target program. It is a server-based system for adaptation and evaluation, and its feature is;  A 30-page document that enables access to the source academic document set and authorized user interaction. user client device,  an application server that includes a microprocessor and memory and runs the main pipeline,  A separate system from the application server that performs document reading and page parsing operations. and a document processing server that can run within it,  enables AI-assisted evaluation of course content, using the local processor / GPU 35 model-based server, on-premises machine learning model, private cloud model, or 9 an AI assessment in the form of an external model service accessed over the network server or service connection,  source course recordings, target curriculum data, matching decisions, and decision a database server that stores its state,  A file that stores source academic documents, document fragments, and numerical summaries. 5 storage unit  Data communication between user client devices, servers, and organizational catalogs a network interface that provides,  A company that produces the exemption and adaptation report as an electronic output or printable document. reporting and output unit, 10  whether the data to be processed contains personal data or publicly available course content taking into account whether it is of a certain nature, the data is stored on the local model server, on-premises. in machine learning model, private cloud or external AI evaluation Data reduction that specifies how data is processed at the server or service connection, Masking and model orientation layer, 15  Text layer running on a document processing server or application server a system that extracts text from documents and processes scanned documents as images. document reading and page separation module,  Course code, course name, and achievement score from the document reading and page separation module. A resource that extracts course content information such as grade, national credit, ECTS, and course content is called Course Inference 20. module,  Fragmented information from transcripts, course syllabi, and supplementary documents, identifying which area each section belongs to. The source area of ​​the document will be tracked in a single source course record. a resource course deduplication that combines and stores this record on the database server module, 25  Course catalog data that is compatible with the target program's language of instruction and curriculum version a teaching language that enables selection and generates a warning for human review in case of uncertainty and curriculum resource validation module,  For each resource course, the course code, course name, previous human-approved equivalencies, and content are provided. Using proximity signals, a limited candidate is selected instead of the entire target program. 30 A candidate target course generation module that defines a target course list.  Model environment defined by the data reduction, masking, and model orientation layer through, the source course and the prospective target course content, learning outcomes, weekly topics, Comparison of content overlap and rationale information in terms of national credit and ECTS. a proven matching assessment module that produces 35  running on the application server's microprocessor and proof-of-matching The output of the evaluation module should not be accepted as a final decision due to content overlap. from threshold, credit / ECTS sufficiency, passing status and decision-score consistency checks a rule engine that passes,  Situations where multiple source courses correspond to a single target course are considered together in 5 to evaluate, calculate the group national credit and ECTS total and credit-weighted a many-to-one equivalence module structured to generate a success score,  source course, target course, credit, ECTS, grade, content overlap, alternative candidates and presenting the alerts to the user's client device; requesting the authorized user to accept the suggestion, a person who enables him to change or reject and finalize the decision 10 Review and finalization interface,  finalized decisions, resources used and target courses, elective slots and total used resource courses that combine exempt ECTS credits into a single calculation. preventing reuse and calculating the adaptation semester / year with the exempt ECTS upper limit. a final exemption ledger managed through this account status, 15  The document on which the decision is based, source course recording, matching proposal, human intervention, and rule engine results; without storing raw personal data and with artificial intelligence evaluation. without restarting the server or service connection, on the database server An audit trail module that shows the recorded decision status. It includes. 20 The structural and characteristic features and all the advantages of the invention are given in the figures below. Thanks to the detailed explanation written with references to the figures, it becomes clearer. This will be understood, and therefore the evaluation should also take these figures and detailed explanations into account. It must be done by taking 25. Figures that will help understand the invention. Figure 1 is a schematic representation of the system that is the subject of the invention. Figure 2 is a schematic representation of document processing and source course recording. 30 Figure 3 is a schematic representation of evidential matching, rule checking, and human confirmation. Figure 4 is a schematic representation of the many-to-one equivalency and weighted grade calculation. Figure 5 is a schematic representation of the final exemption register, categorical adjustment, and upper limit calculation. Figure 6 is a schematic representation of the report and audit trail without rerunning the model. Figure 7 shows a schematic representation of the modules used in the system and method that are the subject of the invention. 35 11 The drawings do not necessarily need to be scaled and are necessary for understanding the invention. Details that are not present may have been overlooked. Furthermore, at least to a large extent... Elements that are identical or at least have substantially identical functions are numbered the same. It is shown. Explanation of Part References 10. Source academic document set 20. Document retrieval and verification module 30. Document reading and page separation module 10 40. Resource lesson drawing module 50. Resource lesson consolidation module 60. Personal data protection sub-module 70. Target curriculum acquisition module 80. Instructional language and curriculum source validation module 15 90. Candidate target course production module 100. Evidence-based matching evaluation module Rule 110 engine 120. Many-to-one equivalence modulus 130. Human review and validation interface 20 140. Final exemption booklet 150. Categorical adaptation module 160. Report generation module 170. Audit trail module 180. Institutional Equivalence Memory 25 190. Job sequence management module 200. User client device 210. Application server 220. Document processing server 230. AI evaluation server or service connection 30 240. Database server 250. File storage unit 260. Network interface 270. Reporting and output unit 280. Data reduction, masking, and model orientation layer 35 12 1001. The source academic document set (10) from the user client device (200) network interface (260) transmitting to the application server (210), document retrieval and verification Checking the file suitability in accordance with module (20) and the document file to be stored in the storage unit (250), 1002. Application of the numerical summary of the document stored in the file storage unit (250) 5 calculated by the server (210) and stored in the database server (240) and detecting the re-upload of the same document through this summary, 1003. Document reading and page separation module (30) document processing server (220) or by running on the application server (210), from documents containing a text layer Text extraction and image processing of scanned documents, resulting in 10% of the data obtained. kept in the file storage unit (250), 1004. In document sets containing documents belonging to more than one student, the document processing server (220) or the document by determining the page ranges by the application server (210) Separation on a student basis 1005. According to the resource lesson extraction module (40), the document processing server (220), application 15 server (210) and data reduction, masking and model routing layer (280) the model environment or AI evaluation server or service it has specified Using the link (230), course code, course name, grade, national credit, ECTS and course Extraction of content