Term translation recommendation method and device, electronic equipment and storage medium

Abstract

The application provides a term translation recommendation method and device, electronic equipment and a storage medium, and belongs to the technical field of natural language processing. The method comprises the following steps: obtaining a sentence text to be translated, and extracting a feature vector of the sentence text to be translated; inputting the feature vector of the sentence text to be translated into a pre-constructed recommendation model to obtain a candidate term translation result corresponding to the sentence text to be translated output by the recommendation model; determining a target candidate term translation result based on the candidate term translation result; obtaining user feedback information in real time, determining a recommendation priority of the target candidate term translation result based on the user feedback information; and the recommendation model is trained based on the feature vector of a sample sentence text to be translated and a candidate term translation result corresponding to the sample sentence text to be translated. The application can effectively improve the translation efficiency and accuracy.

Description

Technical Field

[0001] This invention relates to the field of natural language processing technology, and in particular to a terminology translation recommendation method, apparatus, electronic device, and storage medium. Background Technology

[0002] In computer-aided translation, terminology suggestion is a crucial function. Translators typically connect to multiple terminology databases, and a single term often corresponds to multiple definitions. Existing terminology suggestion functions usually present all possible translations for a term to the translator, requiring the translator to sift through a large number of results to select the correct term. This leads to inefficiency as translators struggle to quickly choose the correct term for use in their translations. Therefore, there is an urgent need for a terminology translation recommendation method that can suggest accurate translations to translators. Summary of the Invention

[0003] This invention provides a terminology translation recommendation method, apparatus, electronic device, and storage medium to address the shortcomings of existing terminology prompt functions, which typically present all translation results for a term to the translator, making it difficult for the translator to quickly select the correct terminology translation result and resulting in low work efficiency.

[0004] In a first aspect, the present invention provides a terminology translation recommendation method, comprising:

[0005] Obtain the text of the sentence to be translated, and extract the feature vector of the text of the sentence to be translated;

[0006] The feature vector of the sentence text to be translated is input into a pre-built recommendation model to obtain the candidate term translation results corresponding to the sentence text to be translated output by the recommendation model.

[0007] Based on the translation results of the candidate terms, the translation result of the target candidate term is determined;

[0008] Real-time acquisition of user feedback information; and determination of the recommendation priority of the translation results of the target candidate terms based on the user feedback information.

[0009] The recommendation model is trained based on the feature vector of the sample sentence text to be translated and the candidate term translation result identifiers corresponding to the sample sentence text to be translated.

[0010] In some embodiments, determining the target candidate term translation result based on the candidate term translation results includes:

[0011] Multiple online translation results corresponding to the sentence text to be translated are obtained in real time from multiple online translation platforms;

[0012] Based on the translation results of the candidate terms and the multiple online translation results, the translation result of the target candidate term is determined.

[0013] In some embodiments, determining the target candidate term translation result based on the candidate term translation results and the plurality of online translation results includes:

[0014] Based on the aforementioned multiple online translation results, multiple online terminology translation results were obtained;

[0015] The candidate term translation results are compared with the multiple online term translation results, and the target candidate term translation result is determined from the candidate term translation results and the multiple online term translation results.

[0016] In some embodiments, determining the recommendation priority of the translation results of the target candidate terms based on the user feedback information includes:

[0017] Based on the user feedback information, a score is obtained for the translation result of the target candidate term;

[0018] Based on the scores of the translation results of the target candidate terms, the recommendation priority of the translation results of the target candidate terms is determined.

[0019] In some embodiments, the process of determining the recommendation model includes:

[0020] Obtain sample sentence text to be translated, and extract the feature vector of the sample sentence text to be translated.

[0021] Determine the candidate term translation result identifiers corresponding to the sample sentence text to be translated;

[0022] Using the feature vector of the sample sentence text to be translated as sample data, and the candidate term translation result identifier corresponding to the sample sentence text to be translated as sample label, an initial recommendation model is trained;

[0023] After the initial recommendation model is trained, the recommendation model is obtained.

