Translation system for natural medicinal material domain-specific knowledge

Abstract

The present application relates to a translation system for natural medicinal material domain-specific knowledge. The system comprises: a front-end program, a first learning model, a search engine and a natural medicinal material domain-specific knowledge base. On the basis of performing coreference primary term annotation, and coreference annotation of standard translations in multiple languages on natural medicinal material terms in the natural medicinal material domain-specific knowledge base, the search engine is used to search in the natural medicinal material domain-specific knowledge base for a natural medicinal material term extracted by the first learning model from text to be translated, the first learning model generates a translated document in an original format on the basis of a search result, and finally, the front-end program displays a translation result in a target language in a mode matching a display mode of a user interaction interface. By means of the translation system in the present application, a user can conveniently and intelligently perform accurate and standardized translation on related terms and knowledge in the domain of natural medicinal materials on the basis of requirements of the user.

Description

A translation system for domain-specific knowledge of natural medicinal materials Technical Field

[0001] The present application relates to the field of natural medicinal material information processing and application, and more specifically, to a translation system for natural medicinal material domain-specific knowledge. Background Art

[0002] Natural medicinal materials (NMMs) have long been recognized as a powerful reservoir of therapeutic agents, their importance reflected in the diversity and biological relevance of the compounds they produce. These compounds play a key role in addressing a wide range of pathological conditions, covering a wide range from infectious diseases to cancer, and continue to serve as a rich source of new drug leads. In addition, NMMs have an extensive history of clinical application around the world, such as in China, India, and the Arab world, demonstrating their enduring relevance in the global healthcare landscape. Standardization of natural medicinal material information and ease of access to this standardized knowledge are key steps to accelerate the pace of research and discovery in this field worldwide.

[0003] However, due to the inherent complexity of natural medicinal materials, even if the species and medicinal parts are the same, but only the processing methods differ, they actually correspond to different natural medicinal materials. However, since their names are not strictly distinguished, and there is no systematic and standardized nomenclature, using commonly used translation tools or translating them online often only yields one-sided or even incorrect translations. For example, when trying to translate the term "ephedra" into English, a variety of different answers may be obtained, including "Ma Huang," "Ephedra," and "Ephedrae Herba." In fact, the "Chinese Pharmacopoeia (2020 Edition)" records in detail that the term "ephedra" refers to natural medicinal materials derived from several different species, including Ephedra sinica (plant grass ephedra), Ephedra intermedia (plant medium ephedra) or Ephedra equisetina (plant horsetail ephedra), but the current naming conventions often tend to classify these different species as the general term "ephedra". Therefore, the inaccuracy or even incorrectness caused by the translation of natural medicinal material terms and knowledge according to the current naming conventions may cause great obstacles to scientific research, and will inevitably undermine the reliability and validity of academic conclusions, and even hinder research progress in this field.

[0004] It can be seen that the current existing technology has not yet found a convenient and intelligent translation tool that can provide various users, including professionals in the field, with accurate and standardized cross-language translation of terms and knowledge related to the field of natural medicinal materials.

[0005] Summary of the Invention

[0006] This application is proposed to solve the above-mentioned problems existing in the prior art. This application intends to provide a translation system for natural medicinal material domain knowledge, which can enable various users, including professionals in the field, to conveniently and intelligently translate the terms and knowledge related to the natural medicinal material field accurately and standardizedly.

[0007] According to the first solution of the present application, a translation system for natural medicinal material domain knowledge is provided, comprising a front-end program, a first learning model, a search engine, and a natural medicinal material domain knowledge base, wherein the front-end program is configured to receive a source language text to be translated containing natural medicinal material terms and a designation of a target language input by a user through a user interaction interface; obtain a translation result containing a standard translation of the natural medicinal material terms in the target language by parsing the translation document in the original format from the first learning model; and present the translation result in the target language on the user interaction interface in a manner that matches the display mode of the user interaction interface. The first learning model is configured to: extract natural medicinal material terms from the text to be translated and transmit them to the search engine; utilize at least one multilingual translation model to generate a translation document in the original format based on the search results returned by the search engine, wherein the translation document in the original format contains the coreference subject of the natural medicinal material terms and the standard translation of the natural medicinal material terms in the target language. The search engine is configured to search the natural medicinal material domain-specific knowledge base for the natural medicinal material term and return search results, wherein the search results include a coreference subject of the natural medicinal material term and a standard translation of the natural medicinal material term in a target language. The natural medicinal material domain-specific knowledge base is configured such that each natural medicinal material term stored in the natural medicinal material domain-specific knowledge base has a coreference subject annotation and a coreference annotation of a standard translation in multiple languages.

