Term translation method and system based on artificial intelligence and storage medium

Abstract

The invention relates to the field of terminology information processing and application, in particular to a terminology translation method and system based on artificial intelligence and a storage medium. The method comprises the steps of extracting a first term text from a first language, combining the first term text with a sentence where the first term text is located to form first input information, and adding a cue word text to the first input information to obtain second input information; processing the second input information through a large language model to obtain first field information; combining the first term text with the first domain information to form a first defined text through a large language model; vectorizing the first definition text to obtain a first semantic vector; and performing similarity comparison after repeating the steps. The model performs semantic analysis on the basis of the terms and the sentences where the terms are located, information parameters in a specific language environment are combined, candidate translations corresponding to the terms are searched firstly on the basis of a large language model in the technical field, the candidate translations are selected, and a translation result with the highest matching degree is output.

Description

Technical Field

[0001] The present invention relates to the field of professional terminology information processing and application, and in particular to a terminology translation method, system and storage medium based on artificial intelligence. Background Art

[0002] At present, terminology translation and management mainly rely on term sets and dictionaries published in professional fields. However, they cannot correctly match polysemous words and require manual judgment and selection. Secondly, term sets are static and have limited update frequency, making it difficult to keep up with the rapid development of professional fields and the emergence of new terms. Furthermore, the high operating and maintenance costs of term databases limit the speed of application and update. And when faced with term expansion in different fields and languages, the support capabilities are insufficient.

[0003] Therefore, it is necessary to provide an artificial intelligence-based terminology translation method, system and storage medium to break the above limitations. Summary of the invention

[0004] The present invention provides a terminology translation method, system and storage medium based on artificial intelligence, including the following three aspects:

[0005] In a first aspect, a terminology translation method based on artificial intelligence comprises the following steps:

[0006] S1, selecting a first language, extracting a first term text from the first language, combining the first term text with the sentence where the first term text is located to form first input information, adding a prompt word text to the first input information to obtain second input information; processing the second input information through a large language model to obtain first domain information;

[0007] S2. Combining the first term text with the first domain information through a large language model to form a first definition text;

[0008] S3, vectorizing the first definition text to obtain a first semantic vector;

[0009] S4, select a second language, and repeat steps S1 to S3 to obtain a second semantic vector;

[0010] S5. After setting a similarity threshold, the similarity between the first semantic vector and the second semantic vector is compared with the threshold as a reference and a judgment is made.

[0011] Furthermore, step S6 is included, specifically, if the similarity is greater than the threshold, it is determined to be the same term, and if the similarity is less than or equal to the threshold, it is determined to be a non-identical term.

[0012] Furthermore, the prompt words are used to ensure that the model is more accurate in processing terminology domain classification.

[0013] Furthermore, the first set threshold is 90%.

[0014] Furthermore, step S2 classifies and retrieves the target language terms from an existing term library according to the field, and performs field and context screening through a large language model.

[0015] Furthermore, before performing the step S5, the system further includes an annotation correction parameter. If the sum of the annotation correction parameter and the similarity value is greater than the first set threshold, it is determined to be the same term.

[0016] Furthermore, the system also includes a terminology evidence-based module, which classifies terms into levels and forms level parameters, and loads the level parameters when generating the second language text.

[0017] In the second aspect, a terminology translation system based on artificial intelligence includes a semantic vectorization algorithm model, a text classification model and a vector comparison model. The semantic vectorization algorithm model is used to vectorize the meaning of text content, the text classification model is used to obtain the classification to which a specific term belongs, and the vector comparison model is used to compare the similarity of different vectors.

[0018] In a third aspect, a readable storage medium is provided, wherein execution instructions are stored in the readable storage medium, and when the execution instructions are executed by a processor, they are used to implement the method described in the first aspect.

[0019] This technical solution adopts advanced large language model semantic understanding technology. After the text is input into the model, the model will perform semantic analysis based on the "term" and the "sentence", and output the information parameters of the "term" in the specific language environment of the "sentence". Based on the technical field, the large language model will first search for the candidate translation corresponding to the term, select the candidate translation, and output the translation result with the highest matching degree. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 It is a logic block diagram;

[0021] Figure 2 Sets the reference image for the threshold. DETAILED DESCRIPTION

[0022] 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 implementation methods. The embodiments of the present application are further described in detail below in conjunction with the accompanying drawings and specific implementation methods, but are not intended to limit the present application.

[0023] The words "first", "second" and similar words used in this application do not indicate any order, quantity or importance, but are only used to distinguish. Words such as "include" or "comprise" and similar words mean that the elements before the word include the elements listed after the word, and do not exclude the possibility of also covering other elements. The execution order of the various steps in the method described in this application in conjunction with the accompanying drawings is not intended to be limiting. As long as it does not affect the logical relationship between the various steps, 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.