[0024] In some embodiments, the training of the initial recommendation model includes:

[0025] The feature vector of the sample sentence text to be translated is input into the initial recommendation model to obtain the candidate term translation prediction result corresponding to the sample sentence text to be translated output by the initial recommendation model;

[0026] Obtain the historical translation results of the sample sentence text to be translated;

[0027] Based on the historical translation results, the translation results of historical terms are obtained;

[0028] The candidate term translation prediction results are compared with the historical term translation results, and the parameters of the initial recommendation model are optimized iteratively.

[0029] In some embodiments, after obtaining the recommendation model, the method further includes:

[0030] Based on the user feedback information, the parameters of the recommendation model are optimized in real time.

[0031] Secondly, the present invention provides a terminology translation recommendation device, comprising:

[0032] The acquisition unit is used to acquire the text of the sentence to be translated and extract the feature vector of the text of the sentence to be translated.

[0033] The recommendation unit is used to input the feature vector of the sentence text to be translated into a pre-built recommendation model to obtain the candidate term translation results corresponding to the sentence text to be translated output by the recommendation model;

[0034] The first determining unit is used to determine the translation result of the target candidate term based on the translation results of the candidate terms;

[0035] The second determining unit is used to acquire user feedback information in real time and determine the recommendation priority of the translation results of the target candidate terms based on the user feedback information.

[0036] The recommendation model is trained based on the feature vector of the sample sentence text to be translated and the candidate term translation result identifiers corresponding to the sample sentence text to be translated.

[0037] Thirdly, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the terminology translation recommendation method as described above.

[0038] Fourthly, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the terminology translation recommendation method as described above.

[0039] This invention provides a terminology translation recommendation method, apparatus, electronic device, and storage medium. By acquiring the text of the sentence to be translated, extracting the feature vector of the text, inputting the feature vector into a pre-built recommendation model, obtaining candidate terminology translation results, determining the target candidate terminology translation result, and determining the recommendation priority of the target candidate terminology translation result by acquiring user feedback information in real time, it can effectively improve translation efficiency and accuracy. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0041] Figure 1 This is a flowchart illustrating the terminology translation recommendation method provided in an embodiment of the present invention;

[0042] Figure 2 This is a flowchart illustrating the process of determining the recommendation model provided in an embodiment of the present invention;

[0043] Figure 3 This is a schematic diagram of the terminology translation recommendation device provided by the present invention;

[0044] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0046] The terms "first," "second," etc., used in this specification are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of the invention can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, and the number of objects is not limited; for example, the first object can be one or more.

[0047] Figure 1 This is a flowchart illustrating the terminology translation recommendation method provided in an embodiment of the present invention. Figure 1 As shown, a terminology translation recommendation method is provided, including the following steps: step 110, step 120, step 130, and step 140. This method's steps are merely one possible implementation of the invention.

[0048] Step 110: Obtain the text of the sentence to be translated and extract its feature vector.

[0049] The text to be translated refers to the text of a single sentence that needs to be translated.

[0050] Optionally, the text of the sentence to be translated can be obtained through various means, such as text uploaded by the user or text obtained by speech recognition of the user's voice data.

[0051] Optionally, the text to be translated, which contains multiple sentences, can be segmented into sentences to obtain multiple sentences to be translated.

[0052] Optionally, the text to be translated may be preprocessed, such as text cleaning, word segmentation, stop word filtering, and stemming, to obtain the preprocessed text to be translated.

[0053] Optionally, based on the preprocessed sentence text to be translated, multiple target words are obtained, each target word is encoded to obtain a feature vector of each target word, and the feature vector of the sentence text to be translated is obtained based on the feature vector of each target word.

[0054] It is understandable that by extracting the feature vector of the sentence to be translated, we can capture the semantic and contextual information of the sentence, which makes it easier to apply to the recommendation model for term translation recommendation, thus helping to improve the accuracy and effectiveness of translation.