[0008] Using the translation system for natural medicinal material domain knowledge according to each embodiment of the present application, the front-end program, the first learning model, the search engine and the natural medicinal material domain knowledge base are jointly applied. First, on the basis of extracting the natural medicinal material terms involved in the text to be translated using the first learning model, the natural medicinal material terms in the natural medicinal material domain knowledge base have the feature of coreference annotation, so that the search engine can search the coreference graph or the dictionary established based on the calculation of the coreference graph, thereby returning the search results containing the coreference of the natural medicinal material term and the standard translation of the target language of the natural medicinal material term; then, the first learning model can generate the original format translation document containing the coreference of the natural medicinal material term and the standard translation of the target language of the natural medicinal material term using different multilingual translation models; finally, the front-end program parses the translation document and provides different modes of translation results on the user interface according to the user's requirements. Thus, using the translation system of the present application, users can conveniently and intelligently translate the terms and knowledge related to the natural medicinal material field accurately and standardizedly according to their needs.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0010] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention, as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In the drawings, which are not necessarily drawn to scale, like reference numerals may describe similar components in different views. The drawings illustrate various embodiments generally by way of example and not limitation, and together with the description and claims, serve to illustrate the disclosed embodiments. Where appropriate, like reference numerals are used throughout the drawings to refer to like or similar parts. Such embodiments are illustrative and are not intended to be exhaustive or exclusive of the embodiments of the present apparatus or method.

[0012] FIG1 shows a schematic diagram of the structure of a translation system for natural medicinal material domain-specific knowledge according to an embodiment of the present application.

[0013] FIG2 shows a schematic diagram of a user interaction interface according to an embodiment of the present application.

[0014] FIG3 shows a schematic diagram of a first user operation according to an embodiment of the present application.

[0015] FIG4( a ) shows a schematic diagram of displaying a translated document in an original format on a user interaction interface according to an embodiment of the present application.

[0016] FIG4( b ) shows a schematic diagram of displaying a translated document in an original format on a user interaction interface according to an embodiment of the present application.

[0017] FIG5 shows a schematic diagram of user evaluation according to an embodiment of the present application. DETAILED DESCRIPTION

[0018] In order to enable those skilled in the art to better understand the technical solution of the present application, the present application is described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present application are further described in detail below in conjunction with the accompanying drawings and specific embodiments, but are not intended to limit the present application.

[0019] The words "first", "second" and similar terms used in this application do not indicate any order, quantity or importance, but are only used to distinguish. Words such as "include" or "comprise" mean that the elements before the word include the elements listed after the word, and do not exclude the possibility of also including other elements. The execution order of the various steps in the method described in this application with reference to the accompanying drawings is not intended to be limiting. As long as the logical relationship between the various steps is not affected, several steps can be integrated into a single step, a single step can be decomposed into multiple steps, and the execution order of the various steps can be changed according to specific needs.

[0020] It should also be understood that the term "and / or" in this application is merely a description of the relationship between associated objects, indicating that three possible relationships exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. Furthermore, the character " / " in this application generally indicates that the associated objects are in an "or" relationship.

[0021] According to an embodiment of the present application, a translation system for natural medicinal material domain knowledge is provided. FIG1 shows a schematic diagram of the composition structure of the translation system for natural medicinal material domain knowledge according to an embodiment of the present application, and FIG2 shows a schematic diagram of the user interaction interface according to an embodiment of the present application.

[0022] As shown in FIG1 , the translation system 100 at least includes a front-end program 101 , a first learning model 102 , a search engine 103 , and a natural medicinal material domain-specific knowledge base 104 .

[0023] Among them, the natural medicinal material domain knowledge base 104 stores structured and standardized natural medicinal material terms and knowledge, and for agricultural NMM (Agricultural NMM, ANMM) and processed NMM (Processed NMM, PNMM), both are named using the natural medicinal material systematic nomenclature created by the inventor of this application. Just as an example, the natural medicinal material systematic nomenclature is composed of a triplet of [natural medicinal material ID, natural medicinal material systematic name, natural medicinal material common name], wherein the natural medicinal material ID, natural medicinal material systematic name and natural medicinal material common name are all globally unique, so agricultural natural medicinal materials and derived processed natural medicinal materials, as well as processed natural medicinal materials using different processing methods, can be distinguished from each other. Furthermore, the natural medicinal material systematic name of agricultural natural medicinal materials at least includes the species origin and medicinal part; the natural medicinal material systematic name of processed natural medicinal materials at least includes the species origin, medicinal part and processing method, and the species origin is expressed in Latin. In other embodiments, the natural medicinal material systematic names of agricultural NMM and processed NMM may further include elements such as special descriptions.