information and storage on the database server (240), 1006. In accordance with submodule (60) on personal data protection, by the application server (210) 20 Protecting identity data through masking or digital hashing, documents in the file storage unit (250) and in the database server of the reduced decision data (240) Data reduction, masking and modeling of sources containing personal data by storing personal data. Data reduction, masking and modeling by the routing layer (280) implementation of guidance, 25 1007. In accordance with the resource course deduplication module (50), fragmented parts from different documents information on a single source course record on the application server (210) combined and stored on the database server with the source information of each field (240) hiding, According to the 1008. Target curriculum acquisition module (70), target program courses, service courses 30 and data on elective course structures by the application server (210) Obtained from the database server (240) or via the network interface (260), 1009. In accordance with the instruction language and curriculum source validation module (80), the application by the server (210) to the language of instruction and curriculum version of the target program suitable course catalog data from the database server (240) or network interface (260) 35 Selecting from a list and generating a human review alert in case of uncertainty, 13 1010. Resources stored on the database server (240) by the application server (210). Checking course registrations in terms of passing status and eligibility for exemption. Maintaining national credit and ECTS values ​​in separate areas, 1011. According to the candidate target course production module (90), by the application server (210) Using course data (240) on the database server and previous approved equivalencies, 5 For each source course, a limited list of candidate target courses is determined. According to the evidential matching evaluation module (100), data reduction, masking and The model environment determined by the model guidance layer (280) according to the nature of the data or via AI assessment server or service connection (230), 10 with the resource course that the application server (210) receives from the database server (240) candidate target courses in terms of content, learning outcomes, national credits and ECTS Comparing content and generating rationale information, 1013. Rule engine (110) on the microprocessor of the application server (210) By running it, the rule set defined in the database server (240) of the generated suggestion Accordingly, the content overlap threshold, credit / ECTS sufficiency, and decision-score consistency are 15. subjected to deterministic controls in terms of 1014. According to the many-to-one equivalence module (120), by the application server (210) Multiple source courses retrieved from the database server (240) into a single target course together with the evaluation, calculation of the total group national credits and ECTS, and Generating a credit-weighted performance score, 20 1015. In accordance with the Human Review and Finalization Interface (130), the application server (210) source / target course, credit, ECTS, grade, received from (240) database server, Reason, alternative and warning information is transmitted to the user client via the network interface (260). to the device (200), 1016. Acceptance given by the authorized user via the user client device (200), modification 25 or rejection inputs to the application server (210) via the network interface (260) transmission and final decision status of the decisions in the database server (240) hiding, 1017. Creation of the final exemption book (140) by the application server (210), finalized decisions, resources used and target courses, elective slots and 30 total exempted ECTS value as single account status in database server (240) storage and prevention of reuse of used resource lessons, According to the 1018. Categorical adaptation module (150), by the application server (210) Categorical adjustment only occurs after final decisions and in the database. This should be done according to the remaining selectable capacity received from the server (240), 35 14 1019. Total exemptions in the final exemption book (140) by the application server (210). Checking the ECTS value against the upper limit defined in the database server (240). and marking the excluded courses, 1020. The total of final exemption ECTS after the upper limit by the application server (210) calculated and transferred to the adaptation semester / year account and stored on the database server 5 (240) storage, 1021. The report production module (160) is run on the application server (210), final decisions and final exemption register (140) in the database server (240) Production of exemption and adaptation report and reporting and output unit (270) presented through, 10 1022. According to the audit trail module (170), the decision by the application server (210) the document it is based on, the matching proposal, human intervention and the rule engine (110) the result of; artificial intelligence evaluation server or service connection (230) Decision status recorded only on the database server (240) without being re-run 15 1023. In accordance with institutional equivalence memory (180), human-approved equivalences are stored on the database server (240) by the application server (210) to be stored and used in candidate target course production in subsequent evaluations. According to module (120) 1014.1 Many-to-one equivalence, 20 by application server (210). from among the source course records stored on the database server (240) a single target course a resource course group is formed by identifying multiple resource courses that may correspond to each other. creation, 1014.2 The resource courses that make up the group are set by the application server (210). joint evaluation of their contents, 25 1014.3 Application server (210) national credit of resource courses forming the group and by summing the ECTS values ​​separately, the total of group national credits and group ECTS is calculated. calculation, 1014.4 Target received by application server (210) from database server (240) According to course credit and ECTS information, the credit / ECTS proficiency requirement of the target course is a single source 30 monitoring based on group totals instead of individual lessons, 1014.5 The success grade to be written to the target course by the application server (210), group Calculation based on the credit-weighted average of the source courses that make up the system. 1014.6 Whether the weight data is found by the application server (210) Checking the source course records (240) on the database server, weight data 35 If it is not found, a warning will be issued instead of generating a passing grade, and the situation will be reported to the human element. left for review. According to module 1022.1 Audit trail (170), the database is accessed by the application server (210). resource course recordings, matching suggestions and finalized 5 stored on the server (240) reading of the decisions, 1022.2 By the application server (210), decision summary, candidate target course summaries, human interventions and rule engine (110) results are retrieved from the database server (240) and single integrating them into a decision-making chain, 1022.3 Audit permission by application server (210), AI evaluation 10 server or service connection (230) without being called again and with raw transcript text Without storing the open identification number, only the decision registered on the database server (240) to be created based on the situation, 1022.4 Reporting of the report and audit trail generated by the application server (210). and its presentation as an electronic output via the output unit (270). 15 Detailed Description of the Invention In this detailed explanation, the preferred configurations of the invention are not merely for better understanding the subject. 20 It is explained. The invention involves combining courses from different institutional, program, or curriculum sources in higher education. academic resources for recognition, exemption, and adaptation assessment within the target program 25 that process data from documents and target curriculum resources in a computer environment It involves a server-based system and method. The elements and functions used in the system and method that are the subject of the invention are as follows: The source academic document set (10) includes the student's transcript, course syllabus, commission form, and Appendix 30. It is a document set consisting of documents. User login is provided from the client device (200) and The file is stored in the storage unit (250). The document import and verification module (20) handles the import of documents into the system, file type / size. It ensures the control and storage of the digital summary of the document. 