Local characteristic culture translation service platform and method

Abstract

The invention discloses a local characteristic culture translation service platform and method, and relates to the technical field of translation services. The platform comprises a database, an intelligent recording module and a translation service module, the characteristic culture translation service platform is in communication connection with the user side and the translation side; local culture custom information is dynamically collected from an open culture database, a tourism website, social media and other platforms through the crawler technology, the collected information is cut into a plurality of culture data strips, a machine translation system is used for rapidly generating preliminary translations, and the translations are translated according to translation fields and keywords. The auditing task of the preliminary translation is accurately allocated to the most suitable translator; the keyword is extracted from the input information of the user side, and the correlation value between the input information and the culture data bar in the database is calculated according to the keyword, so that the translation result most matched with the user demand can be quickly found, and the retrieval accuracy and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of translation services, and in particular to a local cultural translation service platform and method. Background Art

[0002] With the development of globalization, cross-border cultural communication is becoming more and more frequent. Local culture is a unique business card of a region. Translation can promote the dissemination and exchange of culture and enhance cultural identity. The rise of tourism has brought more and more tourists into contact with the unique culture of different regions. Accurate and attractive translation helps tourists understand the local cultural background and enhance their travel experience.

[0003] At present, local cultural translation services usually include introductions to various cultural customs and their corresponding translations. This method can easily lead to cultural customs that are not included in the platform not having corresponding translations, and some translated content has not been reviewed, resulting in inaccurate or incomplete cultural information. Summary of the invention

[0004] The main purpose of the present invention is to provide a local cultural translation service platform and system to overcome the problems of limited coverage and uneven content quality mentioned in the above background technology.

[0005] To achieve the above-mentioned purpose, according to one aspect of the present invention, a local cultural translation service platform is provided, which includes: a database, an intelligent collection module and a translation service module;

[0006] The intelligent collection module uses crawler technology to collect local cultural and custom information on open network platforms, and cuts the information into several cultural data strips, each of which corresponds to a timestamp; based on the cultural data strips, the cultural data strips are translated and collected to obtain corresponding translations, and the cultural data strips and their corresponding translations are collected and stored in the database;

[0007] The translation service module provides translation query services to the user based on the input information of the user. The specific translation service process is as follows:

[0008] The input information of the user end is obtained, and the specific input information includes text, pictures or audio. The input information is obtained and keywords are extracted by natural language processing, image recognition or speech recognition. The keywords of the input information are analyzed for correlation with the keywords of the cultural data items in the database to obtain the correlation value between the input information and each cultural data item; the correlation threshold set by the correlation value is compared and analyzed. If the correlation value is greater than the set correlation threshold, the cultural data item corresponding to the correlation value is used as the relevant information item of the input information; the number of relevant data items is counted. If the number of relevant data items is greater than zero, the relevant data items and their final translations are arranged in descending order according to their corresponding correlation values, and sent to the user end as the search result of the input information;

[0009] If the number of relevant data strips is zero, the input information is integrated into a cultural data strip without translation, and sent to the intelligent collection module for translation task assignment to obtain its corresponding final translation;

[0010] When feedback information is received from the user, a re-examination is performed based on the feedback information.

[0011] Furthermore, the specific process of correlation analysis is as follows:

[0012] Compare the keywords of the input information with the keywords of the cultural data strips in the database to obtain the intersection keywords, count the number of keywords in the input information and the number of keywords in the cultural data strips, and record them as R1 and R2 respectively; use the TF-IDF algorithm to mine the importance scores of each intersection keyword in the input information and the cultural data strips, and record them as Hi and Ti respectively, where i = 1, 2, 3...I, I is a positive integer, I represents the total number of intersection keywords in the keywords of the input information and the keywords of the cultural data strips in the database, and i represents the number of any intersection keyword;

[0013] The number of keywords in the input information R1, the number of keywords in the cultural data strip R2, the importance score Hi of the intersection keywords in the input information, and the importance score Ti of the intersection keywords in the cultural data strip are normalized and their values ​​are taken. The values ​​are analyzed and calculated in a formula to obtain the correlation value RH. The specific calculation formula is:

[0014]

[0015] Among them, λ1 and λ2 are the set weight constants respectively.