[0024] Taking the natural medicinal material "Artemisia annua" as an example, in the natural medicinal material domain knowledge base 104, its agricultural natural medicinal material NMM ID is NMM-0001, and the corresponding NMMSN is Artemisia annual Part-aerial, NMMSN-zh is the aboveground part of Artemisia annua, NMMGN is Qing-hao, and NMMGN-zh is Artemisia annua. That is, the [natural medicinal material ID, natural medicinal material system name, natural medicinal material common name] triple is [NMM-0001, Artemisla annual Part-aerial, Qing-hao]. The NMM ID of a processed natural medicinal material "Artemisia annua" is NMM-0002, the corresponding NMMSN is Artemisla annual Part-aerial Segmented, NMMSN-zh is the segmented aerial part of Artemisia annua, NMMGN is Qing-hao-duan, and NMMGN-zh is Artemisia annua segment. That is, the [natural medicinal material ID, natural medicinal material system name, natural medicinal material common name] triplet is [NMM-0002, Artemisla annual Part-aerial Segmented, Qing-hao-duan].

[0025] In addition, the natural medicinal material domain-specific knowledge base 104 is further configured such that each natural medicinal material term stored in the natural medicinal material domain-specific knowledge base 104 has a coreference annotation of a coreference subject and a coreference annotation of a standard translation in multiple languages.

[0026] It should be noted that in the natural medicinal materials domain knowledge base 104, the storage of terminology and knowledge adopts natural medicinal materials knowledge in a customized multilingual Markdown (MLMD) format. MLMD is a new, customized Markdown syntax designed specifically for multilingual documents. It can organize and embed multiple language content in parallel in a single document to achieve parallel comparison, unified management and optimized display. This method not only makes document writing and editing more convenient, but also provides a structured solution for processing and analyzing multilingual data. MLMD has complete grammatical rules, and co-reference annotation is one of its important core grammars.

[0027] The core linguistic principle of coreference annotation grammar is coreference agreement. Specifically, when we understand a word in a language to refer to a specific concept or entity, we understand that the word essentially corefers with that specific concept or entity. The coreference annotation grammar specifically designed in MLMD is designed to help users easily annotate coreference relationships in text, facilitating human understanding and, in machine translation and text processing, enabling machines to better understand the meaning of text.

[0028] The specific syntax of coreference annotation in MLMD is as follows:

[0029] Use the double bracket [[]] syntax. Basic coreference annotations include the following two forms:

[0030] 1. When a word and its reference are consistent: [[word reference]]

[0031] 2. When a word and its reference are inconsistent: [[word reference|word display]]

[0032] According to the above grammar, after coreference annotation of "Shennong", the following three sentences can be generated:

[0033] (1) Shennong is considered by the Chinese to be the founder of medicine.

[0034] (2) Shennong is considered by the Chinese to be the founder of medicine.

[0035] (3)[[Shennong|Shennong]]is regarded as the founder of medicine and pharmacy by Chinese people.

[0036] This shows that Shennong serves as the subject of the concept "Shennong," while Yandi and Shennong serve as coreferences of the concept. In this case, Shennong can be called the coreference subject of all coreferences of the concept "Shennong." Through the above coreference annotation, even those without relevant knowledge can understand that these three sentences are actually referring to the same person, "Shennong."

[0037] In the natural medicinal material knowledge converted into the multilingual Markdown format, the natural medicinal material terms can also use the above-mentioned coreference annotation syntax to coreference annotate with their corresponding coreference subject. When annotating the natural medicinal material terms corresponding to each natural medicinal material, for convenience, the natural medicinal material ID can be used as the coreference subject. For example, the agricultural natural medicinal material Artemisia annua corresponds to the natural medicinal material with NMM ID NMM-0001. Therefore, the Artemisia annua in sentence (4) can be coreference annotated and converted into sentence (5):

[0038] (4) Artemisia annua is a natural medicinal material.

[0039] (5) [[NMM-0001|Artemisia annua]] is a natural medicinal material.

[0040] In addition, according to the embodiments of the present application, each natural medicinal material term is also annotated with a co-referential annotation of the standard translation in multiple languages. Just as an example, for example, the standard translation of the target language of the natural medicinal material term at least includes the co-referential subject of the natural medicinal material term expressed in the target language, the common name of the natural medicinal material corresponding to the co-referential subject of the natural medicinal material term, and the natural medicinal material system name corresponding to the co-referential subject of the natural medicinal material term. Then, one way of expressing the standard translation of the natural medicinal material term can be: the natural medicinal material system name corresponding to the natural medicinal material co-referential subject (natural medicinal material co-referential subject, natural medicinal material common name corresponding to the natural medicinal material co-referential subject). Taking "ephedra" as an example, its coreference subject in the natural medicinal material domain knowledge base is NMM-0006, the corresponding English language natural medicinal material common name is Ma-huang, and the corresponding English language natural medicinal material system name is Ephedra equisetina vel intermedia vel sinica Stem-herbaceous. Therefore, the standard English translation of ephedra can be expressed as: Ephedra equisetina vel intermedia vel sinica Stem-herbaceous (NMM-0006, Ma-huang). Under the above premise, the coreference annotation of the standard translation of the natural medicinal material term "ephedra" in multiple languages ​​can be as shown in the following (6) or (7):

[0041] (6)[[NMM-0006|Ephedra equisetina vel intermedia vel sinica Stem-herbaceous(NMM-0006,Ma-huang)]]

[0042] (7)[[NMM-0006|Ephedra equisetina vel intermedia vel sinica Stem-herbaceous(NMM-0006,Ma-huang)]]

[0043] In other embodiments, other elements in the natural medicinal material domain knowledge base and forms different from the above can also be selected as standard translations of natural medicinal material terms. This application does not impose any restrictions on this, as long as it is accurate and standardized.