35 16 Document reading and page separation module (30), outputs PDF text, displays scanned documents It is a module that functions as a tool for separating page ranges in documents with many students. The resource lesson extraction module (40) retrieves the course code, name, grade, national credit, ECTS and content from the document. It is the module that extracts the information. 5 The source course unification module (50) tracks transcript and syllabus information with source area tracking. It is a module that combines all course recordings into a single source. Personal data protection submodule (60), which prevents the open storage of sensitive identity information, 10 This module provides auditing capabilities through masking and numerical hashing. Target curriculum acquisition module (70), target program courses, service courses, elective courses It is the module that covers the structures and program curriculum. The language of instruction and curriculum source validation module (80) is the language of instruction of the target program and This module takes into account the suitability of the available curriculum records. Candidate target course generation module (90), for source course target course candidates code, name, previous It is a module that determines using signals such as verified equivalence and content proximity. 20 Evidence-based matching assessment module (100), source and target courses content, learning The output is a module that compares weekly topics, credits, and ECTS credits. Rule engine (110), content overlap, credit / ECTS sufficiency, passing status, upper limit and grade 25 It is the unit that enforces rules such as transformation. Multiple-to-one equivalence module (120), the equivalence of multiple source courses to a single target course. This module generates group credit / ECTS accounts and weighted grades when counted. Human review and confirmation interface (130), the authorized user to accept the proposal, It is the unit that allows the decision to be changed or rejected and finalized. The final exemption book (140) brings together the final decisions in a single account status, again It is the unit that manages the consistency of usage, elective capacity and total ECTS credits. 35 17 Categorical adaptation module (150), successful courses that do not directly match, remaining in the final notebook This module evaluates based on elective capacity. Report generation module (160), exemption from final decisions and final exemption book and This module generates the adaptation report. 5 The audit trail module (170) traces the decision chain without storing raw personal data and model re-engineering. It is a module that displays the information without running it. Institutional equivalence memory (180) 10 previously human-approved equivalencies It is the unit that uses this as a clue in candidate generation. Job sequence management module (190), document processing, target data acquisition, matching and report generation. It is the module that runs processes like these in the background. User client device (200), authorized user can upload, review, make decisions and It is the device used to view reports. This could be a computer, tablet, or terminal device. It is possible. Application server (210), containing microprocessor and memory, main pipeline, document retrieval, 20 Performing the processes of deduplication, candidate generation, rule checking, final ledger and reporting. It is the server. Document processing server (220), PDF text extraction, image-based reading and page It is the processor-based unit that performs the parsing operations. It can be a separate server or an application 25 It can also run within the server (210). AI assessment server or service connection (230), source and target course Local processor / GPU based, enabling AI-powered evaluation of content. model server, on-premises machine learning model, private cloud model, or over the network 30 It is an external model service that is accessed. The database server (240) stores source course records, target curriculum data, and matching data. It is the server that stores the decisions, the final exemption register, and the audit trail records. 35 18 File storage unit (250), source academic documents, document fragments, numerical summaries. It is the storage area that stores report files. Network interface (API link layer) (260), user client device (200), servers (210, 220, Data between institutional catalogs, student information system and artificial intelligence service (230, 240). It is the unit that facilitates communication. The reporting and output unit (270) provides the exemption / adaptation report as an electronic printout, PDF or It is the unit that produces or presents printable documents. Data reduction, masking and model orientation layer (280), personal data of the source document taking into account whether it contains or is of a publicly available course content Is the data on the local model server, in the on-premises machine learning model, or in the private cloud? This unit determines whether the data will be processed by the AI ​​or an external service. Data reduction, Masking and model orientation layer (280), 15 to be sent for model evaluation a model that masks or reduces directly identifiable personal data in the data defining the scope of data to be submitted for evaluation, raw personal data within the decision trail prevents data from being stored externally according to institutional policy or system configuration. model service, private cloud model, on-premises model server, or local model This is the technical layer that determines the processing. In this layer, the Turkish Republic identity number, student number 20 When directly identifying fields such as number and name are masked / reduced; course Data such as name, letter grade, ECTS, national credit, and course content are external to the institution's policy. The model service can be processed in a private cloud, on-premises model, or local model environment. is happening. The invention works on the following principle: The invention covers the course recognition, exemption, and adaptation process from source document acquisition to final report production. It executes as a traceable pipeline. This pipeline runs through the user client device (200), application server (210), document processing server (220), artificial intelligence evaluation 30 server or service connection (230), database server (240), file storage unit (250), network interface (260) and data reduction, masking and model routing layer (280) and can be done with the reporting / output unit (270). Obtaining source documents and checking their processability 35 19 The system uploads transcripts, course syllabi, and commission documents via the user client device (200). Application of source documents such as form or additional academic document via network interface (260). It transmits to the server (210). The microprocessor of the application server (210) processes the file type, size, It performs page count and processability checks. The document file is stored in the file storage unit. (250) can be stored. The numerical summary of the document is calculated by the application server (210) and 5 is kept on the database server (240) so that the same document can be uploaded again later This allows for detection. Document reading and student-based parsing The document processing server (220) or application server (210) processes text in source documents. If there is a text layer, it extracts text directly. If there is no text layer or the document is scanned... If the document is in image format, the pages are processed as images. The document contains the papers of multiple students. If it contains such content, the server microprocessor determines the page ranges and processes the document on a student-by-student basis. It can be separated. Thus, a separate resource course dataset is created for each student. Extraction and deduplication of source course data. Document processing server (220), application server (210) and data reduction where necessary, local model server selected by masking and model routing layer (280) or AI assessment service link (230), course code from documents read, course 20 extracts the information. The application server (210) extracts the course list and course syllabus found in the transcript. or by combining content information found in supplementary documents to create a unique resource for each course. creates course records. These records are stored on the database server (240) and each field is which It is also recorded that it came from the document or data source. Personal data protection and decision data reduction. The application server (210) stores the necessary technical information so that the decision process can be monitored. Personal data such as raw transcript bodies or public identification numbers are used unnecessarily. It can function in a way that will prevent it from being retained. Document summary, summary of evaluation inputs, candidate Lesson summaries, implemented controls and decision changes database server (240) personal 30 It can be recorded in a way that contains no personal data or contains reduced personal data. Data reduction, masking and model orientation layer in practice (280), transcript and note Instead of sending resources that may contain personal data, such as documents, to an external modeling service, send them internally to the organization. to run on the model server or on the local machine learning model Configurable. 