Traditional Chinese medicine culture narrative and intelligent translation propagation method and system oriented to African context

Abstract

The invention provides an African context-oriented traditional Chinese medicine culture narrative and intelligent translation propagation method and system, and relates to the technical field of natural language processing and traditional Chinese medicine internationalization application. According to the method, a traditional Chinese medicine narrative corpus containing a disease view, a body view and a treatment view is constructed, an African context knowledge base fusing religious and health behavior elements is established, and a concept mapping relation is formed; an intermediary narrative representation is constructed on the basis, and unified coding of the traditional Chinese medicine concept and the African local cognitive system is achieved; driving intelligent translation by utilizing the intermediary narrative representation, so that a source language traditional Chinese medicine text is accurately converted into a culture-adaptive target language narrative text; customized propagation content is generated according to scenes such as science popularization, clinical communication and policy propaganda, and self-adaptive updating of the knowledge base and the translation model is realized in combination with audience feedback, so that the understanding degree and acceptability of traditional Chinese medicine cross-culture propagation are improved.

Description

Technical Field

[0001] This invention relates to the fields of natural language processing and international application of traditional Chinese medicine, and in particular to a method and system for the intelligent translation and dissemination of traditional Chinese medicine cultural narratives in an African context. Background Technology

[0002] Currently, the international dissemination of Traditional Chinese Medicine (TCM) mainly relies on human interpreters, general machine translation, and educational materials in a few mainstream languages ​​such as English. Existing systems primarily focus on terminology comparison and sentence-level translation, directly mapping TCM disease names, prescription names, and treatment principles to target language vocabulary, with some platforms supplemented by simple annotations. For TCM dissemination in Africa, it largely depends on English or French intermediary texts, which are then translated a second time by local personnel. There is a lack of systematic technical solutions tailored to African languages ​​and cultures, making it difficult to reach a wider audience, including grassroots patients and traditional practitioners.

[0003] With the development of cross-cultural health communication and narrative medicine research, academia and industry have gradually realized that simple terminology translation is insufficient to support the international understanding of complex concepts such as the holistic view and syndrome differentiation and treatment in traditional Chinese medicine. They have begun to introduce narrative techniques such as storytelling, case studies, and contextualized dialogue. Simultaneously, technologies such as neural network machine translation, domain adaptation, and knowledge-enhanced translation are emerging, enabling translation systems to perform professional text processing by incorporating medical knowledge bases and terminology databases. Intelligent communication platforms for multiple languages ​​and scenarios have become an important direction, but existing work mainly focuses on Chinese-English and Central European languages, with few comprehensive technical approaches specifically targeting African multilingualism and its unique religious beliefs, health concepts, and social structures.

[0004] Current technologies for the international dissemination of Traditional Chinese Medicine (TCM) generally suffer from the following shortcomings: First, there is a lack of mechanisms to model the narrative structure of TCM culture with the African context, failing to achieve conceptual alignment at the levels of disease perspectives, body perspectives, and treatment perspectives, resulting in translated content that is "word-for-word but not logically sound." Second, existing translation systems mostly translate directly sentence by sentence or paragraph, lacking mediating narrative representations oriented towards the target culture, making it difficult to reuse shared semantics across different African languages ​​and support multilingual expansion. Third, the disseminated content is usually a uniform version, failing to dynamically adjust genre and information granularity according to different tasks such as popular science education, clinical communication, and policy promotion, and lacking a closed-loop optimization mechanism based on audience understanding and trust feedback. Therefore, it is necessary to propose an integrated approach encompassing narrative corpus modeling, African contextual knowledge construction, mediating narrative representation generation, intelligent translation and contextualized rewriting, and feedback updates to achieve intelligent translation and dissemination of TCM cultural narratives oriented towards the African context. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, the purpose of this invention is to provide a method and system for the intelligent translation and dissemination of traditional Chinese medicine cultural narratives in the context of African languages. By constructing a cross-cultural narrative semantic layer, the invention achieves systematic alignment of traditional Chinese medicine concepts with the African context, significantly improving the accuracy and cultural compatibility of cross-language communication.

[0006] To achieve the above objectives, the present invention provides the following solution:

[0007] A method for the narrative and intelligent translation of traditional Chinese medicine culture in an African context includes:

[0008] S1. Collect TCM literature and popular science materials to construct a TCM narrative corpus that annotates elements of disease view, body view, and treatment view;

[0009] S2. Based on data on African native languages ​​and health concepts, establish an African contextual knowledge base that includes religious and health behavior elements, and form a conceptual mapping relationship with the aforementioned Chinese medicine narrative corpus;

[0010] S3. Under the constraints of the concept mapping relationship, generate an intermediate narrative representation that uniformly encodes the concepts of traditional Chinese medicine, treatment processes, and African contextual elements, which is used to carry the shared semantics of cross-cultural narrative and translation;

[0011] S4. Using the aforementioned mediating narrative representation, a narrative-driven intelligent translation model for African languages ​​is constructed, enabling the source language TCM text to be transformed into a target language narrative text that combines fidelity and cultural adaptability through the aforementioned mediating narrative representation.

[0012] S5. Set task tags for popular science, clinical communication, and policy promotion for the communication scenarios, and rewrite the target language narrative text and adjust its granularity based on the task tags to obtain customized communication content for the corresponding scenarios.

[0013] S6. Collect audience understanding and trust feedback data, and jointly update the African context knowledge base, the mediated narrative representation, and the intelligent translation model to adaptively optimize the narrative and translation effects of traditional Chinese medicine culture in the African context.

[0014] Preferably, step S1 includes:

[0015] The TCM literature and popular science materials are divided into genres such as case narratives, popular science stories, consultation dialogues and policy texts to obtain a set of TCM narrative units organized by genre.

[0016] Each narrative unit in the TCM narrative unit set is labeled with multi-level tags, including disease naming method, etiology and pathogenesis explanation, description of physical sensations, treatment process, medication basis and prognosis evaluation, to obtain narrative unit labeling data;

[0017] The narrative unit annotation data and corresponding narrative character information are written into the TCM narrative corpus to form the TCM narrative corpus that supports cross-text alignment and narrative structure analysis.