[0055] Step 120: Input the feature vector of the sentence text to be translated into the pre-built recommendation model to obtain the candidate term translation results corresponding to the sentence text to be translated output by the recommendation model.

[0056] Among them, the candidate term translation result refers to the alternative results obtained after translating the terms in the text of the sentence to be translated.

[0057] Step 130: Based on the translation results of the candidate terms, determine the translation results of the target candidate terms.

[0058] Optionally, the translation results of candidate terms can be screened, supplemented, or corrected to obtain the translation results of the target candidate terms.

[0059] For example, the confidence level of the translation results of candidate terms is calculated, and the translation results of candidate terms are filtered based on a preset confidence threshold to obtain the translation results of the target candidate terms.

[0060] Step 140: Obtain user feedback information in real time, and determine the recommendation priority of the translation results of the target candidate terms based on the user feedback information.

[0061] The recommendation model is trained based on the feature vector of the sample sentence text to be translated and the candidate term translation result identifiers corresponding to the sample sentence text to be translated.

[0062] User feedback information includes ratings, preferences, and comments.

[0063] Understandably, by acquiring user feedback in real time and determining the recommendation priority of target candidate term translation results based on this feedback, it can help improve user satisfaction, increase the accuracy of recommendations, and achieve personalized recommendations, thereby enhancing user engagement and providing a better user experience.

[0064] In some embodiments, the recommendation priority of translation results for target candidate terms is determined based on user feedback information, including:

[0065] Based on user feedback, a score is obtained for the translation results of the target candidate terms;

[0066] Based on the scores of the translation results of the target candidate terms, the recommendation priority of the translation results of the target candidate terms is determined.

[0067] Optionally, weights can be set for user feedback information such as user ratings, selection preferences, and feedback comments. By combining user ratings, selection preferences, and feedback comments, the score of the translation result of the target candidate term can be determined.

[0068] Optionally, the translation results of the target candidate terms can be sorted according to their scores, and the priority order of recommendations can be determined based on the sorting results.

[0069] Optionally, the scores of the translation results of the target candidate terms can be updated in real time, and the recommendation priority of the translation results of the target candidate terms can be adjusted.

[0070] Understandably, by determining the recommendation priority based on the scores of the target candidate term translation results, a more personalized and intelligent ranking of term translation results can be provided that better meets the needs and preferences of users, thus better satisfying the needs of different users.

[0071] In this embodiment of the invention, by acquiring the text of the sentence to be translated, extracting the feature vector of the text, using a pre-built recommendation model, obtaining candidate term translation results, determining the target candidate term translation result, and determining the recommendation priority of the target candidate term translation result by acquiring user feedback information in real time, the flexibility of the recommendation is improved. This can help users interact during the translation process and help them quickly and accurately select the appropriate term translation result, thereby improving the efficiency and accuracy of translation.

[0072] It should be noted that each embodiment of the present invention can be freely combined, rearranged, or executed individually, and does not need to rely on or depend on a fixed execution order.

[0073] In some embodiments, determining the target candidate term translation result based on the candidate term translation results includes:

[0074] Multiple online translation results corresponding to the sentence to be translated are obtained in real time from multiple online translation platforms;

[0075] Based on the translation results of candidate terms and multiple online translation results, the translation result of the target candidate term is determined.

[0076] Online translation platforms refer to platforms that provide translation services via the Internet. The working principle of online translation platforms is to use automated translation engines to perform translation tasks by utilizing large-scale corpora and machine learning algorithms.

[0077] Optionally, based on the translation results of candidate terms and multiple online translation results, the translation results of candidate terms can be supplemented, filtered, or corrected to obtain the translation results of the target candidate terms.

[0078] Understandably, by acquiring multiple online translation results corresponding to the sentence text to be translated in real time, and determining the target candidate term translation result based on the candidate term translation results and multiple online translation results, not only is the diversity of recommendations improved, but also the accuracy and credibility of the recommendations are enhanced.