[0044] In some embodiments, the front-end program 101 is configured to receive a text to be translated in a source language containing natural medicinal material terms and a designation of a target language input by a user through a user interaction interface. As shown in Figure 2, a user can input a text to be translated in a source language containing natural medicinal material terms in the first area A1 of the user interaction interface 1011: "Ephedra is a natural medicinal material.", wherein the source language of the text to be translated T1 is Chinese, the natural medicinal material term included is "Ephedra", and the target language is designated as "English" through the control C1 on the user interaction interface 1011, that is, Chinese is translated into English. In other embodiments, the target language can also be designated as multiple different languages ​​(not shown) at the same time, thereby obtaining translation results in multiple target languages ​​at the same time.

[0045] In some embodiments, the first learning model 102 may be configured to extract natural medicinal herb terms from the text to be translated T1 and transmit them to the search engine 103. In T1, the natural medicinal herb term is "ephedra." In other embodiments, the text to be translated may include more than one natural medicinal herb term. The first learning model 102 may be configured to extract all natural medicinal herb terms contained in the text to be translated and transmit them to the search engine 103.

[0046] In some embodiments, the search engine 103 is configured to search the natural medicinal material terms in the natural medicinal material domain knowledge base 104 and return search results, wherein the search results include the coreference subject of the natural medicinal material terms and the standard translation of the natural medicinal material terms in the target language.

[0047] It should be noted that the coreference primary term graph (CPTG) can usually be pre-constructed based on the natural medicinal material relationship set in the natural medicinal material domain knowledge base, wherein each group of natural medicinal material relationships in the natural medicinal material relationship set can be represented as a manually annotated [source object, relationship, target object] triple, wherein the source object and the target object at least include natural medicinal material terms, and the relationship at least includes a synonym relationship, a containment relationship, a derived / superior relationship, a derived / subordinate relationship, a peer relationship, etc. As an example only, a synonym relationship includes, for example: [Traditional Chinese Medicine, Synonym (Synonym), Chinese Medicine]. In addition, the source object and the target object can also be nouns and terms in different languages, etc. For example, the synonym relationship can also include cross-language synonyms, such as [Natural Medicinal Material, Cross-language Synonym, Natural Medicinal Material], etc.

[0048] Based on the natural medicinal material relationship set, a search engine is used to retrieve each group of natural medicinal material relationships with synonymous relationships in the natural medicinal material relationship set, thereby constructing a coreference graph. The constructed coreference graph contains the coreference terms of natural medicinal material terms and can reflect the synonymous relationship between each natural medicinal material term. Therefore, CPTG can basically be considered as a directed acyclic graph (DAG).

[0049] Still taking ephedra as an example, the synonym relationships associated with the term "ephedra" retrieved by search engine may include [Ephedrae Herba, Synonym, Ephedra], [Ephedra, Synonym, Ephedra], [Ma huang, Synonym, Ephedra], as well as [Ephedra, Synonym, nmm-0006], [Ma-huang, Synonym, nmm-0006], [Ephedra equisetina or Ephedra intermedia or Ephedra herbaceous stem, Synonym, nmm-0006], [Ephedra equisetina vel intermedia vel sinica Stem-herbaceous, Synonym, nmm-0006]. Based on the above synonym search results, a search term including Ephedrae Herba, Ephedra, Ma huang, Ephedra, Ma-huang, Ephedra equisetina or Ephedra intermedia or Ephedra herbaceous stem, Ephedra equisetina vel intermedia vel sinica can be constructed. Stem-herbaceous has multiple graph nodes, each of which points to the coreference graph of the coreference term nmm-0006. Once the CPTG is constructed, preliminary graph calculations can be performed to determine the final coreference term corresponding to any term node. The results of these calculations can be stored in a newly created dictionary. It is worth noting that this dictionary also includes directed edges from the coreference term to itself to allow the queried term itself to be identified as a coreference term.

[0050] It can be seen that by searching in the dictionary established based on the calculation of the coreference subject graph mentioned above, the retrieval efficiency can be greatly optimized, the coreference subjects of natural medicinal terms can be obtained quickly and accurately, and the coreference annotations of the standard translations of the coreference subjects can be used to obtain retrieval results containing the coreference subjects of natural medicinal terms and the standard translations of the natural medicinal terms in the target language.