[0018] Preferably, step S2 includes:

[0019] Using a knowledge graph structure, we construct nodes from TCM disease concepts, syndrome names, treatment methods, African local disease names, traditional treatment practices, religious rituals, and daily health behaviors to obtain a set of semantic nodes;

[0020] By constructing causal relationships, class relationships, similarity relationships, and conflict relationships as edges, a semantic edge set is obtained;

[0021] The semantic node set and the semantic edge set are stored as graph structure data to form the African context knowledge base;

[0022] The graph structure data is evaluated for correlation through expert annotation and corpus statistics to obtain the concept mapping relationship, which is then stored in the African context knowledge base.

[0023] Preferably, step S3 includes:

[0024] The mediating narrative representation is divided into multiple slots, including at least the disease development timeline slot, the physical and emotional state slot, the treatment decision and intervention steps slot, and the religious beliefs and social support environment slot.

[0025] Based on the concept mapping relationship and the African context knowledge base, narrative elements corresponding to each slot are extracted from the traditional Chinese medicine narrative corpus and filled into the corresponding slots to obtain structured mediating narrative data.

[0026] The causal and explanatory relationships between each slot are recorded in the form of a structured template or graph structure to form the mediated narrative representation, which is used to support subsequent narrative reorganization and translation generation.

[0027] Preferably, step S4 includes:

[0028] With the support of the TCM narrative corpus and the African context knowledge base, the TCM texts in the source language are segmented into narrative units and semantically analyzed to extract disease-related elements, physical state elements, treatment process elements and cultural context elements, and the mediated narrative representation is generated to obtain the mediated narrative representation data.

[0029] Using the intermediate narrative representation data as an intermediate semantic layer, narrative restructuring and sentence generation are performed on the intermediate narrative representation data according to the grammatical rules, narrative habits and expression paradigms of the target African language, to obtain the target language narrative text that is both faithful and culturally adaptable.

[0030] The narrative-driven intelligent translation model for African languages ​​is an intelligent translation model that uses the mediating narrative representation as the core mediating structure, learns the mapping relationship from the source language to the mediating narrative representation and the mapping relationship from the mediating narrative representation to the target language, and is constrained by the conceptual mapping relationship and the African context knowledge base during the mapping process.

[0031] Preferably, step S5 includes:

[0032] Task tags for popular science, clinical communication, and policy promotion are preset for each communication scenario, and the task tags are matched with the target language narrative text to obtain the task scenario matching results;

[0033] Based on the task scenario matching results, the genre rewriting parameters and information granularity parameters are determined, and the target language narrative text is added to, reorganized, and its language is adjusted based on the genre rewriting parameters and information granularity parameters to obtain the customized dissemination content corresponding to the task tag.

[0034] Specifically, for science popularization tasks, while maintaining the integrity of diagnosis and treatment logic, daily life analogies and risk warnings should be added to reduce the exposure of professional terminology; for clinical communication tasks, diagnostic basis and medication details should be retained and cultural explanations should be appropriately simplified; and for policy promotion tasks, the health benefits of the population and institutional arrangements should be highlighted while controlling information density, terminology visibility and emotional expression intensity.

[0035] Preferably, step S6 includes:

[0036] After the customized communication content is delivered, the audience's understanding accuracy, self-reported credibility score, willingness to accept treatment plan and perception of cultural conflict are collected through questionnaires, interview records or online interaction logs to obtain the audience's understanding and trust feedback data.

[0037] Based on the audience understanding and trust feedback data, a feedback evaluation index system is constructed. The preset target threshold is compared with the actual index value, and update instructions are generated for the African context knowledge base, the mediated narrative representation and the intelligent translation model.

[0038] The updated African context knowledge base, the intermediate narrative representation, and the intelligent translation model are obtained by adjusting the concept weights and relation strengths in the African context knowledge base, correcting the slot settings in the mediated narrative representation that are prone to misunderstanding, and updating the parameters or generation preferences of the intelligent translation model according to the update instructions.

[0039] Preferably, step S4 further includes:

[0040] During the training phase, TCM texts in at least one intercontinental common language and multiple African native languages ​​were selected, and the TCM texts in each language were converted into the aforementioned mediated narrative representations to obtain multilingual mediated narrative training data.

[0041] Based on the multilingual mediated narrative training data, the encoding sub-model and decoding sub-model corresponding to different languages ​​are learned respectively. This enables the narrative-driven intelligent translation model for African languages ​​to achieve multi-directional conversion between different native African languages ​​during the inference stage through the mediated narrative representation constructed once. This reduces the need to build translation models pairwise and improves the efficiency of multilingual expansion.

[0042] Preferably, it further includes:

[0043] Indexes are established for the TCM narrative corpus, the African context knowledge base, the mediated narrative representation, the target language narrative text, the customized communication content, and the audience understanding and trust feedback data, and the data generated in each step are associated and stored with the corresponding indexes.

[0044] When performing steps S1 to S6, the corresponding data is called as input according to the index identifier, and the stored data record is updated with the index identifier as the key after each step is completed, thereby forming a traceable data flow link to support the iterative optimization of the method and the auditing of results.

[0045] A system for the narrative and intelligent translation of traditional Chinese medicine culture in an African context, comprising:

[0046] The TCM narrative corpus construction and element annotation unit is used to collect TCM literature and popular science materials to construct a TCM narrative corpus that annotates elements of disease view, body view, and treatment view.

[0047] The African context knowledge base and concept mapping unit are used to establish an African context knowledge base containing religious and health behavior elements based on African native language and health concept data, and to form a concept mapping relationship with the TCM narrative corpus.

[0048] The mediating narrative representation generation and shared semantic encoding unit is used to generate a mediating narrative representation that uniformly encodes traditional Chinese medicine concepts, treatment processes, and African contextual elements under the constraints of the concept mapping relationship, and is used to carry the shared semantics of cross-cultural narrative and translation.

[0049] The narrative-driven intelligent translation unit for African languages ​​is used to construct a narrative-driven intelligent translation model for African languages ​​using the mediating narrative representation, so that the source language Chinese medicine text is transformed into the target language narrative text with both fidelity and cultural adaptability through the mediating narrative representation.

[0050] The multi-scenario genre rewriting and granularity adjustment unit is used to set task tags for popular science, clinical communication, and policy promotion for dissemination scenarios, and to rewrite and adjust the granularity of the target language narrative text based on the task tags to obtain customized dissemination content for the corresponding scenarios.