[0016] Furthermore, the specific process of re-examination based on the feedback information is as follows:

[0017] Retrieve the matching values ​​of other translation terminals excluding those participating in the review of this cultural data strip, select the translation terminal with the largest matching value as the target translation terminal, and send the cultural data strip, the final translation and feedback information to the target translation terminal. If the translator of the target translation terminal finds that the feedback information is not valid after verification, there is no need to modify it, and the final translation is used as the review content of the target translation terminal; if the translator of the target translation terminal finds that the feedback information is valid after verification, the final translation of the cultural data strip is modified according to the feedback information to form a new final translation, and the new final translation is used as the review content of this review by the target translation terminal; the modification record of the cultural data strip is increased once, thereby obtaining the modification records of each cultural data strip, and integrating them into the translation terminals participating in the review, thereby obtaining several modification records corresponding to each translation terminal, and sending them to the database.

[0018] Furthermore, the modification records of each cultural data item are integrated into the translation end involved in the review. The specific integration process is as follows:

[0019] A modified cultural data strip is randomly selected, and the translation terminals that participated in the review of the cultural data strip before the revision are recorded as relevant translation terminals. The review content of the cultural data strip by each relevant translation terminal is extracted. If the review content is different from the final translation after the revision, the modification record of the cultural data strip by the relevant translation terminal is accumulated as one modification; thereby, the modifications of each modified cultural data strip are accumulated in the modification record of each relevant translation terminal.

[0020] Furthermore, the specific process of translating and collecting each cultural data item is as follows:

[0021] 5-1: For each cultural data item, use machine translation to translate it to generate a preliminary translation, so that each cultural data item corresponds to a preliminary translation; set a characteristic cultural translation service platform to communicate with a number of translators, each translator belongs to a translation field; each translation field corresponds to a number of translation terminals, each translation terminal corresponds to a translator, and the translation terminal of each field corresponds to a number of keywords;

[0022] 5-2: Use natural language processing technology to identify the field to which the cultural data bar belongs and extract the keywords therein, and use the TF-IDF algorithm to mine the importance score of each keyword in the text of the cultural data bar; retrieve several translation ends with the same field as the cultural data bar, and record them as preliminary translation ends; analyze the degree of overlap between the cultural data bar and each preliminary translation end to obtain the overlap value;

[0023] 5-3: Set each translation end to correspond to a historical translation value, and record it as η; obtain the number of translations in the translation list of the preliminary translation end in real time, and record it as N; normalize the overlap value Fα, the historical translation value η and the number of translations N and take their values, and calculate the values ​​in a formula to obtain the matching value NF between the cultural data strip and each preliminary translation end. The specific calculation formula is:

[0024]

[0025] Among them, β1, β2, and β3 are respectively set weight constants, and their values ​​are set by personnel in this field; the preliminary translation end with the largest matching value is selected as the target translation end of this cultural data strip, and this cultural data strip and its corresponding preliminary translation are sent to the target translation end, and the number of translations in the translation list of the target translation end is increased by one; the target translation end reviews the corresponding translation according to the cultural data strip to form a final translation, and collects the data strip and its corresponding final translation into the database for storage; whenever the translation end completes the translation review of a cultural data strip, the number of translations in the corresponding translation list is reduced by one, and an audit record is generated and sent to the database for storage; each specific audit record includes the audit start time, audit end time, data volume and audit content;

[0026] 5-4: Retrieve the latest historical translation records of each translation end from the database. The specific historical translation records include the number of review records, the review start time, review end time and data volume corresponding to each review record, modification records and the number of modifications corresponding to each modification record;

[0027] If the review record is equal to zero, it means that the translator is reviewing the translation for the first time, and the historical translation value is assigned a value of one;

[0028] If the audit record is not equal to zero, the historical audit is cumulatively analyzed to obtain the historical translation value, and it is updated to 5-3;

[0029] 5-5: Repeat the process of step 5-2 to step 5-4 for each cultural data strip in the chronological order of its corresponding timestamp, so as to complete the allocation of the review task of each cultural data strip.