[0024] It should also be understood that the term "and / or" in this application is only a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this application generally indicates that the associated objects before and after are in an "or" relationship.

[0025] Example 1

[0026] refer to Figure 1 This embodiment discloses a terminology translation method based on artificial intelligence, comprising the following steps:

[0027] S1. Select a first language, extract a first term text from the first language, combine the first term text with the sentence where the first term text is located to form first input information, add a prompt word text to the first input information to obtain second input information; process the second input information through a large language model to obtain first domain information.

[0028] In the following text, term 1 is used to simplify and replace the first term text, language A is used to simplify and replace the first language, definition 1 is used to simplify and replace the first definition text, and sentence 1 is used to simplify and replace the sentence in which the first term text is located.

[0029] The explanation is as follows: The first step is to extract the terms of a sentence in language A and obtain term 1. "Term 1" and "the sentence where term 1 is located" are used as the input information of the model. Under the prompt words we designed, the large language model is allowed to determine the specific field of "term 1" based on the input information. Since terms may have different meanings in different fields. For example, the word "apple" may refer to "Apple" in the computer field. In the field of plants, it refers to "a kind of fruit". The purpose is to use the two pieces of information "term 1" and "the sentence where term 1 is located" to allow the fine-tuned large language model to determine the specific field of "term 1" in the sentence.

[0030] S2, combining the first term text with the first domain information through a large language model to form a first definition text;

[0031] As explained below, this step obtains the "term domain" and "term 1" from the model in the previous step. These two pieces of information are input into the large language model. Here we use another set of prompt words to fine-tune the large language model. This allows the large language model to accurately generate the definition of "term 1" under the "term domain". The characteristic of the large language model with a huge knowledge base is utilized here to enable the model to generate definitions of terms. The definition is generated because the term itself is limited by the number of characters, and usually fewer characters are required to express a complex concept. At this time, if only the terms are semantically vectorized, there will be too many words with similar vectors between different languages, and there will be large deviations in the definitions of the words. Therefore, we choose to let the model understand the meaning of the term in the sentence first to determine its domain, and then let the model develop a definition for this term. S3. Vectorize the first definition text to obtain the first semantic vector;

[0032] As described below, this step uses a semantic vectorization model to vectorize the generated "definition of term 1". It is engaged in converting the "term definition" into a "semantic vector" that a computer can use for semantic comparison.

[0033] S4, select a second language, repeat steps S1 to S3 to obtain a second semantic vector; Figure 1 , S5, after setting the similarity threshold, the similarity between the first semantic vector and the second semantic vector is compared with the threshold as a reference and a judgment is made.

[0034] The step S6 is also included, specifically, if the similarity is greater than a threshold, it is determined to be the same term, and if the similarity is less than or equal to the threshold, it is determined to be a non-identical term.

[0035] Hint words are used to ensure that the model is more accurate in processing term domain classification.

[0036] The first threshold is set to 90%. Figure 2 , the vectors in definition vector set 1 are calculated in sequence with the vectors in definition vector set 2 for similarity. If the similarity is higher than the threshold, it is assumed that the two have the same meaning, such as vector A and vector F.

[0037] Step S2 classifies and retrieves target language terms from an existing term library according to the domain, and performs domain and context screening using a large language model.

[0038] Before performing step S5, the system further includes an annotation correction parameter. If the sum of the annotation correction parameter and the similarity value is greater than a first set threshold, it is determined to be the same term.

[0039] The system also includes a terminology evidence-based module, which classifies terms and forms level parameters, and then generates a second language text to load the level parameters.

[0040] The application is described as follows

[0041] #Role:Terminology Management Expert

[0042] ##Profile

[0043] -Language proficiency: Japanese, English

[0044] - Role Description: Terminology Management Specialists are skilled in accurately identifying, researching and maintaining specialized terminology in a multilingual environment to ensure consistency and professionalism in translation.

[0045] -Knowledge reserve: Medicine, neurosurgery

[0046] ##background:

[0047] Vocabulary source: Medical professional works

[0048] Vocabulary Language: Japanese

[0049] ##input:

[0050] |<term>|<sentence>|

[0051] ##task: Complete the following tasks in sequence according to the terms and the sentences in which the terms are located:

[0052] 1. Classify vocabulary according to the classification standards of Chinese professional disciplines.

[0053] ##Rules:

[0054] - If no result can be output, NA is returned.