[0051] The first learning model 102 can be further configured to generate a translation document in the original format based on the search results returned by the search engine 103 using at least one multilingual translation model, wherein the translation document in the original format includes a coreference subject of the natural medicinal material term and a standard translation of the natural medicinal material term in the target language.

[0052] It is worth noting that the first learning model 102 can be composed of multiple sub-models with different functions. For example, the first sub-model is a term extraction model, which is dedicated to extracting natural medicinal material terms, thereby achieving a faster term extraction rate and higher extraction accuracy; the second sub-model is a translation model, which is dedicated to generating translated documents. This multi-model combination method allows each model to focus on its own function, thereby further improving the overall translation performance.

[0053] In some embodiments, the front-end program 101 is further configured to obtain a translation result of a standard translation of the target language containing natural medicinal material terms by parsing the translation document in the original format from the first learning model 102; and present the translation result of the target language on the user interaction interface 1011 in a manner that matches the display mode of the user interaction interface 1011. For example, the translation result can be displayed in the second area A2 as shown in Figure 2 according to the display mode pre-specified by the user. It is worth noting that the layout of the first area A1 for user input and the second area A2 for presentation of translation results is only exemplary, and can also be any other applicable manner, or support user customization and adjustment, etc., and this application does not impose any restrictions on this.

[0054] According to the translation system for natural medicinal material domain knowledge in various embodiments of the present application, various components such as the front-end program, the first learning model, the search engine and the natural medicinal material domain knowledge base are collaboratively designed to cooperate with each other and be closely linked: on the basis of the natural medicinal material domain knowledge base annotating the various natural medicinal material terms stored therein with co-referential subjects and providing multi-language standard translation annotations for each co-referential subject, accordingly, the search engine is enabled to search in the term dictionary created based on the co-referential subject graph search, so that the target language standard translation of the natural medicinal material terms in the search results is obtained based on the co-referential subjects of the natural medicinal material terms, thereby ensuring the accuracy, consistency and standardization of the translation of natural medicinal material terms. At the same time, it can also effectively improve the interpretability of the translation, facilitate user understanding, tracing, and correction, and contain richer natural medicinal material related knowledge that can be deeply utilized in the translation results.

[0055] According to the embodiment of the present application, the natural medicinal materials in the natural medicinal materials domain knowledge base are configured to include all natural medicinal material terms in the "Chinese Pharmacopoeia: 2020 Edition: Volume 1", and when the version of the "Chinese Pharmacopoeia" is updated, the natural medicinal materials domain knowledge base of this application will also be updated accordingly, thereby ensuring that the natural medicinal materials-related text therein can include the updated implementation version of the "Chinese Pharmacopoeia".

[0056] In some embodiments, the front-end program may be further configured to receive, via the user interface, a first user operation on the standard translation of the natural medicinal material term in the target language. The first user operation may be, for example, hovering a mouse over, clicking, double-clicking, or selecting a translation from a right-click menu on the standard translation of the natural medicinal material term in the target language, or any other applicable operation, which is not specifically limited in this application.

[0057] In some embodiments, the front-end program may be further configured to: in response to the first user operation, present a summary page including summary information of the natural medicinal material term and access to a detailed page of the natural medicinal material term on the user interaction interface.

[0058] FIG3 shows a schematic diagram of a first user operation according to an embodiment of the present application. In FIG3 , the first user operation is to hover the mouse over the standard translation of the target language of the natural medicinal material term as an example. When the user hovers the mouse over the standard English translation of the natural medicinal material term "Ephedra equisetina vel intermedia vel sinica Stem-herbaceous (NMM-0006, Ma-huang)" shown in the dotted box 301 in FIG3 , in response to the first user operation, summary information 302 and a summary page 303 containing the natural medicinal material term will be presented on the user interaction interface 1011. The user can further enter the detailed page of the natural medicinal material term via the summary page 303, such as a knowledge page related to the natural medicinal material term in the natural medicinal material domain knowledge base. In some embodiments, the summary page 303 can be a floating box containing the title 304 of the natural medicinal material term and the summary information 302. In some embodiments, the language used when presenting the summary information 302 can be consistent with the page where the user interaction interface 1011 is located. In addition, whether it is the translation result in the target language presented in the second area A2 shown in FIG. 2 or the summary information 302 about the natural medicinal material term presented in the summary page 303 shown in FIG. 3 , the standard translation of the natural medicinal material term in the target language can be highlighted in a manner such as a color different from the context, so that the user can better identify and understand the translation result.

[0059] Furthermore, in some embodiments, the front-end program may be further configured to receive a second user operation on the summary page 303, wherein the second user operation may include, for example, clicking on the title 304 of the natural medicinal material term in the summary page 303. Furthermore, in response to the second user operation, a detailed page (not shown) of the natural medicinal material term is presented to the user, wherein the detailed page includes a knowledge set associated with the natural medicinal material term in the natural medicinal material domain-specific knowledge base.