[0079] In some embodiments, the target candidate term translation result is determined based on candidate term translation results and multiple online translation results, including:

[0080] Based on multiple online translation results, multiple online terminology translation results were obtained;

[0081] The candidate term translation results are compared with multiple online term translation results to determine the target candidate term translation result from the candidate term translation results and multiple online term translation results.

[0082] Optionally, the terms in the sentence text to be translated are identified, and online term translation results are extracted from each online translation result.

[0083] For example, the sentence to be translated is "The concept of artificial intelligence has gained widespread attention in recent years". One online translation result for this sentence is "In recent years, the concept of artificial intelligence has received widespread attention". The term in this sentence is "artificial intelligence", and the extracted online term translation result is "artificial intelligence".

[0084] Optionally, the candidate term translation results can be compared with multiple online term translation results to verify the candidate term translation results, calculate the confidence level of the candidate term translation results, and select candidate term translation results with a confidence level higher than a preset threshold.

[0085] Optionally, the candidate term translation results can be compared with multiple online term translation results, and the multiple online term translation results can be filtered to obtain the filtered online term translation results.

[0086] Optionally, the filtered candidate term translation results can be merged with the filtered online term translation results to obtain the target candidate term translation results.

[0087] Optionally, the candidate term translation results can be compared with multiple online term translation results, and the candidate term translation results can be corrected to obtain the target candidate term translation result.

[0088] Understandably, by comprehensively considering multiple online terminology translation results and candidate terminology translation results corresponding to the sentence to be translated, the accuracy, diversity, and reliability of terminology translation recommendations can be improved, resulting in better translation outcomes.

[0089] Figure 2 This is a flowchart illustrating the process of determining the recommendation model provided in an embodiment of the present invention. Figure 2 As shown, the process of determining the recommendation model includes the following steps: steps 210, 220, 230, and 240. This method flow is merely one possible implementation of the present invention.

[0090] Step 210: Obtain the sample sentence text to be translated and extract the feature vector of the sample sentence text to be translated.

[0091] Alternatively, sample sentences to be translated can be obtained through online translation platforms, corpora, or historical translation databases.

[0092] Optionally, sample text to be translated can be obtained, and the sample text to be translated can be segmented into sentences to obtain multiple sample sentence texts to be translated.

[0093] Optionally, the sample sentence text to be translated can be preprocessed, including text cleaning, word segmentation, stop word filtering, and stemming, to obtain the preprocessed sample sentence text to be translated.

[0094] Optionally, based on the preprocessed sample sentence text to be translated, multiple sample target words are obtained, each sample target word is encoded to obtain the feature vector of each sample target word, and the feature vector of the sample sentence text to be translated is obtained based on the feature vector of each sample target word.

[0095] Step 220: Determine the candidate term translation result identifiers corresponding to the sample sentence text to be translated.

[0096] Optionally, terms can be extracted from the sample sentence text to be translated, and candidate term translation results corresponding to the sample sentence text can be obtained through professional dictionaries, terminology databases, historical translation databases, etc. The candidate term translation results corresponding to the sample sentence text to be translated can be proofread, and the candidate term translation result identifier corresponding to the sample sentence text to be translated can be determined.

[0097] Step 230: Using the feature vector of the sample sentence text to be translated as sample data, and the candidate term translation result identifier corresponding to the sample sentence text to be translated as sample label, train the initial recommendation model.

[0098] Understandably, by training an initial recommendation model based on the feature vector of the sample sentence to be translated and the corresponding candidate term translation result identifier, the model can automatically provide recommendations for candidate term translations for the sentence to be translated, reducing the burden on translators and improving translation efficiency.

[0099] In some embodiments, training the initial recommendation model includes:

[0100] The feature vector of the sample sentence text to be translated is input into the initial recommendation model to obtain the candidate term translation prediction results corresponding to the sample sentence text to be translated output by the initial recommendation model;

[0101] Obtain the historical translation results of the sample sentence text to be translated;

[0102] Based on the historical translation results, the translation results of historical terms are obtained;

[0103] The candidate term translation prediction results are compared with the historical term translation results, and the parameters of the initial recommendation model are optimized iteratively.