[0055] ##Print:

[0056] |<Term>|<Field Information>|

[0057] ##Initialization

[0058] As a role <role>have <profile>The capability, reference <background>Provide the information to complete <task>, strictly abide by <rules>,by <print>Take a deep breath before jumping to conclusions.

[0059] #Role: Neurosurgery terminologist

[0060] ##Profile

[0061] - Proficient in languages: Japanese, Chinese

[0062] - Role Description: Terminology Management Specialists are skilled in accurately identifying, researching and maintaining specialized terminology in a multilingual environment to ensure consistency and professionalism in translations.

[0063] -Knowledge reserve: Medicine, neurosurgery

[0064] ##background:

[0065] - Vocabulary source: Medical professional works

[0066] - Vocabulary language: Japanese

[0067] ##Input:

[0068] |<Term>|<Field Information>|

[0069] ##task: Complete the following tasks in sequence:

[0070] - Based on the input <domain information>, determine the definition of the term. The definition of the term should include the following key contents: the core concept of the term, the attributes and characteristics of the term, and the scope and limitations of the term.

[0071] ##Rules:

[0072] - If no result can be output, NA is returned.

[0073] ##Print:

[0074] |<Term>|<Definition>|

[0075] ##Initialization

[0076] As a role <role>have <profile>The capability, reference <background>Information provided, received <input> Information, complete <task>, strictly abide by <rules>,by <print>Please think step by step and take a deep breath before drawing conclusions.

[0077] Example 2

[0078] Based on Example 1, this example describes an artificial intelligence-based terminology translation system, including a semantic vectorization algorithm model, a text classification model, and a vector comparison model. The semantic vectorization algorithm model is used to vectorize the meaning of text content, the text classification model is used to obtain the classification to which a specific term belongs, and the vector comparison model is used to compare the similarity of different vectors. This example describes the system using a computer to compare words in two different languages ​​that refer to the same concept, specifically for analyzing whether a term is consistent between texts in two different languages.

[0079] For example

[0080] A: An intracranial aneurysm is an abnormal dilation of a blood vessel wall that may have serious consequences for the patient’s health. (Sentence 1)

[0081] Language B: Intracranial aneurysms are abnormal dilations of blood vessels that can have severe consequences for patients' health. (Sentence 2)

[0082] For the two words marked in yellow in language A and language B, non-professionals cannot confirm whether the two words refer to the same concept. This system can be used to determine whether the two words refer to the same concept.

[0083] The specific steps are as follows:

[0084] 1. Input "intracranial aneurysm" (term 1) and "sentence 1" into the large language model (term classification model) that has been fine-tuned with specific prompt words. The large language model will determine and output the professional field of the term "intracranial aneurysm" - "neurosurgery" based on the meaning of "intracranial aneurysm" in "sentence 1".

[0085] 2. Input "neurosurgery" and "intracranial aneurysm" into the large language model (term definition model) that has been fine-tuned with specific prompt words. The large language model will generate a definition of "intracranial aneurysm" in the field of "neurosurgery". Thus, the "term definition" of "intracranial aneurysm" in language A is obtained - "intracranial aneurysm is a vascular disease in the field of neurosurgery, which refers to the abnormal local expansion or protrusion of the cerebral artery, usually appearing at the bifurcation of the blood vessels. It may be congenital or caused by acquired factors such as arteriosclerosis and infection. If an intracranial aneurysm ruptures, it may cause subarachnoid hemorrhage, which is seriously life-threatening. Therefore, it is crucial to detect and treat intracranial aneurysms in time to prevent potential serious complications." This is definition 1.

[0086] 3. Input "Intracranial aneurysms" (term 2) and "sentence 2" into the large language model (term classification model) that has been fine-tuned with specific prompt words. The large language model will judge and output the professional field of the term "Intracranial aneurysms" - "neurosurgery" based on the meaning of "Intracranial aneurysms" in "sentence 2".

[0087] 4. Input "neurosurgery" and "Intracranial aneurysms" into the large language model (term definition model) that has been fine-tuned with specific prompt words. The large language model will generate the definition of "Intracranial aneurysms" in the field of "neurosurgery". Thus, the "term definition" of "Intracranial aneurysms" in language B is obtained - "In neurosurgery," Intracranial aneurysms" refer to localized, abnormal expansions or bulges in the walls of the arteries within the brain, often occurring at bifurcations of blood vessels. These aneurysms can be congenital or caused by factors such as arteriosclerosis or infection. If ruptured, they can lead to subarachnoid hemorrhage, a severe medical condition requiring prompt diagnosis and treatment." (Definition 2)

[0088] 5. Through the semantic vectorization algorithm, the semantic vectors of "Definition 1" and "Definition 2" are obtained. Our system first uses a multilingual text embedding model to obtain the embedding vector of each definition word. This model is trained using a large number of multi-corpora to capture and retain the semantic information of the same concept in different languages.