[0030] Furthermore, the specific process of analyzing the degree of overlap between cultural data and each translation end is as follows:

[0031] Retrieve the keywords corresponding to the preliminary translation end, and compare them with the keywords corresponding to the cultural data strip, identify the intersection keywords and union keywords between the keywords of the preliminary translation end and the keywords of the cultural data strip, and count the number of intersection keywords and the number of union keywords, respectively, and record them as F1 and F2, thereby obtaining the importance score of each intersection keyword in the text of the cultural data strip;

[0032] The importance scores corresponding to the intersection keywords are averaged and recorded as F3. The average score is normalized with the number of intersection keywords F1 and the number of union keywords F2 and the numerical value is taken. The numerical value is analyzed and calculated in a formula to obtain the overlap value Fα between the cultural data strip and the preliminary translation end. The specific calculation formula is:

[0033]

[0034] Where α1 and α2 are respectively set weight constants, from which the overlap value Fα between the cultural data strip and each preliminary translation end can be obtained.

[0035] Furthermore, the specific method of generating the audit content is as follows:

[0036] Compare the preliminary translation with the final text obtained after review by the translation end. If there is any inconsistency, the final text obtained after review by the translation end will be used as the review content of this review; if they are exactly the same, the preliminary translation will be used as the review content.

[0037] Furthermore, the cumulative analysis process of historical audit is as follows:

[0038] 8-1: Analyze the audit efficiency of the translation end based on the audit records to obtain the audit efficiency value;

[0039] 8-2: Analyze the review quality of the translation end based on the modification records to obtain the review quality value;

[0040] 8-3: Normalize the audit efficiency value and the audit quality value and take their numerical values, and perform weighted calculation on the numerical values ​​to obtain the historical translation value.

[0041] Furthermore, the specific process of analyzing the audit efficiency of the translation end based on the audit records is as follows:

[0042] The time difference between the audit start time and the audit end time of each audit record is calculated to get the audit duration, and then the data volume is divided by the audit duration to get the audit speed, thus obtaining the audit speed corresponding to each audit record;

[0043] Compare and analyze the audit speed with the set speed range. When the audit speed is greater than the maximum value in the set speed range, an efficient audit is accumulated; when the audit speed is less than the set efficiency range, a medium-efficiency audit is accumulated; when the audit speed is less than the minimum value in the set speed range, an inefficient audit is accumulated; the cumulative number of efficient audits, medium-efficiency audits and inefficient audits are counted respectively, and recorded as S1, S2 and S3 respectively; the audit speed corresponding to each audit record is averaged to obtain the audit average value, recorded as S4;

[0044] The audit efficiency value Sb is obtained by formulating and analyzing the cumulative number of efficient audits S1, the cumulative number of medium-efficiency audits S2, the cumulative number of inefficient audits S3 and the audit mean S4. The specific calculation formula is:

[0045]

[0046] Wherein b1, b2, and b3 are respectively set weight constants, and b1>b2>b3>1.

[0047] Furthermore, the specific process of analyzing the review quality of the translation end based on the modification records is as follows:

[0048] The total number of audit records is recorded as L, and the number of modifications of each modification record is compared with the set number of modifications. When the number of modifications is greater than the maximum value in the set number interval, a high-frequency modification is accumulated; when the number of modifications is within the set number interval, a medium-frequency modification is accumulated; when the number of modifications is less than the minimum value in the set number interval, a low-frequency modification is accumulated; the total number of high-frequency modifications, medium-frequency modifications and low-frequency modifications are counted respectively, and recorded as P1, P2 and P3 respectively; the number of modifications corresponding to high-frequency modifications, medium-frequency modifications and low-frequency modifications are summed to obtain high-frequency values, medium-frequency values ​​and low-frequency values, and recorded as P4, P5 and P6 respectively;

[0049] The total number of audit records L, the total number of high-frequency modifications P1, the total number of medium-frequency modifications P2, the total number of low-frequency modifications P3, the high-frequency value P4, the medium-frequency value P5 and the low-frequency value P6 are calculated and analyzed to obtain the audit quality value LP; the specific calculation formula is:

[0050]

[0051] b4, b5, and b6 are respectively set weight constants, and b4>b5>b6>1.