[0051] The audience feedback-driven joint adaptive optimization unit is used to collect audience understanding and trust feedback data, and jointly update the African context knowledge base, the mediated narrative representation and the intelligent translation model to adaptively optimize the narrative and translation effects of traditional Chinese medicine culture in the African context.

[0052] The present invention discloses the following technical effects:

[0053] (1) This invention constructs a corpus of TCM narratives that incorporates elements such as disease perspective, body perspective, and treatment perspective, and achieves structured narrative expression based on genre organization and multi-level tagging system. This effectively overcomes the shortcomings of existing technologies that rely solely on terminology comparison and are difficult to convey the holistic view and syndrome differentiation and treatment ideas of TCM. It enables cultural connotations to enter the subsequent translation and dissemination process in the form of calculable narrative units, thereby improving the consistency and reusability of cross-cultural interpretation.

[0054] (2) By establishing an African contextual knowledge base that includes local disease names, traditional therapies, religious rituals and health behaviors, and constructing a conceptual mapping relationship with TCM concepts, this invention solves the problem of the lack of a systematic cross-contextual knowledge alignment mechanism in the prior art, so that TCM narratives can be repositioned and interpreted within the African cultural framework, fundamentally reducing the risk of cultural conflict, misunderstanding and expression breakdown.

[0055] (3) The mediated narrative representation proposed in this invention, as a unified semantic layer across languages ​​and cultures, breaks through the translation path of existing technologies that rely solely on sentence-level or term-level transformations. It enables TCM texts to be first transformed into narrative prototypes that are integrated with the African context, and then generate target language narrative texts, realizing a new path of "aligning semantic structures first and then performing translation", thereby significantly improving the interpretive depth and translation stability of cross-cultural professional texts.

[0056] (4) This invention uses a task tag-driven genre rewriting and granularity adjustment mechanism to automatically adjust the expression mode for different communication scenarios such as popular science, clinical communication, and policy promotion. This solves the defects of existing technologies that use a single text version and cannot dynamically adapt to different scenario needs. It enables the generated content to achieve a higher balance between comprehensibility, professionalism and cultural adaptability, and significantly improves the communication effect and audience acceptance.

[0057] (5) By introducing a closed-loop update mechanism for audience understanding and trust feedback data, this invention achieves dynamic adaptive optimization of the African context knowledge base, mediated narrative representation and intelligent translation model, effectively overcoming the shortcomings of existing technologies in lacking verifiability and sustainable improvement capabilities. This enables the system to continuously iterate with changes in audience cultural preferences and contextual differences in different regions, thereby constructing a long-term evolutionary, multilingual TCM culture dissemination system. Attached Figure Description

[0058] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 A flowchart of the method provided in an embodiment of the present invention;

[0060] Figure 2 This is a schematic diagram of the system structure provided in an embodiment of the present invention. Detailed Implementation

[0061] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0062] The purpose of this invention is to provide a method and system for the intelligent translation and dissemination of traditional Chinese medicine (TCM) cultural narratives in the African context. With mediating narrative representation as the core, it structures TCM knowledge and deeply integrates it with the local African cognitive system, effectively enhancing the understanding and acceptance of cross-cultural translation and dissemination.

[0063] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0064] Figure 1The method flowchart provided in the embodiments of the present invention is as follows: Figure 1 As shown, this invention provides a method for the narrative and intelligent translation of traditional Chinese medicine culture in an African context, characterized by comprising:

[0065] S1. Collect TCM literature and popular science materials to construct a TCM narrative corpus that annotates elements of disease view, body view, and treatment view;

[0066] S2. Based on data on African native languages ​​and health concepts, establish an African contextual knowledge base that includes religious and health behavior elements, and form a conceptual mapping relationship with the TCM narrative corpus.

[0067] S3. Under the constraint of concept mapping relationship, generate a mediating narrative representation that uniformly encodes the concepts of traditional Chinese medicine, treatment process and African context elements, so as to carry the shared semantics of cross-cultural narrative and translation;

[0068] S4. Using mediated narrative representation, construct a narrative-driven intelligent translation model for African languages, so that the source language TCM text can be transformed into a target language narrative text that is both faithful and culturally adaptable through mediated narrative representation.

[0069] S5. Set task tags for popular science, clinical communication, and policy promotion for the communication scenarios. Based on the task tags, rewrite the target language narrative text in terms of genre and adjust the granularity to obtain customized communication content for the corresponding scenarios.

[0070] S6. Collect audience understanding and trust feedback data, and jointly update the African context knowledge base, mediated narrative representation and intelligent translation model to adaptively optimize the narrative and translation effects of traditional Chinese medicine culture in the African context.

[0071] Specifically, step S1 in this embodiment includes:

[0072] First, the data collection process configured in this embodiment is used to acquire traditional Chinese medicine (TCM) literature and popular science materials, which serve as the foundational data for constructing a TCM narrative corpus. The TCM literature mentioned in this embodiment may include publicly published journal articles, textbook chapters, case compilations, etc., while the popular science materials may include TCM brochures, health education manuals, and online popular science articles for the general public. To ensure the representativeness of the corpus, this embodiment sets filtering conditions according to disease category, target audience, and publication time during data collection. For example, at least fifty texts are selected from each of three or more common disease categories. The collected texts are then uniformly formatted and encoded before being stored in the original text database. In this embodiment, the "TCM narrative corpus" refers to a structured corpus set established for subsequent analysis and translation. This corpus set not only preserves the original text content but also synchronously stores narrative-related tag information and character information to support retrieval, statistics, and modeling.

[0073] After the original text is entered into the database, this embodiment classifies the TCM literature and popular science materials according to genre characteristics, resulting in a set of TCM narrative units organized by genre. In this embodiment, a "narrative unit" refers to a text fragment that is semantically relatively complete and can independently describe a diagnosis and treatment process or a health event, such as an outpatient medical record, a doctor-patient dialogue, or a popular science story. Each narrative unit can consist of one or more original text fragments. Specifically, this embodiment uses a combination of rule matching and manual verification to classify the text into four genres: medical record narrative, popular science story, consultation dialogue, and policy text. For example, paragraphs containing information on the onset of illness, diagnosis, treatment, and follow-up are classified as medical record narratives, and doctor-patient question-and-answer content is classified as consultation dialogue. After the genre division is completed, this embodiment annotates each narrative unit in the narrative unit set with multi-level tags, including disease naming method, etiology and pathogenesis explanation, description of physical sensation, treatment process, medication basis and prognosis evaluation. Among them, multi-level tags refer to recording multiple dimensions such as concept category, expression mode and pragmatic function in the same narrative unit. For example, the disease naming method tag records not only the specific name of the disease, but also whether the name is a traditional Chinese medicine term, a modern medical term or a mixed usage. In this embodiment, it is preferred that each narrative unit contains at least six types of tags to form narrative unit annotation data.