[0089] The core of this model uses the deep learning algorithm of "neural network language model", and its goal is to learn a function that can map text content to a continuous vector space, so that semantically similar words are also close to each other in this space. In fact, this process is a kind of "semantic vectorization".

[0090] In the specific operation, we will input the definition of each term in language A and language B into this multilingual word embedding model, and the model will output the corresponding embedding vector. "Definition 1 vector" and "Definition 2 vector" are obtained in this way.

[0091] 6. Measure the similarity between "Definition 1 vector" and "Definition 2 vector". The specific measurement method used is to calculate the cosine similarity between the two. The calculation formula is as follows:

[0092]

[0093] Among them, A and B are the two vectors we want to compare, '·' represents the dot product operation, and ‖A‖2 and ‖B‖2 represent the bi-norm (i.e., the length of the vector) of vector A and B respectively.

[0094] At this time, these two vectors are generated by our system, which uses a multilingual word embedding model to generate "Definition 1 Vector" and "Definition 2 Vector". In this model, the embedding vectors of the same or similar concepts in different languages ​​should be close. This is based on the distribution hypothesis theory, which believes that the meaning of a word can be defined by the context of the words around it, that is, if two words have similar roles in their respective contexts, then their semantics should also be similar. Therefore, in a cross-lingual environment, even if the words are from different languages, as long as they refer to the same or similar concepts, their embedding vectors should be similar. In our system, if the cosine value between "Definition 1 Vector" and "Definition 2 Vector" is greater than or equal to 0.85, we will consider that the words represented by the two vectors are almost semantically consistent, that is, they refer to the same concept. For example, if the vector cosine similarity of the two words "intracranial aneurysm" in language A and "Intracranialaneurysms" in language B exceeds 0.85, then we can assume that their meanings are consistent.

[0095] Example 3

[0096] The present embodiment is based on the present embodiment 2 in that the present embodiment describes the technology for establishing a term base. The difference from the present embodiment 2 is that the present embodiment shows the scenario in which the technology is used in "two texts in different languages ​​but with similar contents". The present technology can automatically find corresponding terms in two texts and quickly establish a term base. When the user has two documents in different languages ​​but with similar contents, there may be multiple terms with the same concept but in different languages.

[0097] For example:

[0098] When you have both the "Guidelines for the Clinical Management of Unruptured Intracranial Aneurysms in China (2024 Edition)" (Document 1) and the "2023 Guideline for the Management of Patients With Aneurysmal Subarachnoid Hemorrhage: A Guideline From the American Heart Association / American Stroke Association" (Document 2). Since both documents are about the clinical treatment of aneurysms, there may be many corresponding Chinese and English terms in them. (It should be noted that these two documents are not translated, but were drafted by experts from their respective countries). Users can use this patented technology to quickly establish Chinese and English term pairs from these two documents, thereby establishing a term library. The specific steps are as follows.

[0099] 1. For the text in Document 1 and Document 2, use the punctuation mark to separate the sentences (the punctuation mark can be defined by yourself. The default is period, exclamation mark, question mark). Get the sentence set of Document 1 (sentence set 1) and the sentence set of Document 2 (set 2)

[0100] 2. Use the term extraction model to extract terms from all sentences in set 1 and set 2. Obtain all "terms" and "term sentences" in set 1, and all "terms" and "term sentences" in set 2.

[0101] Term extraction uses a large language model. The model is fine-tuned with cue words. The cue words used are as follows.

[0102] #Role:Terminology Management Expert

[0103] ##Profile

[0104] -Language skills: Japanese, English

[0105] - Role Description: Terminology Management Specialists are skilled in accurately identifying, researching and maintaining specialized terminology in a multilingual environment to ensure consistency and professionalism in translation.

[0106] ##background:

[0107] -Content Area: Medicine

[0108] -Languages ​​included: Japanese, English

[0109] -Terminology definition: The terminology should be professional, scientific, unambiguous and systematic. ##Task:

[0110] -Please identify and extract all the technical terms from the given text.

[0111] ##Rules:

[0112] - Returns NA if there is no term.

[0113] ##Print:

[0114] <Term>

[0115] ##Initialization:

[0116] As a role <role>have <profile>The capability, reference <background>Provide the information to complete <task>, strictly abide by <rules>,by <print>3. Perform the step (125 or 345) of obtaining the "term definition vector" in "Example 1" for each term. At this time, the definition vectors of all terms in document 1 (definition vector set 1) and the definition vectors of all terms in document 2 (definition vector set 2) are obtained.