[0060] In addition, in order to facilitate users to discover interaction methods such as the first user operation and the second user operation, user prompts can be provided in a variety of ways on the user interaction interface 1011, including but not limited to the cursor changing shape when the user's mouse passes over the relevant activation area, and / or, prompting the function of the activation area in the form of floating text, etc., which are not listed here one by one.

[0061] Therefore, the translation system according to the embodiment of the present application can not only provide accurate and standardized text translation results, but also provide the possibility of acquiring knowledge related to natural medicinal material terms through friendly and intuitive user interaction methods, further improving the interpretability, credibility and user satisfaction of the translation results, and enabling the translation system to also have the ability to explore, learn and discover natural medicinal material-specific knowledge.

[0062] In other embodiments, the front-end program may be further configured to receive user-added or modified custom translations of words and sentences through the user interaction interface, and add the newly added or modified custom translations of words and sentences to the user's custom translation set.

[0063] Still taking Figure 2 as an example, the user can add any custom translation of words and phrases including natural medicinal material terms through the control C2 on the user interaction interface 1011, or modify and activate / deactivate the custom translation of existing words and phrases through the control C3. In particular, an activation check box can be set before each entry of the custom translation method added by each user. Checking the check box represents that the custom translation method is activated and will be applied to the next translation. In this way, the user can customize and selectively use different custom translation combinations in different translations, which further increases the customizability of the translation.

[0064] Accordingly, the first learning model can be further configured to generate a translated document in its original format based on the search results returned by the search engine and the user's customized translation set. Furthermore, when the source language text to be translated contains words or phrases from the customized translation set, the customized translations from the customized translation set are preferentially used when generating the translated document in its original format. This allows for translation results that better meet the user's personalized needs. Furthermore, as the user uses the translation system, they can gradually accumulate their own customized translation set, resulting in translation results that increasingly meet their desired objectives, allowing them to efficiently obtain accurate and personalized text translations.

[0065] In some embodiments, the front-end program is further configured to, when presenting the target language translation results on the user interface, differentiate between portions of the translation results using the standard translation method and portions using the user's customized translation method, for example by color. This display method allows users to more clearly distinguish between the standard translation method and the customized translation method, thereby better understanding and verifying the translation results.

[0066] In some embodiments, the front-end program may be further configured to receive, via a user interaction interface, a third user operation for specifying translation performance, wherein the translation performance includes at least accuracy priority and speed priority. Still taking FIG. 2 as an example, the user may utilize control C4 to set a translation mode, with different translation modes having different translation performances. For example, when the translation mode is set to “more accurate translation,” accuracy is prioritized while translation speed has a lower priority. However, when the translation mode is set to “faster translation” (not shown), speed is prioritized while accuracy has a lower priority.

[0067] Among them, when the translation performance specified by the third user operation prioritizes accuracy, the first learning model can be enabled to use a first multilingual translation model with higher translation accuracy through methods such as parameter passing; conversely, when the translation performance specified by the third user operation prioritizes speed, the first learning model can be enabled to use a second multilingual translation model with higher translation speed.

[0068] By allowing the user to specify translation performance, the translation system according to the embodiment of the present application is given greater flexibility, allowing the user to select translation performance in a targeted manner based on specific characteristics such as the length and purpose of the text to be translated, thereby enabling the translation system to better match the user's actual needs.

[0069] In some embodiments, the natural medicinal material domain-specific knowledge base is further configured as follows: each natural medicinal material term stored in the natural medicinal material domain-specific knowledge base also has a coreference annotation of multiple standard translations in multiple languages ​​including different deformations, and the deformations at least include part of speech changes and singular and plural changes.

[0070] Accordingly, the first learning model can be further configured to generate a translation document in the original format based on the search results returned by the search engine, wherein the translation document in the original format includes a coreference of the natural medicinal material term and multiple standard translations of the natural medicinal material term in the target language, including different variations. In this way, a user can be provided with multiple alternative translations of the natural medicinal material term, so that the user can select the most correct, most consistent with the original meaning, or most suitable for the context.

[0071] In an embodiment of the present application, a user can configure the display mode of the user interaction interface, wherein the display mode of the user interaction interface may include at least a combination of a first mode and a second mode, wherein the first mode includes at least one of a Chinese display mode, an English display mode, an English-Chinese display mode, and a Chinese-English display mode; and the second mode includes displaying the translated document in the original format and not displaying the translated document in the original format.