[0052] In order to achieve the above object, according to another aspect of the present invention, a method for translating local culture is provided, the method comprising the following steps:

[0053] Q1: By using crawler technology to collect local cultural and custom information on open network platforms, the information is cut into several cultural data strips, each of which corresponds to a timestamp; according to the cultural data strips, the cultural data strips are translated and collected to obtain corresponding translations, and the cultural data strips and their corresponding translations are collected and saved; the specific process of translation and collection is as follows:

[0054] 5-1: For each cultural data item, use machine translation to translate it to generate a preliminary translation, so that each cultural data item corresponds to a preliminary translation; assume that there are several translators, each translation field corresponds to several translation terminals, each translation terminal corresponds to one translator, and the translators in each field correspond to several keywords;

[0055] 5-2: Use natural language processing technology to identify the field to which the cultural data bar belongs and extract the keywords therein, and use the TF-IDF algorithm to mine the importance score of each keyword in the text of the cultural data bar; retrieve several translators who are in the same field as the cultural data bar and record them as preliminary translators; analyze the degree of overlap between the cultural data bar and each preliminary translator to obtain the overlap value;

[0056] 5-3: Set each translator to correspond to a historical translation value, and record it as η; obtain the number of translations in the translation list of the preliminary translator in real time, and record it as N; normalize the overlap value η, the historical translation value Fα and the number of translations N and take their values, and perform formulaic calculation on the values ​​to obtain the matching value NF between the cultural data strip and each preliminary translator; the specific calculation formula is:

[0057]

[0058] Among them, β1, β2, and β3 are respectively set weight constants, and their values ​​are set by personnel in this field; the preliminary translator with the largest matching value is selected as the target translator of this cultural data strip, and this cultural data strip and its corresponding preliminary translation are sent to the target translator, and the number of translations in the translation list of the target translator is increased by one; the target translator reviews the corresponding translation according to the cultural data strip to form a final translation, and the data strip and its corresponding final translation are collected and saved; whenever the translator completes the translation review of a cultural data strip, the number of translations in the corresponding translation list is reduced by one, and an audit record is generated; each specific audit record includes the audit start time, audit end time, data volume and audit content;

[0059] 5-4: Retrieve the latest historical translation records of each translator from the database. The specific historical translation records include the number of review records, the review start time, review end time and data volume corresponding to each review record, modification records and the number of modifications corresponding to each modification record;

[0060] If the review record is equal to zero, it means that the translator is reviewing the translation for the first time, and the historical translation value is assigned a value of one;

[0061] If the audit record is not equal to zero, the historical audit is cumulatively analyzed to obtain the historical translation value, and it is updated to 5-3;

[0062] 5-5: Repeat the process of step 5-2 to step 5-4 for each cultural data strip in the order of the time stamps corresponding to them, so as to complete the distribution of the review tasks for each cultural data strip.

[0063] Q2: Provide translation query services based on user input information, or generate new translations based on user needs; the specific translation service process is as follows:

[0064] Obtaining user input information, which specifically includes text, pictures or audio, using natural language processing, image recognition or speech recognition to obtain input information and extract keywords, performing correlation analysis on the keywords of the input information and the keywords of the cultural data items to obtain correlation values ​​between the input information and each cultural data item; performing comparative analysis on the correlation threshold set for the correlation value, and if the correlation value is greater than the set correlation threshold, taking the cultural data item corresponding to the correlation value as the relevant information item of the input information; counting the number of relevant data items, and if the number of relevant data items is greater than zero, arranging the relevant data items and their final translations in descending order according to their corresponding correlation values, and sending them to the user as the search result of the input information;

[0065] If the number of relevant data strips is zero, the input information is integrated into a cultural data strip without translation, and sent to Q1 for translation and collection to obtain its corresponding final translation;

[0066] When receiving feedback from users, re-examine based on the feedback and execute Q3;

[0067] Q3: Retrieve the matching values ​​of other translators who have not participated in the review of this cultural data strip, select the translator with the largest matching value as the target translator, and send the cultural data strip, final translation and feedback information to the target translator. If the target translator finds that the feedback information is not valid after verification, there is no need to modify it, and the final translation will be used as the review content of the target translator; if the target translator finds that the feedback information is valid after verification, the final translation of the cultural data strip is modified according to the feedback information to form a new final translation, and the new final translation will be used as the review content of the target translator's review this time; the modification record of the cultural data strip is increased once, thereby obtaining the modification records of each cultural data strip, and integrating them into the translators involved in the review, thereby obtaining several modification records corresponding to each translator.