[0074] After completing multi-level tagging, this embodiment writes the tagged data of the narrative units and the corresponding narrative role information into the TCM narrative corpus to form a TCM narrative corpus that supports cross-text alignment and narrative structure analysis. In this embodiment, "narrative role information" refers to the different subjects participating in the narrative and their roles, such as doctors, patients, family members, policymakers, or educators, and indicates the speaking position, behavior type, and narrative perspective of each role in the narrative. For example, the party proposing the diagnosis and giving the treatment plan is identified as the doctor role, and the party stating the symptoms and expressing feelings is identified as the patient role. This embodiment establishes a unified record entry for each narrative unit, storing the genre type, original text of the narrative unit, multi-level tags, and narrative role information in the same corpus table in the form of structured fields. This allows for subsequent cross-text alignment analysis between different texts based on disease naming methods, treatment processes, or narrative roles. For example, it can automatically align the disease naming methods and treatment processes of the same disease in different cultural expressions in more than ten medical case narratives, thereby providing directly callable basic data support for subsequent construction of mediated narrative representations and narrative-driven intelligent translation.

[0075] Specifically, in this embodiment, step S2 first collects local expressions and contextual information related to Traditional Chinese Medicine (TCM) based on African native languages ​​and health concepts, which serves as the foundational data for constructing an African contextual knowledge base. The African native languages ​​and health concepts data mentioned in this embodiment may include a summary table of folk disease names recorded in at least three African native languages, interview texts with traditional healers, health education materials published by religious sites such as churches or mosques, and health education manuals compiled by non-governmental organizations. In organizing the above data, this embodiment extracts statements related to diseases, symptoms, treatments, lifestyle behaviors, and religious practices, and manually labels their functions in context, such as marking a phrase as a disease name, a description as a treatment method, or a text as health instruction in a religious ceremony. To facilitate subsequent unified organization, this embodiment uses a "knowledge graph structure" to store and manage this information. A knowledge graph structure refers to a data organization method that represents concepts and their relationships in the form of "points" and "connections," used to support cross-language and cross-cultural concept retrieval and reasoning.

[0076] In this embodiment, after determining to adopt a knowledge graph structure, entries related to TCM disease concepts, syndrome names, and treatment methods, as well as entries related to African native disease names, traditional treatment practices, religious rituals, and daily health behaviors, are constructed as nodes to obtain a semantic node set. In this embodiment, the "semantic node set" refers to a group of points in the knowledge graph used to represent specific concepts or objects. For example, a TCM disease "liver stagnation syndrome" can be abstracted as a node, and a certain African native disease name, a certain herbal decoction therapy, and a certain prayer ritual can be abstracted as different nodes. Each node is accompanied by attribute fields such as name, language, source document number, and concept category. This embodiment further constructs a semantic edge set by using descriptions in the materials and expert experience to construct edges representing causal, class, similarity, and conflict relationships between concepts. For example, when an interview describes a traditional herbal remedy for relieving symptoms similar to "rheumatism" in Traditional Chinese Medicine, this embodiment establishes a similarity or substitution relationship edge between the "rheumatism" node and the traditional herbal remedy node. When a religious text indicates that a certain behavior is considered a cause of disease, this embodiment establishes a causal relationship edge between the behavior node and the corresponding disease node. Through the above steps, this embodiment forms a graph structure data consisting of a semantic node set and a semantic edge set, and stores this graph structure data in a preset data storage, thereby forming an African context knowledge base.

[0077] After the initial construction of the graph structure data, this embodiment further evaluates the relevance of the graph structure data through expert annotation and corpus statistics to generate concept mapping relationships in order to achieve effective integration with the TCM narrative corpus. In this embodiment, "concept mapping relationship" refers to the one-to-one, one-to-many, or many-to-one association established between TCM concepts and African contextual concepts, used to characterize the comparability of disease categories, symptom experiences, or treatment intentions in the two cultural systems. For example, this embodiment invites at least three experts with TCM backgrounds and at least three experts with African traditional medicine backgrounds to manually match a batch of candidate pairs containing more than one hundred TCM disease concepts and more than one hundred African native disease names, giving a similarity level between each pair, and calculating a relevance score for ranking based on the frequency of co-occurrence in the narrative corpus; the higher the relevance score, the more stable the correspondence between the TCM concept and the African concept in actual use. This embodiment combines expert annotation results with corpus statistics to identify the mapping relationship type for concept pairs with high relevance. The mapping information between the corresponding TCM nodes and African nodes (including corresponding types, application scenarios, typical example sentence numbers, etc.) is written into the African context knowledge base so that it can be directly called when generating mediating narrative representations and performing narrative-driven intelligent translation, thereby achieving systematic alignment between TCM narratives and African contexts.

[0078] Furthermore, in this embodiment, step S3 first designs the overall structure of the mediated narrative representation based on the aforementioned concept mapping relationship and the African context knowledge base, enabling it to simultaneously carry TCM concepts, treatment processes, and African context elements. The "mediated narrative representation" described in this embodiment refers to a neutral narrative structure situated between the source language text and the target language text, used to record disease development, physical sensations, treatment decisions, and religious and social environments in a unified manner, thereby forming a cross-culturally shared semantic carrier. To facilitate detailed expression of the above content, this embodiment divides the mediated narrative representation into multiple "slots," where a slot refers to a preset information location used to accommodate a certain type of narrative element, including at least a disease development timeline slot, a physical and emotional state slot, a treatment decision and intervention step slot, and a religious belief and social support environment slot. For example, the disease development timeline slot is used to record the onset time, disease stages, and key turning points, while the religious belief and social support environment slot is used to record narrative content related to religious practices, family support, or community opinion.