35 Acquisition and validation of target curriculum data. Application server (210), via network interface (260) or from database server (240) The student receives information about the target program's courses, service courses, elective course structures, and curriculum. The language of instruction of the program is taken into consideration. For Turkish programs, Turkish course catalog data is used. For English programs, English course catalog data is preferred. Official effective year or 5 In applications that contain curriculum version information, the system can perform verification based on this information. If no information is available, find course content and program links among the existing curriculum records. The most suitable record is selected; in case of uncertainty, a warning can be generated for human review. Preliminary suitability check of source courses 10 The application server (210) is taken into consideration for exemption evaluation of the source course. It conducts initial checks to determine if it can be taken. Failed courses, preparatory courses, grade. Insufficient courses or courses not covered by exemption are included in the matching totals. It is not accepted or is marked as requiring human review. National credit and ECTS are separate. They are preserved in sections and used separately in subsequent accounts. 15 Determining candidate target courses. The application server (210) targets each resource course that can be evaluated. determines the candidate courses in the program. During this determination, the database server (240) Course code, course name, previously approved equivalencies, course content similarity, and target 20. Signals such as the course order in the program can be used together. This allows for AI-powered solutions. The evaluation, instead of freely working on the entire target program, is technically focused. This is done from a prepared and limited list of candidates. Proven matching and deterministic controls 25 The application server (210) database contains source course and candidate target course data. It receives data reduction, masking and model routing layer (280) from the server (240). Depending on the nature of the data, local model server, on-premises machine learning model, private cloud model or external artificial intelligence evaluation server or service connection (230) It can specify that the evaluation should be done based on this. AI-powered evaluation, every 30 The evaluation may generate content overlap, acceptance or rejection grounds, and warnings for the candidate course. with fixed or deterministic parameters that support producing the same result for the same inputs It can be run. The application server (210) accepts the model output directly as the final decision. It does not. Content overlap threshold, credit / ECTS sufficiency, course sequence number compatibility, score and decision. It performs checks such as consistency, passing grade, and course status on its own microprocessor. 35 21 Matching multiple source courses to a single target course. If necessary, the application server (210) combines multiple source courses into a single target course. They can be evaluated together. In this case, the content of the source courses is considered together. ECTS and National credit values ​​are summed up, and the credit / ECTS qualification requirement group total of the target course is calculated. It is checked via this. The success grade to be written for the target course is the weighted average of the credits of the source courses. It can be calculated using the average. If there is no national credit, the ECTS value can be used as a weight. If no weighting data is available, the system does not generate a score and leaves the matter to human review. Human review and confirmation The application server (210) displays the AI-powered suggestions and the results of the rule checks in 10 The authorized user provides the resource to the client device (200) via the network interface (260). course, target course, national credit, ECTS, passing grade, content overlap, alternative target courses and The user sees the alerts together. The user can accept or reject the suggestion, or change the target course. or you can edit the match score. Once the review is complete, the decisions are final. The status is stored on the database server (240). 15 Final exemption booklet After finalization, the application server (210) creates the final exemption book and It stores the finalized equivalence decisions in a single database server (240). It aggregates the information in the account. Resource courses used, resource course groups, exempted target 20. courses, elective slots, direct equivalencies, many-to-one equivalencies, categorical adaptation decisions, elective capacity utilization, total exempted ECTS, upper limit control and adaptation The semester / yearly accounts are managed from this ledger. The ledger's technical functions include reporting and various accounting modules. The goal is to ensure data consistency between them. Categorical adaptation and selective capacity control The application server (210) can only determine categorical adaptation based on final decisions. It runs. Courses that are considered directly equivalent or have been used in the source course group. It cannot be reused in categorical adaptation. During categorical adaptation, in the target program... The elective slot or pool capacity is obtained from the database server (240). A resource course 30 If the ECTS value exceeds the remaining capacity and partial credit application is not defined, the system will do this. It does not automatically place the course in the relevant elective pool and generates a warning. Exemption ECTS upper limit and adaptation calculation The application server (210) states that the total exempted ECTS in the final exemption booklet is in accordance with institutional rule 35. It checks whether the student's preference information exceeds the upper limit specified in the set. If it does, the student's preference information is displayed. 22 If available, this information can be taken into consideration. If student preference information is not available, the system will sort by grade. They can make a choice and present this situation as a warning requiring human review. Year of adaptation. The calculation is also done based on the final ECTS total after exceeding the same upper limit. Generating reports and audit trails without rerunning the model 5 The report generation module is executed by the application server's (210) microprocessor and Resource course records, matches, human stored on the database server (240) It uses finalizations, the final exemption register, and rule engine results. Reporting and auditing. trace, AI assessment server or service connection (230) without being called again, Reporting and output unit (270) is generated via the registered decision status. Decision summary 10 Model name, numerical summary of the evaluation system, candidate course summaries, pre- and post-decision. The system may store elements that do not contain personal data, such as implemented controls and warnings. Concrete application example In one example application, the student's "Physics I" course taken at his previous institution is worth 3 national credits and 4.15 The ECTS credits for the course "Physics Laboratory I" are equivalent to 1 national credit and 2 ECTS credits. Target The course "General Physics I" is offered in the program with 4 national credits and 6 ECTS credits. Commission These two resource courses can be evaluated together using a symbol or a rule defined in the system. This In this case, the application server (210) stores the resource course on the database server (240). By reviewing the records, the target course requirement is 20, totaling 4 national credits and 6 ECTS credits. It determines whether it meets the content overlap threshold when course contents are evaluated together. If provided, it generates a multiple-to-one equivalency proposal. The success grades of the source courses are credits. It is calculated based on weight. If the necessary weight data is not available, the system does not generate a score, and the human factor is not present. It gives a warning regarding the review. The target course is counted once in the final exemption register. Same source. Lessons cannot be reused in another direct or categorical match. 