[0068] Beneficial effects of the present invention:

[0069] The present invention uses crawler technology to dynamically collect local cultural and custom information from open cultural databases, travel websites, social media and other platforms, thereby realizing automatic data collection and greatly reducing the cost and time of manual information collection. The collected information is cut into several cultural data strips, and a timestamp is added to each data strip, which helps to track the source and timeliness of the data and ensure the real-time and accuracy of the translated content.

[0070] The present invention uses machine translation systems such as Google Translate and Deep 1 to quickly generate preliminary translations, thereby improving translation speed. In particular, when processing large amounts of data, preliminary translation results can be quickly generated. According to the translation field and keywords, the review task of the preliminary translation is accurately assigned to the most suitable translator, ensuring that each translator can give full play to his or her professional expertise and guaranteeing the correctness of the preliminary translation review results, thereby improving the overall translation efficiency and quality. At the same time, the review efficiency and quality of each translation end are comprehensively evaluated through the historical translation records of the translation end, and the historical translation value of the translation end is dynamically updated based on multiple indicators such as the calculated review time, review speed, and number of modifications, thereby ensuring the continuous improvement of translation quality.

[0071] The present invention uses NLP, image recognition or speech recognition technology to extract keywords from the input information (text, picture or audio) of the user end, and calculates the correlation value between the input information and the cultural data items in the database based on the keywords, so as to quickly find the translation result that best matches the user's needs, thereby improving the accuracy and efficiency of the search; if there is no cultural data item matching the user input in the database, the input information can be integrated into a new cultural data item, and the final translation can be generated through the intelligent collection module, thereby achieving high flexibility of the translation service; at the same time, the user gives feedback on the search results and points out the problems in the translation, and the platform will reallocate the translation tasks according to the feedback information to ensure the accuracy of the final translation, and accumulate the updated final translation by modifying the records;

[0072] In summary, the intelligent collection module and translation service module have significantly improved the speed, accuracy and user experience of local cultural translation through automated data collection, precise task allocation, real-time retrieval and efficient feedback and correction mechanism. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0074] Figure 1 is a schematic diagram of a platform of the present invention;

[0075] Figure 2 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0076] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0077] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0078] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so as to describe the embodiments of the present invention described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0079] It should be noted that the relevant information (including but not limited to user information) and data (including but not limited to data for display, data for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.

[0080] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0081] According to an embodiment of the present invention, Figure 1 As shown, a local characteristic cultural translation service platform is provided, which includes: a database, an intelligent collection module and a translation service module; the characteristic cultural translation service platform is connected to the user end and the translation end in communication, and the user can access the characteristic cultural translation service platform through the user end (such as a PC or a mobile phone, etc.) to obtain the translation in real time;

[0082] The intelligent collection module uses crawler technology to dynamically collect local cultural customs information from open cultural databases, travel websites, social media and other platforms, and cuts the information into several cultural data strips. Each cultural data strip corresponds to a timestamp, which is the time when the intelligent collection module obtains it from the platform. By cutting the collected information into several cultural data strips and adding a timestamp to each data strip, it helps to track the source and timeliness of the data, ensuring the real-time and accuracy of the translated content.