[0079] After determining the slot structure, this embodiment extracts narrative elements corresponding to each slot from the TCM narrative corpus based on concept mapping relationships and an African context knowledge base. These elements are then aggregated and filled according to slot type to form structured mediated narrative data. Specifically, when processing a TCM narrative unit, this embodiment first identifies text fragments related to disease naming, symptom evolution, emotional reactions, treatment plans, and religious or family reactions, and determines which slot each fragment fits into. Subsequently, using the aforementioned concept mapping relationships, TCM disease concepts or syndrome names are aligned with African local disease names or traditional understandings, ensuring that the content filled into the slot retains the original TCM meaning while also including corresponding expressions in the African context. For example, in the disease development timeline slot, this embodiment can simultaneously record descriptions of stages such as "acute onset" and "remission period" from a TCM perspective, as well as colloquial descriptions of the same stage in the African local context, thus presenting bidirectional semantic information within a single slot.

[0080] After filling each slot with content, this embodiment records the causal and explanatory relationships between slots in the form of a structured template or graph structure, ultimately forming a mediated narrative representation to support subsequent narrative reorganization and translation generation. The "structured template" in this embodiment refers to arranging the content of each slot sequentially according to a fixed field order, with relationship markers attached between the fields. For example, "a certain religious taboo behavior" is marked as a causal relationship with "symptom aggravation," and "increased family support" is marked as a facilitating relationship with "improved treatment adherence." The "graph structure" treats key elements in each slot as nodes, recording their sequential, transitional, explanatory, or comparative relationships as lines, ensuring the narrative structure remains clear in both time and logic dimensions. Through this method, the mediated narrative representation constructed in this embodiment can be directly invoked in the subsequent narrative reorganization process, flexibly rearranging the narrative order and highlighting different elements according to different target languages ​​and communication scenarios, thereby providing a complete and traceable semantic foundation for narrative-driven intelligent translation.

[0081] Furthermore, in this embodiment, the purpose of step S4 is to utilize the intermediate narrative representation as an intermediate semantic layer, based on the already constructed TCM narrative corpus, African context knowledge base, and intermediate narrative representation, to complete the conversion of the source language TCM text into the target African language narrative text. This embodiment constructs a narrative-driven intelligent translation model for African languages, ensuring that the translation process no longer relies solely on surface-level sentence correspondences, but rather unfolds around a narrative structure jointly composed of TCM disease perspectives, treatment processes, and the local African context. This guarantees that the generated target language narrative text possesses high fidelity and adaptability at both the medical meaning and cultural expression levels.

[0082] In this embodiment, the narrative unit segmentation and semantic analysis of the source language TCM text are first performed with the support of a TCM narrative corpus and an African context knowledge base. "Narrative unit segmentation" refers to dividing continuous text into several segments with complete semantics based on treatment events, symptom changes, or key dialogue nodes. For example, a medical record can be divided into multiple narrative units: chief complaint, present illness, past medical history, treatment plan, and follow-up results. "Semantic analysis" refers to identifying disease-related elements, physical state elements, treatment process elements, and cultural context elements within each narrative unit. For example, "headache recurring for three months" is identified as disease development information, "fear of losing one's job" as emotional state information, "using acupuncture combined with oral decoction" as treatment process information, and "family opposition to acupuncture" as social support environment information. In this embodiment, these elements are filled into the mediating narrative representation according to the aforementioned slot structure, forming mediating narrative representation data, which serves as the unified input for subsequent translation generation.

[0083] After constructing the mediated narrative representation data, this embodiment uses this data as an intermediate semantic layer. Following the grammatical rules, narrative habits, and expression paradigms of the target African language, it performs narrative restructuring and sentence generation on the mediated narrative representation data. "Narrative restructuring" refers to rearranging the presentation order of the disease's cause, symptom development, treatment decisions, and outcome feedback according to common narrative styles in the target culture, rather than directly adopting the word order and text structure of the source language text. For example, in some African contexts, the family and religious background may be emphasized first, followed by the introduction of the disease and treatment process. In sentence generation, this embodiment selects appropriate sentence structures and forms of address based on the grammatical structure and polite expressions of the target language. For example, when expressing a doctor's advice, it uses euphemistic and persuasive language commonly used in the local culture, thereby generating a target language narrative text that is both medically faithful and culturally natural.

[0084] The "Narrative-Driven Intelligent Translation Model for African Languages" described in this embodiment refers to a type of translation model that uses a mediating narrative representation as its core mediating structure. It learns the mapping relationship between the source language TCM text and the mediating narrative representation, as well as the mapping relationship between the mediating narrative representation and the target African language narrative text. In this type of translation model, the former mapping process focuses on learning how to correctly extract and organize narrative elements from the source language text, while the latter mapping process focuses on learning how to retell these narrative elements according to the requirements of the target language and culture. This embodiment is consistently constrained by conceptual mapping relationships and an African contextual knowledge base during the mapping process. For example, when a TCM syndrome appears in the mediating narrative representation, the translation model will prioritize referencing the corresponding African native disease name or symptom combination in the conceptual mapping relationship, and combine this with information on common narrative methods for that disease name in the African contextual knowledge base, thereby avoiding situations where a literal translation is not understood by the local audience.

[0085] To ensure the narrative-driven intelligent translation model has cross-language scalability, this embodiment selects TCM texts in at least one intercontinental common language and multiple African native languages ​​during the training phase. The TCM texts in each language are converted into mediated narrative representations, resulting in multilingual mediated narrative training data. Based on this, this embodiment learns "encoding sub-models" and "decoding sub-models" for different languages. The encoding sub-model refers to the sub-process of learning how to map TCM narrative text in a source language to a mediated narrative representation, while the decoding sub-model refers to the sub-process of learning how to generate narrative text in a target language from the mediated narrative representation. For example, three encoding sub-processes and three decoding sub-processes can be learned for one common language and two African native languages ​​respectively. This ensures that during the inference phase, when any supported source language text is input, only one encoding is needed to obtain the mediated narrative representation, and then the target text is generated through the corresponding language's decoding sub-process. This eliminates the need to train a translation model separately for each language pair, significantly reducing the need to build translation models pairwise and improving multilingual scalability.

[0086] Optionally, in this embodiment, step S5 first presets three types of task tags—popular science, clinical communication, and policy promotion—based on different purposes of disseminating traditional Chinese medicine culture in an African context, and then performs scene identification and matching on the target language narrative text accordingly. The "task tags" mentioned in this embodiment refer to marking information used to identify the main application scenarios of the target language narrative text. For example, text used for public health education is marked as a popular science task, text used for doctor-patient communication is marked as a clinical communication task, and text used for departmental presentations or legal interpretations is marked as a policy promotion task. In specific implementation, this embodiment can determine the target language narrative text based on the text source channel, target audience type, and usage environment. For example, translation results from clinic follow-up records are matched as clinical communication tasks, materials used for village presentations are matched as popular science tasks, and materials used as attachments to health department documents are matched as policy promotion tasks, thereby obtaining task scene matching results and providing a basis for subsequent targeted text adjustments.