[0104] The historical translation results include one or more translation results.

[0105] Optionally, historical translation results of the sample sentence text to be translated can be obtained from a historical translation database.

[0106] Optionally, based on historical translation results, historical terms are identified, and translation results of historical terms are obtained.

[0107] It is understandable that by inputting the feature vector of the sample sentence text to be translated into the initial recommendation model, the candidate term translation prediction result corresponding to the sample sentence text to be translated output by the initial recommendation model is obtained. The historical translation results of the sample sentence text to be translated are obtained. Based on the historical translation results, the historical term translation results are obtained. The candidate term translation prediction results are compared with the historical term translation results, and then the confidence of the candidate term translation prediction results can be calculated. In this way, the parameters of the initial recommendation model can be optimized iteratively to improve the performance of the recommendation model.

[0108] Step 240: After the initial recommendation model is trained, the recommendation model is obtained.

[0109] Optionally, metrics such as accuracy and recall can be used to evaluate the recommendation model.

[0110] In some embodiments, after obtaining the recommendation model, the method further includes:

[0111] Based on user feedback, the parameters of the recommendation model are optimized in real time.

[0112] Understandably, by dynamically adjusting model parameters based on user feedback during real-time recommendation, it is possible to adapt to changes in user interests, improve recommendation effectiveness, and facilitate personalized recommendations to users.

[0113] It should be noted that in the process of determining the recommendation model, parameters are optimized by utilizing historical translation data and real-time user feedback information. This enables the recommendation model to better understand user needs and recommend more accurate candidate term translation results, which helps to make flexible recommendations based on different contexts and needs.

[0114] In this embodiment of the invention, by acquiring sample sentence text to be translated, extracting the feature vector of the sample sentence text to be translated, determining the candidate term translation result identifier corresponding to the sample sentence text to be translated, using the feature vector of the sample sentence text to be translated as sample data, and using the candidate term translation result identifier corresponding to the sample sentence text to be translated as sample label, an initial recommendation model is trained. After the initial recommendation model is trained, a recommendation model is obtained, which can improve the performance of the recommendation model and improve the efficiency and accuracy of translation.

[0115] The terminology translation recommendation device provided in the embodiments of the present invention will be described below. The terminology translation recommendation device described below can be referred to in correspondence with the terminology translation recommendation method described above.

[0116] Figure 3 A schematic diagram of the terminology translation recommendation device provided by the present invention is shown below. Figure 3 As shown, the device 300 includes:

[0117] The acquisition unit 310 is used to acquire the text of the sentence to be translated and extract the feature vector of the text of the sentence to be translated.

[0118] Recommendation unit 320 is used to input the feature vector of the sentence text to be translated into the pre-built recommendation model and obtain the candidate term translation results corresponding to the sentence text to be translated output by the recommendation model;

[0119] The first determining unit 330 is used to determine the translation result of the target candidate term based on the translation results of the candidate terms;

[0120] The second determining unit 340 is used to acquire user feedback information in real time and determine the recommendation priority of the translation results of the target candidate terms based on the user feedback information.

[0121] The recommendation model is trained based on the feature vector of the sample sentence text to be translated and the candidate term translation result identifiers corresponding to the sample sentence text to be translated.

[0122] Optionally, based on the translation results of the candidate terms, the translation results of the target candidate terms are determined to include:

[0123] Multiple online translation results corresponding to the sentence to be translated are obtained in real time from multiple online translation platforms;

[0124] Based on the translation results of candidate terms and multiple online translation results, the translation result of the target candidate term is determined.

[0125] Optionally, based on the translation results of candidate terms and multiple online translation results, the translation result of the target candidate term is determined, including:

[0126] Based on multiple online translation results, multiple online terminology translation results were obtained;

[0127] The candidate term translation results are compared with multiple online term translation results to determine the target candidate term translation result from the candidate term translation results and multiple online term translation results.