[0072] Accordingly, the front-end program is further configured to receive, via the user interaction interface, a fourth user operation specifying a first mode and a fifth user operation specifying a second mode; and in response to the fourth user operation and the fifth user operation, present the translation result of the target language on the user interaction interface in a display mode that matches the combination of the first and second modes specified by the user. Regarding the first display mode, i.e., the language display mode, as previously described, it can be consistent with the language setting on the page where the user interaction interface is located, or it can be additionally set by the user to generate a customized translation document that meets the user's requirements.

[0073] Figures 4(a) and 4(b) illustrate schematic diagrams of displaying a translated document in its original format on a user interface according to an embodiment of the present application. The translation results presented in the second area A2 of Figure 4(a) correspond to a second mode in which the translated document in its original format is not displayed. The user can use control 401 to toggle between displaying the translated document in its original format and not displaying it. When the mouse is moved over control 401, a floating box prompt such as "Show MLMD Source Code" appears.

[0074] The translated document presented in the second area A2 in FIG4(b) corresponds to the second mode, which displays the translated document in its original format. In the original format, natural medicinal material terms using the standard translation are displayed in a [[subject|standard translation]] format similar to coreference annotations, while those using user-defined translations are displayed in a [[user-defined translation]] format. As shown in FIG4(b), in the original format, "Ma Huang" uses the standard translation and is displayed as [[nmm-0006|Ephedra equisetina vel intermedia vel sinica Stem-herbaceous(NMM-0006,Ma-huang)]], while "Natural Medicinal Material" uses the user-defined translation and is displayed as [[Natural Medicinal Material]]. By allowing the translated documents in their original format to be displayed in the user interaction interface, the interpretability of the translation results can be further enhanced. For professional users who deeply use the translation system of this application, the translated documents in their original format can be stored in a private database, and can also be easily integrated with other codes or documents written in the same format, providing convenience for users.

[0075] In some embodiments, the front-end program is further configured such that a user calls the front-end program by accessing a web page.

[0076] In other embodiments, the front-end program is further configured to provide the user with the option of collecting, downloading and quoting all the previous translation records on the user interaction interface, so that the user can add all the previous translation records to the user's private information, or download the original data associated with the natural medicinal material terminology, or quote the pages of each translation record. In particular, for professional users who use the translation system of the present application in depth, the translation system of the present application can not only be used as a translation tool, but also as a recording tool for their academic research and learning, and as the number of uses increases, the translation can become more and more accurate, and the relevant private information will become more and more abundant.

[0077] In addition, the front-end program can be further configured to: set a user evaluation option on the user interaction interface so as to improve the training of the first learning model using the evaluation information submitted by the user. Figure 5 shows a user evaluation schematic diagram according to an embodiment of the present application. As shown in Figure 5, the user can use the control 501 set in the second area A2 to give a user evaluation. The user evaluation options can include two types of good and bad reviews, or three types of good, general, and bad reviews (not shown). Evaluation options of any multiple levels can also be set (not shown). This application does not make specific restrictions on this. It is worth noting that the user evaluation given by the user will be fed back to the first learning model for improved training of the first model. Therefore, with the use and evaluation of the translation system in this application, the performance of the translation system will also gradually improve.

[0078] The above description is intended to be illustrative and not restrictive. For example, the above examples (or one or more schemes thereof) can be used in combination with each other. Other embodiments may be used by those skilled in the art when viewing the above description. Moreover, in the above detailed description, various features may be combined together to simplify the application. This should not be interpreted as intending that disclosed features that are not claimed for protection are essential to any claim. Rather, the subject matter of the invention may lie in a feature combination that is less than all the features of a disclosed embodiment. Therefore, the claims are hereby incorporated into the detailed description as examples or embodiments, wherein each claim is independently a separate embodiment, and it is conceivable that these embodiments can be combined with each other in various combinations or permutations. The scope of the present invention should be determined with reference to the claims and the full scope of equivalents to which these claims are assigned.

Claims

1. A translation system for natural medicinal material domain knowledge, characterized in that: It includes front-end program, first learning model, search engine and natural medicinal materials domain knowledge base, among which, The front-end program is configured as follows: receiving, through a user interaction interface, a text to be translated in a source language including natural medicinal material terms and a designation of a target language input by a user; Obtaining translation results of the target language containing the standard translation of natural medicinal material terms by parsing the translation document in the original format from the first learning model; Presenting the translation result in the target language on the user interaction interface in a manner matching the display mode of the user interaction interface; The first learning model is configured as: Extracting natural medicinal material terms from the text to be translated and transmitting them to the search engine; Using at least one multilingual translation model, based on the search results returned by the search engine, a translation document in an original format is generated, wherein the translation document in the original format includes a coreference of the natural medicinal material term and a standard translation of the natural medicinal material term in a target language; The search engine is configured to: Searching the natural medicinal material term in the natural medicinal material domain knowledge base and returning search results, wherein the search results include the coreference subject of the natural medicinal material term and the standard translation of the natural medicinal material term in the target language; The natural medicinal materials domain-specific knowledge base is configured as follows: Each natural medicine term stored in the natural medicine domain-specific knowledge base has a coreference annotation of a common subject and a coreference annotation of standard translations in multiple languages.