[0117] 4. Traverse "Definition vector set 1" and calculate the cosine similarity with the vectors in "Definition vector set 2" in turn. If the cosine value is greater than 0.85, it is considered that the two have the same meaning and constitute a term pair, refer to Example 1.

[0118] The pseudo code example is as follows:

[0119]

[0120]

[0121] Example 4

[0122] The difference from Example 3 is that this example demonstrates the application of this technology in a translation scenario. First, a text with translation in a certain language can be analyzed, and then a search can be performed in the term library to quickly obtain an accurate translation of the term.

[0123] For example:

[0124] When preparing to translate the "Chinese Unruptured Intracranial Aneurysm Clinical Management Guidelines (2024 Edition)" (Document 1) into English, we assume that we already have a corresponding terminology database. The terminology database data structure must include any one of the two items: "term string itself", "term example sentence" or "term definition". Or the terminology database itself is a vector database.

[0125] The specific steps are as follows.

[0126] 1. Use the sentence segmentation model to segment the text in document 1. Get the sentence set of document 1 (sentence set 1)

[0127] 2. For all sentences in set 1, use the term extraction model to extract terms. Get all the "terms" and "term sentences" in set 1.

[0128] 3. For each term in set 1, the step of obtaining the "term definition vector" in "Example 1" (125 or 345) is performed. At this time, the definition vectors of all terms in document 1 are obtained (definition vector set 1)

[0129] 4. Use the term definition vector obtained in 3 to complete term retrieval in the term library.

[0130] 5. If the terms extracted from document 1 are stored in the term base, the corresponding terms in the term base will be obtained based on semantics.

[0131] Because this method refers to semantics, it can find the corresponding terms in the term base more accurately than traditional string retrieval.

[0132] Example 5

[0133] This embodiment illustrates a readable storage medium, in which execution instructions are stored. When the execution instructions are executed by a processor, they are used to implement the method in Embodiment 1.

[0134] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.< / print> < / rules> < / task> < / background> < / profile> < / role> < / print> < / rules> < / task> < / background> < / profile> < / role> < / print> < / rules> < / task> < / background> < / profile> < / role>

Claims

1. A terminology translation method based on artificial intelligence, characterized in that: The steps include: S1, selecting a first language, extracting a first term text from the first language, combining the first term text with the sentence where the first term text is located to form first input information, adding a prompt word text to the first input information to obtain second input information; processing the second input information through a large language model to obtain first domain information; S2. Combining the first term text with the first domain information through a large language model to form a first definition text; S3, vectorizing the first definition text to obtain a first semantic vector; S4, select a second language, and repeat steps S1 to S3 to obtain a second semantic vector; S5. After setting a similarity threshold, the similarity between the first semantic vector and the second semantic vector is compared with the threshold as a reference and a judgment is made.

2. The term translation method based on artificial intelligence according to claim 1, characterized in that: The step S6 is also included, specifically, if the similarity is greater than the threshold, it is determined to be the same term, and if the similarity is less than or equal to the threshold, it is determined to be a non-identical term.

3. The domain term translation method based on artificial intelligence according to claim 1 is characterized in that: The hint words are used to ensure that the model is more accurate in processing term domain classification.

4. The term translation method based on artificial intelligence according to claim 2 is characterized in that: The first set threshold is 90%.

5. The term translation based on artificial intelligence according to any one of claims 1 to 4 The method is characterized in that Step S2 classifies and retrieves the target language terms from an existing term library according to the field, and performs field and context screening using a large language model.

6. The terminology translation method based on artificial intelligence according to claim 5 is characterized in that: Before performing step S5, the system further includes an annotation correction parameter. If the sum of the annotation correction parameter and the similarity value is greater than the first set threshold, it is determined to be the same term.

7. The terminology translation method based on artificial intelligence according to claim 5, characterized in that: The system also includes a terminology evidence-based module, which classifies terms into levels and forms level parameters, and loads the level parameters when generating the second language text.

8. A terminology translation system based on artificial intelligence, characterized in that: It includes a semantic vectorization algorithm model, a text classification model and a vector comparison model. The semantic vectorization algorithm model is used to vectorize the meaning of text content, the text classification model is used to obtain the category to which a specific term belongs, and the vector comparison model is used to compare the similarity between different vectors.

9. A readable storage medium, characterized in that: The readable storage medium stores execution instructions, which are used to implement the method according to any one of claims 1 to 6 when executed by a processor.

Images