[0087] After obtaining the task scenario matching results, this embodiment determines the corresponding genre rewriting parameters and information granularity parameters based on different task tags, and adds, deletes, reorganizes, and adjusts the terminology of the target language narrative text to generate customized dissemination content corresponding to the task tags. The "genre rewriting parameters" mentioned in this embodiment refer to a set of control elements used to adjust the text's style and organization, such as whether to use a storytelling approach, whether to present it in a question-and-answer format, and whether to use formal terminology or colloquial expressions. The "information granularity parameters" refer to elements used to control the level of detail in the information, such as whether to only give the disease name or provide a detailed explanation of the diagnostic criteria for a diagnosis, and whether to only mention the treatment method or list the specific drug name and dosage range for drug information. For example, in the context of popular science tasks, this embodiment guides the text to use more everyday narratives and metaphors through genre rewriting parameters, replaces professional terms with more colloquial expressions, and controls the level of detail through information granularity parameters, retaining only symptoms and general treatment paths that are closely related to public understanding; in the context of clinical communication tasks, this embodiment retains diagnostic basis and medication details, but appropriately simplifies the cultural explanations to make the text both professional and not to increase the communication burden.

[0088] Building upon the above, this embodiment further differentiates the intensity of emotional expression and group-oriented information based on different task tags to generate customized communication content. The "customized communication content" mentioned in this embodiment refers to the target language text result that is reorganized and re-expressed for a specific scenario while maintaining the logic of TCM diagnosis and treatment and the semantics of the mediating narrative. For example, for science popularization tasks, this embodiment adds everyday analogies and risk warnings without changing the causes of diseases and treatment principles, while reducing the frequency of unnecessary professional terms, making it easier for the general audience to understand and adopt the advice. For clinical communication tasks, this embodiment emphasizes individual conditions, treatment plans, and precautions, enhancing patient trust through appropriate reassuring expressions and weakening macro-policies and group-oriented statements. For policy promotion tasks, this embodiment highlights group health benefits, institutional arrangements, and behavioral norms, compressing individual case information into typical examples, and controlling information density, terminology visibility, and emotional expression intensity to make the text suitable for use in conference presentations, announcements, or guidance documents, thereby achieving precise dissemination of the same mediating narrative in different application scenarios.

[0089] As an optional implementation, in this embodiment, step S6 first involves organizing feedback collection based on the audience's understanding and trust levels after the customized dissemination content is actually delivered. This embodiment obtains feedback information from different dissemination scenarios through three channels: questionnaires, key interview records, and online interaction logs. For example, short paper questionnaires are distributed after village presentations; patients' understanding of the explained content is recorded during outpatient follow-ups; and the audience's reading time, commenting tendencies, and willingness to forward the disseminated content are recorded on online platforms. This embodiment unifies the feedback information from these different sources into several indicators, including comprehension accuracy rate, self-reported credibility score, willingness to accept treatment plans, and perception of cultural conflict. "Audience understanding and trust feedback data" refers to a comprehensive set of feedback data composed of these indicators closely related to the degree of understanding and trust, used to reflect the actual effect of the current dissemination content on different groups. For example, the comprehension accuracy rate can be estimated by the proportion of correct answers to key questions, and the self-reported credibility score can be recorded using a grading system from one to ten.

[0090] After obtaining audience understanding and trust feedback data, this embodiment further constructs a feedback evaluation index system to uniformly assess the communication effects across different dimensions. The "feedback evaluation index system" described in this embodiment refers to an index structure that organizes various feedback indicators within the same framework, sets target values, and compares deviations. For example, this embodiment can set a target threshold of over 80% for comprehension accuracy, a target threshold of over 7 points for self-reported credibility, and a lower target upper limit for cultural conflict perception. By comparing the actually observed index values ​​with the pre-set target thresholds, this embodiment can determine which disease topics, audiences, or communication scenarios exhibit insufficient understanding, insufficient trust, or prominent cultural conflicts. Based on this, this embodiment generates "update instructions," which are adjustment requirements for the African context knowledge base, mediated narrative representation, and intelligent translation model. For example, it may instruct the weakening of certain metaphors, the enhancement of certain background explanations, or the change of word choice in specific religious contexts.

[0091] This embodiment adjusts the concept weights and relation strengths in the African context knowledge base, the slot settings in the mediation narrative representation, and the parameters or generation preferences of the intelligent translation model item by item according to the generated update instructions to obtain the updated knowledge and model state. The "concept weight" mentioned in this embodiment refers to a numerical value reflecting the importance of a concept in a specific context. For example, when feedback shows that a traditional disease name is easier for the audience to understand, the weight of that disease name in the knowledge base can be appropriately increased. "Relationship strength" refers to a measure of the closeness of the association between different concepts. For example, when a religious ritual and treatment adherence show a strong connection in multiple feedbacks, the relationship strength between the two can be correspondingly increased. This embodiment also corrects the slot settings in the mediation narrative representation that are prone to misunderstanding based on the feedback results, such as adding more detailed description options for slots related to emotional states or cultural taboos. Meanwhile, this embodiment adjusts the "generation preference" of the intelligent translation model. Generation preference refers to the model's tendency to choose among multiple legal expressions, such as the trade-off between professional terminology and colloquial expressions, and the choice between direct and indirect expressions. By appropriately biasing towards expressions that are more easily accepted by the target group based on feedback, the subsequently generated target language narrative text is more in line with the local audience's understanding habits and trust expectations while maintaining the stability of the medical meaning.

[0092] Furthermore, in this embodiment, to ensure the location and traceability of the data generated in each step during subsequent use, corresponding index identifiers are established for the aforementioned data when constructing the TCM narrative corpus, the African context knowledge base, the mediating narrative representation, the target language narrative text, the customized communication content, and the audience understanding and trust feedback data. The "index identifier" mentioned in this embodiment refers to the marking information used to uniquely identify a data entry or a version of a data set. It can adopt a numbering format composed of numbers. For example, a batch of TCM narrative corpus is set as Corpus Batch No. 1, the corresponding African context knowledge base version is set as Knowledge Version No. 1, and the mediating narrative representation, target language narrative text, and customized communication content generated in the same batch are each assigned a subdivided number based on the prefix No. 1. In this embodiment, during data storage, each data entry and its index identifier are recorded together in a preset data storage structure, so that the corresponding data can be quickly located through the index identifier during subsequent retrieval.