25 In the system that is the subject of the invention, source academic document set (10) user client device (200) It is received via and transmitted to the application server (210) via the network interface (260). Document Instructions of the receiving and verifying module (20), microprocessor of the application server (210) It is executed by [company name]. File type, size, number of pages, and processability are checked. Document 30 The file is in the file storage unit (250), and the numerical summary of the document is in the database. can be stored on the server (240). Document reading and page separation module (30), document processing server (220) or application It is run on the server (210). If the document contains a text layer, direct text extraction is performed. 35 If it is a scanned image, image-based reading is performed. For multiple students 23 When documents are in the same package, page ranges are separated on a student-by-student basis. Individual assessments are provided for each student. The source lesson extraction module (40) comes from the document reading and page separation module (30). The data includes course code, course name, grade, national credit, ECTS, course content, and learning outcome 5. and extracts weekly lesson plan information. Resource lesson deduplication module (50), same resource Combines different pieces of information about the course on the application server (210) and the resulting course It stores the records on the database server (240). The personal data protection submodule (60), student It prevents the unnecessary and overt storage of sensitive information such as identity numbers. Target curriculum acquisition module (70) is the program curriculum of the target higher education institution, course list, service courses, elective course structures and, if applicable, the Bologna course catalog. It retrieves its data from the database server (240) or via the network interface (260). Language of instruction and curriculum source validation module (80), course compatible with the language of instruction of the target program. It enables the use of data. Information on the official effective year or open curriculum version is available. This information is used in the applications where it is available; if this information is not available, the system selects the most suitable curriculum record. It selects and generates a warning in case of uncertainty. The candidate target course generation module (90) determines the target course candidates for each source course. Proven matching assessment module (100), 20 between source course and candidate target course content overlap, learning outcome similarity, weekly topic alignment, national credit and ECTS It evaluates its competence. The AI-powered component is an in-house machine learning model. private cloud model, local AI evaluation server or service connection (230) This can be done with the data reduction, masking and model orientation layer (280), Resources that may contain personal data, such as transcripts and grade reports, may be made available internally or publicly on 25% of the institutional platform. Open course content can be presented in a local or external model environment, depending on institutional policy. It can enable the evaluation of these suggestions. However, these suggestions are implemented by the rule engine (110). Content overlap threshold, credit / ECTS sufficiency on the server's (210) microprocessor, The grade is checked against criteria such as success score, course status, and decision-score consistency. The multiple-to-one equivalence module (120) is the equivalence of multiple source courses to a single target course. In such cases, it is activated on the application server (210). This module, source They jointly evaluate the course content, calculate the total national credits and ECTS credits, and determine the target. It checks the course's proficiency requirements based on the group total. The passing grade to be written for the target course, The grade is calculated using the weighted average of the credit hours of the source courses. If the required weighting data is not available, the grade is 35. It is not fabricated and a warning is issued regarding human scrutiny. 24 Human review and refinement interface (130), produced by application server (210) The authorized user presents the suggestions to the user client device (200). The source course, target course, The authorized person sees ECTS credits, national credits, achievement scores, content overlap, alternative candidates, and warnings. The user can accept or reject the suggestion, change the target course, or lower the matching score to 5. can be edited. When the review is complete, the final decisions are stored in the database server. (240) is kept. After finalization, the application server (210) creates the final exemption book (140). Final exemption booklet, source courses used, target courses used, multiple-choice courses 10 groups, elective course slots, categorical adaptation decisions, total exempted ECTS, upper limit Courses excluded due to reason and final ECTS credits to be used in the adaptation semester / year calculation. It stores the total as a single account statement. The categorical adaptation module (150) only works with finalized decisions. Previously 15 Resource courses used in directly matched or multiple-choice groups are categorized. It cannot be reused during adaptation. The system uses the selective slot or pool in the target program. It obtains its capacity from the database server (240). If the remaining capacity is insufficient, the relevant course It is not placed automatically and an alert is generated for human inspection. Report generation module (160), exemption and adaptation report final exemption register (140) It produces it through. The audit trail module (170) determines which documents and which courses the report is based on. from the records, which matching suggestions, which human interventions, and which rules This indicates that the audit trail consists of engine results. This audit trail is either raw transcript text or clear It is generated without storing the identification number and without rerunning the AI ​​model. 25 The report is available as an electronic file, PDF or printable via the reporting and output unit (270). It can be presented as output. The invention can be implemented as a standalone, server-based system, or as a... It can also operate within a higher education management system or a university accreditation platform. 30 In one application, the system could be a student affairs or application management system, or a university course catalog. or Bologna information system, target program curriculum database, source university course curriculum resources, committee or faculty review interface, report generation and The archiving system works together with the audit and decision monitoring infrastructure. These components are networked. Data exchange with the application server (210) via the interface (260) and API link layer 35 He can. Thanks to this holistic structure, the system goes beyond being just a tool for matching courses. Data source and decision status for exemption and adaptation decisions in higher education institutions. ensuring consistent implementation of the rule, human review, and report output. It becomes a technical infrastructure. 