[0128] Optionally, based on user feedback, the recommendation priority of the translation results for the target candidate terms is determined, including:

[0129] Based on user feedback, a score is obtained for the translation results of the target candidate terms;

[0130] Based on the scores of the translation results of the target candidate terms, the recommendation priority of the translation results of the target candidate terms is determined.

[0131] Optionally, the process of determining the recommendation model includes:

[0132] Obtain the sample sentence text to be translated, and extract the feature vector of the sample sentence text to be translated;

[0133] Determine the translation result identifiers of candidate terms corresponding to the sample sentence text to be translated;

[0134] The initial recommendation model is trained using the feature vector of the sample sentence text to be translated as the sample data and the candidate term translation result identifier corresponding to the sample sentence text as the sample label.

[0135] After the initial recommendation model is trained, the recommendation model is obtained.

[0136] Optionally, training the initial recommendation model includes:

[0137] The feature vector of the sample sentence text to be translated is input into the initial recommendation model to obtain the candidate term translation prediction results corresponding to the sample sentence text to be translated output by the initial recommendation model;

[0138] Obtain the historical translation results of the sample sentence text to be translated;

[0139] Based on the historical translation results, the translation results of historical terms are obtained;

[0140] The candidate term translation prediction results are compared with the historical term translation results, and the parameters of the initial recommendation model are optimized iteratively.

[0141] Optionally, the terminology translation recommendation device also includes:

[0142] The optimization unit is used to optimize the parameters of the recommendation model in real time based on user feedback.

[0143] It should be noted that the terminology translation recommendation device provided in this embodiment of the invention can implement all the method steps implemented in the above-described terminology translation recommendation method embodiment, and can achieve the same technical effect. Therefore, the parts and beneficial effects that are the same as those in the method embodiment will not be described in detail here.

[0144] Figure 4 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 4As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440, wherein the processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a terminology translation recommendation method. This method includes: acquiring the text of the sentence to be translated and extracting its feature vector; inputting the feature vector of the sentence to be translated into a pre-built recommendation model to obtain the candidate terminology translation results corresponding to the sentence to be translated output by the recommendation model; determining the target candidate terminology translation result based on the candidate terminology translation results; acquiring user feedback information in real time and determining the recommendation priority of the target candidate terminology translation result based on the user feedback information; wherein the recommendation model is trained based on the feature vector of the sample sentence to be translated and the identifier of the candidate terminology translation result corresponding to the sample sentence to be translated.

[0145] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0146] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon. When executed by a processor, the computer program implements the terminology translation recommendation method provided by the above methods. The method includes: acquiring the text of the sentence to be translated and extracting the feature vector of the text of the sentence to be translated; inputting the feature vector of the text of the sentence to be translated into a pre-built recommendation model to obtain the candidate terminology translation result corresponding to the text of the sentence to be translated output by the recommendation model; determining the target candidate terminology translation result based on the candidate terminology translation result; acquiring user feedback information in real time and determining the recommendation priority of the target candidate terminology translation result based on the user feedback information; wherein the recommendation model is trained based on the feature vector of the sample text of the sentence to be translated and the identifier of the candidate terminology translation result corresponding to the sample text of the sentence to be translated.