2. The translation system according to claim 1, characterized in that: The front-end program is further configured as follows: receiving, through the user interaction interface, a first user operation of a user for a standard translation of the natural medicinal material term in a target language; In response to the first user operation, presenting a summary page including summary information of the natural medicinal material term and access to a detailed page of the natural medicinal material term on the user interaction interface; A second user operation on the summary page is received, and in response to the second user operation, a detailed page of the natural medicinal material term is presented to the user, wherein the detailed page includes a knowledge set associated with the natural medicinal material term in the natural medicinal material-specific knowledge base.

3. The translation system according to claim 2, characterized in that The first user operation includes hovering over the standard translation of the natural medicinal material term in the target language; the summary page is a floating box containing the title and summary information of the natural medicinal material term; the second user operation includes clicking on the title of the natural medicinal material term.

4. The translation system according to any one of claims 1 to 3, characterized in that: The front-end program is further configured to: receive a user-defined translation of a word or sentence added or modified by the user through the user interaction interface, and add the user-defined translation of the word or sentence added or modified to the user's defined translation set; The first learning model is further configured to generate a translation document in the original format based on the retrieval results returned by the search engine and the user's custom translation set, and, when the source language text to be translated contains words and sentences in the custom translation set, give priority to using the custom translations in the custom translation set when generating the translation document in the original format.

5. The translation system according to claim 4, characterized in that: The front-end program is further configured as follows: When the translation result of the target language is presented on the user interaction interface, the part of the translation result using the standard translation method and the part using the user's customized translation method are displayed separately.

6. The translation system according to any one of claims 1 to 3, characterized in that: The front-end program is further configured as follows: receiving, through the user interaction interface, a third user operation of the user specifying translation performance, wherein the translation performance includes at least accuracy priority and speed priority; When the translation performance specified by the third user operation is accuracy priority, causing the first learning model to use a first multilingual translation model with higher translation accuracy; When the translation performance designated by the third user operation prioritizes speed, the first learning model is caused to use a second multilingual translation model having a higher translation speed.

7. The translation system according to any one of claims 1 to 3, characterized in that: The natural medicinal material domain-specific knowledge base is further configured as follows: each natural medicinal material term stored in the natural medicinal material domain-specific knowledge base also has a coreference annotation of multiple standard translations in multiple languages ​​including different deformations, wherein the deformations at least include part of speech changes and singular and plural changes; The first learning model is further configured to generate a translation document in an original format based on the retrieval results returned by the search engine, wherein the translation document in the original format includes a coreference subject of the natural medicinal material term and a plurality of standard translations of the natural medicinal material term in a target language including different variations.

8. The translation system according to any one of claims 1 to 3, characterized in that: The display mode of the user interaction interface includes a combination of a first mode and a second mode, wherein the first mode includes at least one of a Chinese display mode, an English display mode, an English-Chinese display mode, and a Chinese-English display mode; and the second mode includes displaying the translated document in the original format and not displaying the translated document in the original format; The front-end program is further configured as follows: Receiving, through the user interaction interface, a fourth user operation for specifying the first mode and a fifth user operation for specifying the second mode; In response to the first user operation and the second user operation, the translation result in the target language is presented on the user interaction interface in a display mode that matches a combination of the first mode and the second mode specified by the user.

9. The translation system according to any one of claims 1 to 3, characterized in that: The standard translation of the natural medicinal material term in the target language at least includes the co-referential subject of the natural medicinal material term expressed in the target language, the common name of the natural medicinal material corresponding to the co-referential subject of the natural medicinal material term, and the systematic name of the natural medicinal material corresponding to the co-referential subject of the natural medicinal material term.

10. The translation system according to any one of claims 1 to 3, characterized in that: The target language is a plurality of different languages.

11. The translation system according to any one of claims 1 to 3, characterized in that: The front-end program is further configured as follows: The user interaction interface provides the user with the options of collecting, downloading and citing all previous translation records, so that the user can add all previous translation records to the user's private information, or download the original data associated with the natural medicinal material terms, or quote the pages of each translation record.

12. The translation system according to any one of claims 1 to 3, characterized in that: The front-end program is further configured as follows: A user evaluation option is set on the user interaction interface so as to improve and train the first learning model using the evaluation information submitted by the user.

13. The translation system according to any one of claims 1 to 3, characterized in that: The front-end program is further configured such that a user calls the front-end program by accessing a web page.

14. The translation system according to any one of claims 1 to 3, characterized in that: The natural medicinal materials domain-specific knowledge base is configured to include all natural medicinal materials terms in the "Chinese Pharmacopoeia: 2020 Edition: Volume 1" or the updated and implemented version of the "Chinese Pharmacopoeia".

Images