5 The steps involved in the process carried out with the system that is the subject of the invention are listed below:  The source academic document set (10) from the user client device (200) network interface (260) Transmission to the application server (210) via the document retrieval and verification module (20) Accordingly, the file's suitability is checked and the document is stored in the file storage unit on 10 days. (250) storage (1001),  Application of the numerical summary of the document stored in the file storage unit (250). calculated by the server (210) and stored on the database server (240) and Detection of the re-upload of the same document through this summary (1002),  Document reading and page separation module (30) document processing server (220) or 15 By running on the application server (210), text from documents containing a text layer is extracted. Extraction and processing of scanned documents as images, and the resulting data being compiled into a file. (1003), to be kept in the storage unit (250),  In document sets containing documents belonging to more than one student, the document processing server (220) or by setting the page ranges by the application server (210) the document 20 separation on a student basis (1004),  According to the source lesson drawing module (40), the document processing server (220), application server (210) and data reduction, masking and model routing layer (280) the model environment or AI evaluation server or service connection it has specified Using (230), course code, course name, grade, national credit, ECTS and course content information can be obtained. Extraction and storage on the database server (240) (1005),  In accordance with the personal data protection submodule (60), the identity is determined by the application server (210). protecting data through masking or digital hashing, file documents in the storage unit (250) and in the database server of the reduced decision data (240) data reduction, masking and model 30 by storing sources containing personal data. Data reduction, masking and model routing by the routing layer (280) implementation (1006),  In accordance with the source course unification module (50), fragmented information from different documents combined into a single resource course recording on the application server (210) and each (1007), 35 The source information of the field is stored on the database server (240). 26  In accordance with the target curriculum acquisition module (70), target program courses, service courses and Data on elective course structures are stored in the database by the application server (210). (1008) to be received from the server (240) or via the network interface (260),  In accordance with the language of instruction and curriculum resource validation module (80), the application server (210) Course 5 suitable for the language of instruction and curriculum version of the target program. catalog data from the database server (240) or via the network interface (260) selection and generation of a human review warning in case of uncertainty (1009),  Resources stored on the database server (240) by the application server (210). Checking course registrations in terms of passing status and eligibility for exemption. Preservation of national credit and ECTS values ​​in separate fields (1010), 10  According to the candidate target course production module (90), by the application server (210) Using course data (240) on the database server and previous approved equivalencies, each Determining a limited list of candidate target courses for the source course (1011),  According to the evidence matching evaluation module (100), data reduction, masking and The model guidance layer (280) determines the model environment based on the nature of the data or 15 via artificial intelligence evaluation server or service connection (230), application candidate target with the resource course received by the server (210) from the database server (240) The content of the courses is compared in terms of content, learning outcomes, national credits, and ECTS credits. overlap and production of justification information (1012),  The rule engine (110) is located on the microprocessor of the application server (210) 20 by running, the generated proposal is in accordance with the rule set defined in the database server (240) content overlap threshold, credit / ECTS sufficiency and decision-score consistency passing through deterministic controls (1013),  According to the many-to-one equivalence module (120), by the application server (210) Multiple source courses retrieved from the database server (240) compared to a single target course 25 joint evaluation, calculation of the group national credit and ECTS total, and credit Production of weighted success grade (1014),  In accordance with the human review and confirmation interface (130), the application server (210) Received source / target course, credit, ECTS, grade, justification from database server (240), Alternative and warning information is sent to the user client device (200) via the network interface (260) 30 presentation (1015),  Acceptance, modification or given by the authorized user via the user client device (200) Transmission of rejection inputs to the application server (210) via the network interface (260) and storing the decisions as final decision status on the database server (240) (1016), 35 27  Creation of the final exemption book (140) by the application server (210), finalized decisions, resources used and target courses, elective slots, and total exempt ECTS value as single account status in database server (240) storage and prevention of reuse of used resource courses (1017),  In accordance with the categorical adaptation module (150), categorical 5 is applied by the application server (210). adaptation only after final decisions and from the database server (240) to be done according to the remaining elective capacity received (1018),  Total exempted ECTS (140) in the final exemption booklet by the application server (210) its value is checked against the upper limit defined in the database server (240) and excluded Marking of dropped courses (1019), 10  The total of final exempted ECTS credits after the upper limit by the application server (210) calculated and transferred to the adaptation semester / year account and in the database server (240) storage (1020),  By running the report generation module (160) on the application server (210), Final decisions and final exemption register (140) in the database server (240) 15 Production of exemption and adaptation report and reporting and output unit (270) presented through (1021),  According to the audit trail module (170), the decision is made by the application server (210). the result of the document on which it is based, the matching proposal, human intervention and rule engine (110); AI evaluation server or service connection (230) again 20 without being run, only through the decision status recorded on the database server (240) creation and presentation through reporting and output unit (270) (1022),  In accordance with institutional equivalence memory (180), human-approved equivalencies stored by the application server (210) on the database server (240) and Use in candidate target course production in subsequent evaluations (1023). 25 The sub-steps performed under step 1014 are as follows:  According to the many-to-one equivalence module (120), by the application server (210) a single target among the resource course records stored on the database server (240) Multiple resource courses that could correspond to the course are identified, and one resource course is 30. formation of the group (1014.1),  The resource courses that form the group are determined by the application server (210). joint evaluation of their contents (1014.2),  The application server (210) provides the national credits for the resource courses that make up the group. The ECTS values ​​are added together separately to obtain a total of 35 group national credits and group ECTS. calculation (1014.3), 28  Target received by the application server (210) from the database server (240) Based on course credit and ECTS information, the credit / ECTS proficiency requirement of the target course is singular. Monitoring based on group total instead of source course (1014.4),  The success grade to be written to the target course by the application server (210), group Calculation of the weighted average of the source courses (1014.5), 5  Whether the weight data is found by the application server (210) Checking the source course records (240) on the database server, If weighting data is missing, a success score will not be generated and a warning will be issued. leaving the situation to human investigation (1014.6). The sub-steps performed under step 1022 are as follows:  According to the audit trail module (170), the database is controlled by the application server (210). resource course recordings, matching suggestions and stored on the server (240) Reading of final judgments (1022.1),  By the application server (210), decision summary, candidate target course summaries, human 15 interventions and rule engine (110) results from the database server (240) taking and combining into a single decision chain (1022.2),  Audit permission by the application server (210), artificial intelligence evaluation server or service connection (230) without being called again and raw transcript without storing the text and the explicit identification number, only on the database server 20 (240) creation based on registered decision status (1022.3),  Reporting of the report and audit trail generated by the application server (210) and its presentation as an electronic output via the output unit (270) (1022.4).