[0147] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0148] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A terminology translation recommendation method, characterized in that, include: Obtain the text of the sentence to be translated, and extract the feature vector of the text of the sentence to be translated; The feature vector of the sentence text to be translated is input into a pre-built recommendation model to obtain the candidate term translation results corresponding to the sentence text to be translated output by the recommendation model. Based on the translation results of the candidate terms, the translation result of the target candidate term is determined; Real-time acquisition of user feedback information; and determination of the recommendation priority of the translation results of the target candidate terms based on the user feedback information. The recommendation model is trained based on the feature vector of the sample sentence text to be translated and the candidate term translation result identifiers corresponding to the sample sentence text to be translated. The determination of the target candidate term translation result based on the candidate term translation results includes: Multiple online translation results corresponding to the sentence text to be translated are obtained in real time from multiple online translation platforms; Based on the translation results of the candidate terms and the multiple online translation results, the translation result of the target candidate term is determined; The process of determining the target candidate term translation result based on the candidate term translation results and the multiple online translation results includes: Based on the aforementioned multiple online translation results, multiple online terminology translation results were obtained; The candidate term translation results are compared with the plurality of online term translation results, and the target candidate term translation result is determined from the candidate term translation results and the plurality of online term translation results; The step of determining the target candidate term translation result based on the candidate term translation results further includes: The confidence level of the candidate term translation results is calculated, and the candidate term translation results are filtered based on a preset confidence threshold to obtain the target candidate term translation results.

2. The terminology translation recommendation method according to claim 1, characterized in that, The step of determining the recommendation priority of the translation results of the target candidate terms based on the user feedback information includes: Based on the user feedback information, a score is obtained for the translation result of the target candidate term; Based on the scores of the translation results of the target candidate terms, the recommendation priority of the translation results of the target candidate terms is determined.

3. The terminology translation recommendation method according to claim 1, characterized in that, The process of determining the recommendation model includes: Obtain sample sentence text to be translated, and extract the feature vector of the sample sentence text to be translated. Determine the candidate term translation result identifiers corresponding to the sample sentence text to be translated; Using the feature vector of the sample sentence text to be translated as sample data, and the candidate term translation result identifier corresponding to the sample sentence text to be translated as sample label, an initial recommendation model is trained; After the initial recommendation model is trained, the recommendation model is obtained.

4. The terminology translation recommendation method according to claim 3, characterized in that, The initial recommendation model for training includes: The feature vector of the sample sentence text to be translated is input into the initial recommendation model to obtain the candidate term translation prediction result corresponding to the sample sentence text to be translated output by the initial recommendation model; Obtain the historical translation results of the sample sentence text to be translated; Based on the historical translation results, the translation results of historical terms are obtained; The candidate term translation prediction results are compared with the historical term translation results, and the parameters of the initial recommendation model are optimized iteratively.

5. The terminology translation recommendation method according to claim 3, characterized in that, After obtaining the recommendation model, the process further includes: Based on the user feedback information, the parameters of the recommendation model are optimized in real time.

6. A terminology translation recommendation device, characterized in that, include: The acquisition unit is used to acquire the text of the sentence to be translated and extract the feature vector of the text of the sentence to be translated. The recommendation unit is used to input the feature vector of the sentence text to be translated into a pre-built recommendation model to obtain the candidate term translation results corresponding to the sentence text to be translated output by the recommendation model; The first determining unit is used to determine the translation result of the target candidate term based on the translation results of the candidate terms; The second determining unit is used to acquire user feedback information in real time and determine the recommendation priority of the translation results of the target candidate terms based on the user feedback information. The recommendation model is trained based on the feature vector of the sample sentence text to be translated and the candidate term translation result identifiers corresponding to the sample sentence text to be translated. The determination of the target candidate term translation result based on the candidate term translation results includes: Multiple online translation results corresponding to the sentence text to be translated are obtained in real time from multiple online translation platforms; Based on the translation results of the candidate terms and the multiple online translation results, the translation result of the target candidate term is determined; The process of determining the target candidate term translation result based on the candidate term translation results and the multiple online translation results includes: Based on the aforementioned multiple online translation results, multiple online terminology translation results were obtained; The candidate term translation results are compared with the plurality of online term translation results, and the target candidate term translation result is determined from the candidate term translation results and the plurality of online term translation results; The step of determining the target candidate term translation result based on the candidate term translation results further includes: The confidence level of the candidate term translation results is calculated, and the candidate term translation results are filtered based on a preset confidence threshold to obtain the target candidate term translation results.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the terminology translation recommendation method as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the terminology translation recommendation method as described in any one of claims 1 to 5.

Images