[0093] In this embodiment, during the execution of steps S1 to S6, index identifiers are used as clues for data retrieval and updates. Before entering a certain step, this embodiment retrieves data corresponding to the index identifier from the Traditional Chinese Medicine narrative corpus, the African context knowledge base, or the mediating narrative representation, based on the index identifier of the current processing object. For example, when generating target language narrative text for a certain patient group, the traditional Chinese medicine narrative corpus and the African context knowledge base version of the same batch are retrieved based on the first batch index of the corpus corresponding to that group. After completing this step, this embodiment uses the same index identifier as the key to write the newly generated mediating narrative representation, target language narrative text, or customized communication content into the storage structure, realizing a one-to-one correspondence from the original corpus to the translation result. Similarly, when collecting audience understanding and trust feedback data in a certain scenario, this embodiment associates and records the feedback data with the index identifier corresponding to the customized communication content delivered at that time, so that the specific source can be accurately traced when updating the knowledge base or translation model in the future.

[0094] This embodiment, through the aforementioned method, establishes a "data flow link" that runs from corpus collection, contextual modeling, narrative representation construction, translation generation, scenario-based dissemination, to feedback updates, all under the guidance of index identifiers. This "data flow link" refers to a logical chain that uses index identifiers as the main thread to systematically record the transmission relationships between inputs and outputs at each step. This ensures that each translation and dissemination result can be traced back to the corpus version used, the knowledge base status, and the content of the intermediate narrative representation. For example, when an African native language community experiences a significant increase in comprehension accuracy and a significant decrease in cultural conflict perception over a period of time, this embodiment can use index identifiers to reverse-engineer the African context knowledge base version and translation generation strategy used at that time, and solidify these experiences into a new round of parameter settings. Conversely, when feedback in a scenario shows a low trust score, this embodiment can also trace along the data flow link to the specific semantic mapping or narrative settings that caused the problem, allowing for targeted adjustments and supporting iterative optimization and result auditing of the method.

[0095] Figure 2 This is a schematic diagram of the system structure provided in the embodiments of the present invention, such as... Figure 2 As shown, this invention also provides a system for the narrative and intelligent translation of traditional Chinese medicine culture in an African context, comprising:

[0096] The TCM narrative corpus construction and element annotation unit is used to collect TCM literature and popular science materials to construct a TCM narrative corpus that annotates elements of disease view, body view, and treatment view.

[0097] The African context knowledge base and concept mapping unit are used to establish an African context knowledge base containing religious and health behavior elements based on African native language and health concept data, and to form a concept mapping relationship with the TCM narrative corpus.

[0098] The mediating narrative representation generation and shared semantic encoding unit is used to generate a mediating narrative representation that uniformly encodes traditional Chinese medicine concepts, treatment processes, and African contextual elements under the constraints of the concept mapping relationship, and is used to carry the shared semantics of cross-cultural narrative and translation.

[0099] The narrative-driven intelligent translation unit for African languages ​​is used to construct a narrative-driven intelligent translation model for African languages ​​using the mediating narrative representation, so that the source language Chinese medicine text is transformed into the target language narrative text with both fidelity and cultural adaptability through the mediating narrative representation.

[0100] The multi-scenario genre rewriting and granularity adjustment unit is used to set task tags for popular science, clinical communication, and policy promotion for dissemination scenarios, and to rewrite and adjust the granularity of the target language narrative text based on the task tags to obtain customized dissemination content for the corresponding scenarios.

[0101] The audience feedback-driven joint adaptive optimization unit is used to collect audience understanding and trust feedback data, and jointly update the African context knowledge base, the mediated narrative representation and the intelligent translation model to adaptively optimize the narrative and translation effects of traditional Chinese medicine culture in the African context.

[0102] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0103] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for the narrative and intelligent translation of traditional Chinese medicine culture in an African context, characterized by: include: S1. Collect TCM literature and popular science materials to construct a TCM narrative corpus that annotates elements of disease view, body view, and treatment view; S2. Based on data on African native languages ​​and health concepts, establish an African contextual knowledge base that includes religious and health behavior elements, and form a conceptual mapping relationship with the aforementioned Chinese medicine narrative corpus; S3. Under the constraints of the concept mapping relationship, generate an intermediate narrative representation that uniformly encodes the concepts of traditional Chinese medicine, treatment processes, and African contextual elements, which is used to carry the shared semantics of cross-cultural narrative and translation; S4. Using the aforementioned mediating narrative representation, a narrative-driven intelligent translation model for African languages ​​is constructed, enabling the source language TCM text to be transformed into a target language narrative text that combines fidelity and cultural adaptability through the aforementioned mediating narrative representation. S5. Set task tags for popular science, clinical communication, and policy promotion for the communication scenarios, and rewrite the target language narrative text and adjust its granularity based on the task tags to obtain customized communication content for the corresponding scenarios. S6. Collect audience understanding and trust feedback data, and jointly update the African context knowledge base, the mediated narrative representation, and the intelligent translation model to adaptively optimize the narrative and translation effects of traditional Chinese medicine culture in the African context.

2. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context as described in claim 1, characterized in that, Step S1 includes: The TCM literature and popular science materials are divided into genres such as case narratives, popular science stories, consultation dialogues and policy texts to obtain a set of TCM narrative units organized by genre. Each narrative unit in the TCM narrative unit set is labeled with multi-level tags, including disease naming method, etiology and pathogenesis explanation, description of physical sensations, treatment process, medication basis and prognosis evaluation, to obtain narrative unit labeling data; The narrative unit annotation data and corresponding narrative character information are written into the TCM narrative corpus to form the TCM narrative corpus that supports cross-text alignment and narrative structure analysis.

3. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context as described in claim 1, characterized in that, Step S2 includes: Using a knowledge graph structure, we construct nodes from TCM disease concepts, syndrome names, treatment methods, African local disease names, traditional treatment practices, religious rituals, and daily health behaviors to obtain a set of semantic nodes; By constructing causal relationships, class relationships, similarity relationships, and conflict relationships as edges, a semantic edge set is obtained; The semantic node set and the semantic edge set are stored as graph structure data to form the African context knowledge base; The graph structure data is evaluated for correlation through expert annotation and corpus statistics to obtain the concept mapping relationship, which is then stored in the African context knowledge base.

4. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context according to claim 1, characterized in that, Step S3 includes: The mediating narrative representation is divided into multiple slots, including at least the disease development timeline slot, the physical and emotional state slot, the treatment decision and intervention steps slot, and the religious beliefs and social support environment slot. Based on the concept mapping relationship and the African context knowledge base, narrative elements corresponding to each slot are extracted from the traditional Chinese medicine narrative corpus and filled into the corresponding slots to obtain structured mediating narrative data. The causal and explanatory relationships between each slot are recorded in the form of a structured template or graph structure to form the mediated narrative representation, which is used to support subsequent narrative reorganization and translation generation.

5. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context as described in claim 1, characterized in that, Step S4 includes: With the support of the TCM narrative corpus and the African context knowledge base, the TCM texts in the source language are segmented into narrative units and semantically analyzed to extract disease-related elements, physical state elements, treatment process elements and cultural context elements, and the mediated narrative representation is generated to obtain the mediated narrative representation data. Using the intermediate narrative representation data as an intermediate semantic layer, narrative restructuring and sentence generation are performed on the intermediate narrative representation data according to the grammatical rules, narrative habits and expression paradigms of the target African language, to obtain the target language narrative text that is both faithful and culturally adaptable. The narrative-driven intelligent translation model for African languages ​​is an intelligent translation model that uses the mediating narrative representation as the core mediating structure, learns the mapping relationship from the source language to the mediating narrative representation and the mapping relationship from the mediating narrative representation to the target language, and is constrained by the conceptual mapping relationship and the African context knowledge base during the mapping process.

6. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context according to claim 1, characterized in that, Step S5 includes: Task tags for popular science, clinical communication, and policy promotion are preset for each communication scenario, and the task tags are matched with the target language narrative text to obtain the task scenario matching results; Based on the task scenario matching results, the genre rewriting parameters and information granularity parameters are determined, and the target language narrative text is added to, reorganized, and its language is adjusted based on the genre rewriting parameters and information granularity parameters to obtain the customized dissemination content corresponding to the task tag. Specifically, for science popularization tasks, while maintaining the integrity of diagnosis and treatment logic, daily life analogies and risk warnings should be added to reduce the exposure of professional terminology; for clinical communication tasks, diagnostic basis and medication details should be retained and cultural explanations should be appropriately simplified; and for policy promotion tasks, the health benefits of the population and institutional arrangements should be highlighted while controlling information density, terminology visibility and emotional expression intensity.

7. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context as described in claim 1, characterized in that, Step S6 includes: After the customized communication content is delivered, the audience's understanding accuracy, self-reported credibility score, willingness to accept treatment plan and perception of cultural conflict are collected through questionnaires, interview records or online interaction logs to obtain the audience's understanding and trust feedback data. Based on the audience understanding and trust feedback data, a feedback evaluation index system is constructed. The preset target threshold is compared with the actual index value, and update instructions are generated for the African context knowledge base, the mediated narrative representation and the intelligent translation model. The updated African context knowledge base, the intermediate narrative representation, and the intelligent translation model are obtained by adjusting the concept weights and relation strengths in the African context knowledge base, correcting the slot settings in the mediated narrative representation that are prone to misunderstanding, and updating the parameters or generation preferences of the intelligent translation model according to the update instructions.

8. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context according to claim 1, characterized in that, Step S4 also includes: During the training phase, TCM texts in at least one intercontinental common language and multiple African native languages ​​were selected, and the TCM texts in each language were converted into the aforementioned mediated narrative representations to obtain multilingual mediated narrative training data. Based on the multilingual mediated narrative training data, the encoding sub-model and decoding sub-model corresponding to different languages ​​are learned respectively. This enables the narrative-driven intelligent translation model for African languages ​​to achieve multi-directional conversion between different native African languages ​​during the inference stage through the mediated narrative representation constructed once. This reduces the need to build translation models pairwise and improves the efficiency of multilingual expansion.

9. The method for narrative and intelligent translation of traditional Chinese medicine culture in an African context according to claim 1, characterized in that, Also includes: Indexes are established for the TCM narrative corpus, the African context knowledge base, the mediated narrative representation, the target language narrative text, the customized communication content, and the audience understanding and trust feedback data, and the data generated in each step are associated and stored with the corresponding indexes. When performing steps S1 to S6, the corresponding data is called as input according to the index identifier, and the stored data record is updated with the index identifier as the key after each step is completed, thereby forming a traceable data flow link to support the iterative optimization of the method and the auditing of results.

10. A system for the narrative and intelligent translation of traditional Chinese medicine culture in an African context, characterized in that: include: The TCM narrative corpus construction and element annotation unit is used to collect TCM literature and popular science materials to construct a TCM narrative corpus that annotates elements of disease view, body view, and treatment view. The African context knowledge base and concept mapping unit are used to establish an African context knowledge base containing religious and health behavior elements based on African native language and health concept data, and to form a concept mapping relationship with the TCM narrative corpus. The mediating narrative representation generation and shared semantic encoding unit is used to generate a mediating narrative representation that uniformly encodes traditional Chinese medicine concepts, treatment processes, and African contextual elements under the constraints of the concept mapping relationship, and is used to carry the shared semantics of cross-cultural narrative and translation. The narrative-driven intelligent translation unit for African languages ​​is used to construct a narrative-driven intelligent translation model for African languages ​​using the mediating narrative representation, so that the source language Chinese medicine text is transformed into the target language narrative text with both fidelity and cultural adaptability through the mediating narrative representation. The multi-scenario genre rewriting and granularity adjustment unit is used to set task tags for popular science, clinical communication, and policy promotion for dissemination scenarios, and to rewrite and adjust the granularity of the target language narrative text based on the task tags to obtain customized dissemination content for the corresponding scenarios. The audience feedback-driven joint adaptive optimization unit is used to collect audience understanding and trust feedback data, and jointly update the African context knowledge base, the mediated narrative representation and the intelligent translation model to adaptively optimize the narrative and translation effects of traditional Chinese medicine culture in the African context.

Images