Claims

29 REQUESTS 1. Courses in higher education that come from different institutions, programs, or curriculum sources. a server-based system for recognition, exemption, and adaptation assessment within the target program It is a system, and its feature is; 5  Obtaining the source academic document set (10) and authorized user interaction a user client device (200),  an application server that includes a microprocessor and memory and runs the main pipeline (210),  The application server (210) performs document reading and page parsing operations. a document processing server (220) that can run separately or within it, 10  Local processor / GPU that enables AI-assisted evaluation of course content. model-based server, on-premises machine learning model, private cloud model, or an AI assessment in the form of an external model service accessed over the network server or service connection (230),  source course recordings, target curriculum data, matching decisions, and decision 15 a database server that stores its status (240),  a file that stores source academic documents, document fragments, and numerical summaries storage unit (250),  User client devices (200), servers (210, 220, 230, 240) and institutional catalogs a network interface that enables data communication between (260), 20  whether the data to be processed contains personal data or publicly available course content taking into account whether it is of a certain nature, the data is stored on the local model server, on-premises. in machine learning model, private cloud or external AI evaluation Data reduction that determines processing at the server or service connection (230), Masking and model orientation layer (280), 25  Text that runs on the document processing server (220) or application server (210) Extracts text from documents containing layers and scanned documents as images. a working document reading and page separation module (30),  Course code, course name, from the data coming from the document reading and page separation module (30), A resource course 30 that extracts information on achievement grade, national credit, ECTS, and course content. inference module (40),  Fragmented information from transcripts, course syllabi, and supplementary documents, identifying which area each section belongs to. The source area of ​​the document will be tracked in a single source course record. a resource course that combines and stores this record on the database server (240) deduplication module (50), 35  Course catalog data that is compatible with the target program's language of instruction and curriculum version a teaching language that enables selection and generates a warning for human review in case of uncertainty and curriculum source validation module (80),  For each resource course, the course code, course name, previous human-approved equivalencies, and content are provided. Using proximity signals, a limited candidate is selected instead of the entire target program. 5 A candidate target course generation module (90) that determines a target course list.  The model determined by the data reduction, masking and model orientation layer (280) through the medium, the source course and the candidate target course content, learning outcomes, weekly The subject involves comparing content overlap and rationale in terms of national credit and ECTS. an evidence-based matching assessment module that produces information (100), 10  The application server (210) runs on the microprocessor and uses proof-of-matching. not to accept the output of the evaluation module (100) as a final decision and content overlap threshold, credit / ECTS sufficiency, passing status, and decision-score consistency a rule engine that passes through checks (110),  Situations where multiple source courses correspond to a single target course are considered together. 15 to evaluate, calculate the group national credit and ECTS total and credit-weighted a many-to-one equivalence module structured to produce a success grade (120),  source course, target course, credit, ECTS, grade, content overlap, alternative candidates and presenting warnings to the user client device (200); authorized user accepting the suggestion a 20 that allows it to make, change or reject and finalize the decision. Human review and confirmation interface (130),  finalized decisions, resources used and target courses, elective slots and total used resource courses that combine exempt ECTS credits into a single calculation. preventing reuse and calculating the adaptation semester / year with the exempt ECTS upper limit. a final exemption ledger (140) managing this account status, 25  The document on which the decision is based, source course recording, matching proposal, human intervention, and rule engine (110) result; without storing raw personal data and artificial intelligence without restarting the evaluation server or service connection (230), an audit trail showing the decision status recorded on the database server (240) module (170) 30 It includes.

2. The system is compliant with Request 1 and its feature is that the source academic document set (10) is uploaded to the system. It enables the retrieval of file type, file size, page count, and processability checks. 35 by running and storing the digital summary of the document on the database server (240) 31 a document retrieval and verification module that enables detection of re-upload (20) It includes.

3. The system complies with Request 1 and its characteristic is that it contains documents belonging to more than one student. By defining page ranges in the sets, they will divide the document on a student-by-student basis, and each student will receive 5 pages. Document reading and page reading are structured to create separate source course datasets for each. It contains a separation module (30).

4. The system compliant with Claim 1 is characterized by the fact that sensitive personal information is stored transparently. Personal data that prevents, masks and enables control through numerical aggregation 10 protection submodule (60) and external model service for sources that may contain personal data without sending it, on the in-house model server or on the local machine learning model a data reduction, masking, and model orientation layer configured to operate (280) is included.

5. The system complies with Claim 1, and its features include: target program courses, service courses, and elective courses. course structures and program curriculum from the database server (240) or network interface (260) includes the target curriculum acquisition module (70).

6. The system complies with Claim 1 and its feature is that it only displays successful courses that do not directly match. after the final decisions and the remaining electives in the final exemption book (140) Evaluation based on capacity, automatic placement if capacity is insufficient. It contains a categorical adaptation module (150) that does not produce a warning.

7. The system complies with Claim 1, and its feature is that it provides an electronic printout of the exemption and adaptation report or 25 It includes a reporting and output unit (270) that produces a printable document.

8. The system is compliant with Claim 1, and its feature is; final decisions and final exemption register (140) the unit that produces exemption and adaptation reports and reports and outputs these reports (270) It includes a report production module (160) that is presented through. 30 9. A system that complies with Claim 1, and whose characteristic is that it has been previously validated by humans. storing the equivalencies and giving them as hints in the candidate target course production module (90) It includes an institutional equivalence memory (180) that is available for use. 35 32 10. The system compliant with Claim 1 is characterized by its ability to process documents, acquire target data, and match data. a job sequence management module that runs report generation processes in the background (190) It includes.

11. A system that conforms to claim 1, and its characteristic is that it produces the same result for the same inputs. 5 configured to operate with fixed or deterministic parameters to support It includes the evidence matching evaluation module (100).

12. The system complies with Claim 1, and its characteristic is that ECTS is applied in the absence of national credit data. It will use the value as a weight; if no weight data is available, the success rate will be 10. It will not generate a report and will leave the situation to human review, and a human review warning will be issued. It includes a many-to-one equivalence module (120) structured to create.

13. The system is compliant with Request 1, and its features are: the network interface (260), the application server. (210) a student information system, application management system, university course catalog or 15 An API connection layer that enables data exchange with the Bologna Information System. It includes.

14. The source academic document set (10) received from the user client device (200), 20 courses from different institutions, programs or curriculum sources in higher education computerized recognition, exemption and adaptation assessment within the target program It is an applied method, and its characteristic is;  Document reading and page separation module (30) document processing server (220) or by running on the application server (210), from documents containing a text layer Text extraction and image processing of scanned documents, resulting in 25% of the data obtained. (1003), kept in the file storage unit (250),  According to the source lesson drawing module (40), the document processing server (220), application server (210) and data reduction, masking and model routing layer (280) the model environment or AI evaluation server or service it has specified Using the link (230), course code, course name, grade, national credit, ECTS and course 30 Extraction of content information and storage on the database server (240) (1005),  In accordance with the source course unification module (50), fragmented parts from different documents information on a single source course record on the application server (210) combined and stored on the database server with the source information of each field (240) storage (1007), 35 33  In accordance with the instruction language and curriculum source validation module (80), the application by the server (210) to the language of instruction and curriculum version of the target program suitable course catalog data from the database server (240) or network interface (260) Selection via a system and generation of a human review alert in case of uncertainty. (1009), 5  Stored on the database server (240) by the application server (210) Checking the course recordings for passing status and eligibility for exemption. by protecting national credit and ECTS values ​​as separate areas (1010),  According to the candidate target course production module (90), by the application server (210) Using course data (240) on the database server and previous approved equivalencies, 10 Determining a limited list of candidate target courses for each source course (1011),  According to the evidence matching evaluation module (100), data reduction, masking and The model environment determined by the model guidance layer (280) according to the nature of the data or via AI assessment server or service connection (230), 15 with the resource course that the application server (210) receives from the database server (240) candidate target courses in terms of content, learning outcomes, national credits and ECTS producing content overlap and rationale information by comparison (1012),  The rule engine (110) is on the microprocessor of the application server (210) By running it, the rule set defined in the database server (240) of the generated suggestion Accordingly, content overlap threshold, credit / ECTS sufficiency and decision-score consistency 20 subjected to deterministic controls in terms of (1013),  According to the many-to-one equivalence module (120), by the application server (210) Multiple source courses retrieved from the database server (240) into a single target course together with the evaluation, calculation of the total group national credits and ECTS, and Generating credit-weighted performance scores (1014), 25  In accordance with the human review and confirmation interface (130), the application server (210) source / target course, credit, ECTS, grade, received from (240) database server, Reason, alternative and warning information is transmitted to the user client via the network interface (260). (1015), presented to the device (200),  Acceptance given by the authorized user via the user client device (200), modification 30 or rejection inputs to the application server (210) via the network interface (260) transmission and final decision status of the decisions in the database server (240) storage (1016),  Creation of the final exemption book (140) by the application server (210), finalized decisions, resources used and target courses, elective slots and 35 total exempted ECTS value as single account status in database server (240) 34 preserving and preventing the reuse of used resource lessons (1017),  The total exemptions in the final exemption book (140) by the application server (210) Check the ECTS value against the upper limit defined in the database server (240). marking of the courses to be excluded and excluded (1019), 5  The total of final exempted ECTS credits after the upper limit by the application server (210) calculated and transferred to the adaptation semester / year account and in the database server (240) storage (1020), It includes the steps of the process.

15. The method is in accordance with Request 14 and its feature is; the source academic document set (10) user from client device (200) to application server (210) via network interface (260) file suitability check in accordance with the transmission, document receiving and verification module (20) the step of (1001) to do and store the document in the file storage unit (250) It includes. 15 16. The method is in accordance with Request 14 or 15, and its characteristic is; In the file storage unit (250) The numerical summary of the stored document is provided by the application server (210) calculated and stored on the database server (240) and the same document again This includes the step of identifying the loading through this summary (1002). 20 17. This method complies with Request 14 and its characteristic is: a document containing documents belonging to more than one student. in the sets, the page by the document processing server (220) or application server (210) It includes the step of separating the document on a student basis by determining the intervals (1004).

18. The method is in accordance with Request 14 and its characteristic is; in accordance with the personal data protection submodule (60), Masking or numeric hashing of identity data by the application server (210) by way of protection of documents in the file storage unit (250) and reduced decision data of sources containing personal data is stored on the database server (240). Data reduction by reduction, masking and model orientation layer (280), 30 It includes the step of applying masking and model orientation (1006).

19. The method is in accordance with Claim 14 and its characteristic is; according to the target curriculum acquisition module (70), Application of data regarding target program courses, service courses and elective course structures. by the server (210) from the database server (240) or network interface (260) 35 It includes the step of taking it over (1008). 35 20. The method is in accordance with Claim 14 and its characteristic is; in accordance with the Categorical adaptation module (150), Categorical adaptation by the application server (210) is only for finalized decisions. then and according to the remaining selectable capacity received from the database server (240) It includes the step of doing (1018). 5 21. The method is in accordance with Request 14, and its feature is; Application of the Report generation module (160). By running on server (210), the finalized database server (240) Production of exemption and adaptation report through decisions and final exemption book (140) and includes the step of presenting through the reporting and output unit (270) (1021). 10 22. The method is in accordance with Claim 14 and its feature is; implementation in accordance with the audit trail module (170). document on which the decision is based by the server (210), matching proposal, human intervention and the result of the rule engine (110); artificial intelligence evaluation server or service 15 registered only on the database server (240) without restarting the connection (230). through the decision status and reporting and output unit (270) The presentation includes (1022) processing steps.

23. The method is in accordance with Claim 14 and its characteristic is; In accordance with the institutional equivalence memory (180), 20 by the application server (210) of human-approved equivalencies stored on the database server (240) and candidate target in subsequent evaluations It includes the step of using it in lesson production (1023).

24. The method is compliant with claim 14 and its characteristic is that the database is accessed by the application server (210). According to the target course credit and ECTS information received from the server (240), the target course is 25 Credit / ECTS proficiency requirement based on group totals instead of individual source courses. It includes the auditing step (1014.4).

25. The method is compliant with request 14 or 23, and its characteristic is that it is performed by the application server (210). whether weight data is available or not in the database server (240) source lesson 30 Checked against records; if weighting data is missing, the success grade will be determined. issuing a warning by not producing anything and leaving the situation to human investigation (1014.6) It includes the step.

26. This method complies with Claim 14 and is characterized by its audit trail, raw transcript text, and clear identification. 35 Without storing the number, the decision summary, candidate target registered on the database server (240) 36 creating lesson summaries, implemented controls and warnings (step 1022.3) It includes.

Images