[0083] According to the cultural data strips, they are translated to obtain corresponding translations, and the cultural data strips and their corresponding translations are collected into a database for storage; it should be noted that, since English is an international language, the specific application scenario of the present invention is mainly Chinese-English translation, that is, the local cultural information is in Chinese and the translation is in English; the specific translation collection process is:

[0084] Step 1: For each cultural data item, use machine translation (such as Google Translate, Deep L, etc.) to translate it to generate a preliminary translation, so that each cultural data item corresponds to a preliminary translation; set the characteristic cultural translation service platform to communicate and connect several translation fields, each translation field corresponds to several translation terminals, and each translator belongs to a translation field; the translation terminal of each field corresponds to several keywords, and the keywords can reflect the professional expertise of the translator and improve the accuracy and rationality of the translation task allocation; it should be noted that specific translation fields include art, folklore, history, and food, etc. For example, translators in the folklore field can add keywords based on their expertise, and specific keywords can be: festivals, customs, weddings, etc.; when adding a new translation field, translation terminal or translator, you only need to update the corresponding keyword library and field configuration, and the platform has strong scalability;

[0085] By using crawler technology to dynamically collect information on local cultural customs from open cultural databases, travel websites, social media and other platforms, we have achieved automated data collection, greatly reducing the cost and time of manual information collection.

[0086] Step 2: Use natural language processing (NLP) technology to identify the field to which the cultural data bar belongs and extract the keywords therein, and use the TF-IDF algorithm to mine the importance score of each keyword in the text of the cultural data bar;

[0087] Retrieve several translation ends in the same field as the cultural data strip, and record them as preliminary translation ends; retrieve the keywords corresponding to the preliminary translation ends, and compare them with the keywords corresponding to the cultural data strip, identify the intersection keywords and union keywords between the keywords of the preliminary translation ends and the keywords of the cultural data strip, and count the number of intersection keywords and the number of union keywords, and record them as F1 and F2 respectively; specifically, the keywords of the preliminary translation ends are: A1, A2, A3, B4, B5 and B6; the keywords of the cultural data strip are: The keywords of the cultural data strip are: A0, A2, A3, B1, B2 and B4, then the intersection keywords between the keywords of the preliminary translation end and the keywords of the cultural data strip are: A2, A3, B4, and the union keywords are: A0, A1, A2, A3, B1, B2, B4, B5 and B6. The number of intersection keywords is 3, and the number of union keywords is 9. It should be noted that the greater the number of intersection keywords, the greater the degree of overlap between the two. Therefore, the importance score of each intersection keyword in the text of the cultural data strip can be obtained;

[0088] The importance scores corresponding to the intersection keywords are averaged and recorded as F3. The average score is normalized with the number of intersection keywords F1 and the number of union keywords F2 and the numerical value is taken. The numerical value is analyzed and calculated in a formula to obtain the overlap value Fα between the cultural data strip and the preliminary translation end. The specific calculation formula is:

[0089]

[0090] Wherein α1 and α2 are respectively set weight constants, and their values ​​are set by those skilled in the art. For example, α1 is , α2 is 2.391, and α3 is 3.055; thus, the overlap value Fα between the cultural data strip and each preliminary translation end can be obtained;

[0091] Step 3: Set each translation end to correspond to a historical translation value, and record it as η; obtain the number of translations in the translation list of the preliminary translation end in real time, and record it as N; normalize the overlap value Fα, the historical translation value η and the number of translations N and take their values, and perform formulaic calculation on the values ​​to obtain the matching value NF between the cultural data strip and each preliminary translation end. The specific calculation formula is:

[0092]

[0093] Among them, β1, β2, and β3 are respectively set weight constants, and their values ​​are set by those in this field. For example, β1 is 0.25, β2 is 0.41, and β3 is 0.34; it can be seen from the formula that the larger the historical translation value is, the higher the translation quality of the preliminary translation end is, the faster the translation speed is, and the larger the matching value is; the larger the overlap value is, the greater the correlation between the cultural data strip and the preliminary translation end is, and the larger the matching value is; the smaller the number of translations in the translation list of the preliminary translation end is, the larger the matching value is; the preliminary translation end with the largest matching value is selected as the target translation end of this cultural data strip, and this cultural data strip and its corresponding preliminary translation are sent to the target translation end, and the number of translations in the translation list of the target translation end is increased by one; the target translation end is based on According to the cultural data strip, the corresponding translation is reviewed to form the final translation. The specific review includes review correction or adding annotations or notes (this method of adding notes or notes can make outsiders or foreigners who do not understand the local culture more clearly understand, which is also more beneficial to the promotion of the culture at home and abroad). The data strip and its corresponding final translation are stored in the database; whenever the translation end completes the translation review of a cultural data strip, the number of translations in the corresponding translation list is reduced by one, and an audit record is generated and sent to the database for storage; each specific audit record includes the audit start time, audit end time and data volume (data volume refers to the number of bytes of the cultural data strip) and audit content. The specific generation method of the audit content is:

[0094] Compare the preliminary translation with the final text obtained after review by the translation end. If there is any inconsistency, the final text obtained after review by the translation end will be used as the review content of this review; if they are exactly the same, it means that the translator on the translation end fully agrees with the preliminary translation and uses the preliminary translation as the review content;

[0095] Step 4: retrieve the latest historical translation records of each translation end from the database. The specific historical translation records include the number of review records, the review start time, review end time and data volume corresponding to each review record, modification records and the number of modifications corresponding to each modification record;

[0096] If the review record is equal to zero, it means that the translator is reviewing the translation for the first time, and the historical translation value is assigned a value of one;

[0097] If the audit record is not equal to zero, a cumulative analysis of historical audits is performed, specifically:

[0098] The time difference between the audit start time and the audit end time of each audit record is calculated to obtain the audit duration, and then the data volume is divided by the audit duration to obtain the audit speed, thereby obtaining the audit speed corresponding to each audit record; the audit speed is compared and analyzed with the set speed range. When the audit speed is greater than the maximum value in the set speed range, an efficient audit is accumulated; when the audit speed is less than the set efficiency range, a medium-efficiency audit is accumulated; when the audit speed is less than the minimum value in the set speed range, an inefficient audit is accumulated; the cumulative number of efficient audits, medium-efficiency audits and inefficient audits are counted respectively, and they are recorded as S1, S2 and S3 respectively; the audit speed corresponding to each audit record is averaged to obtain the audit mean value recorded as S4; the cumulative number of efficient audits S1, the cumulative number of medium-efficiency audits S2, the cumulative number of inefficient audits S3 and the audit mean S4 are formulaically calculated and analyzed to obtain the audit efficiency value Sb. The specific calculation formula is:

[0099]

[0100] b1, b2, and b3 are respectively the set weight constants, and b1>b2>b3>1. From the formula, we can see that in the historical translation records of the translation end, the faster the review speed, the greater the corresponding review efficiency value, which is closer to the actual situation than simply directly calculating the average value, and can more accurately measure the efficiency status of the translator's review;

[0101] The total number of audit records is recorded as L, and the number of modifications of each modification record is compared with the set number of modifications. When the number of modifications is greater than the maximum value in the set number interval, a high-frequency modification is accumulated; when the number of modifications is within the set number interval, a medium-frequency modification is accumulated; when the number of modifications is less than the minimum value in the set number interval, a low-frequency modification is accumulated; the total number of high-frequency modifications, medium-frequency modifications and low-frequency modifications are counted respectively, and recorded as P1, P2 and P3 respectively; the number of modifications corresponding to high-frequency modifications, medium-frequency modifications and low-frequency modifications are summed to obtain high-frequency values, medium-frequency values ​​and low-frequency values, and recorded as P4, P5 and P6 respectively;

[0102] The total number of audit records L, the total number of high-frequency modifications P1, the total number of medium-frequency modifications P2, the total number of low-frequency modifications P3, the high-frequency value P4, the medium-frequency value P5 and the low-frequency value P6 are calculated and analyzed to obtain the audit quality value LP; the specific calculation formula is:

[0103]

[0104] b4, b5, and b6 are respectively the set weight constants, and b4>b5>b6>1. It can be seen from the formula that when the total number of audit records is smaller and the number of modification records is more, the audit quality value is smaller; when the number of modifications of each modification record is more, the audit quality value is smaller; specifically, the historical translation records of a certain translation end include 22 audit records, namely J1, J2, J3...J20, and 4 modification records, namely; J6, J9, J10, J18; among them, J6 was modified 5 times, J9 was modified 1 time, J10 was modified 2 times, and J18 was modified 7 times;

[0105] Normalize the audit efficiency value and the audit quality value and take their numerical values, perform weighted calculation on the numerical values ​​to obtain the historical translation value, and update it to step three; it should be noted that the greater the audit efficiency and the greater the audit quality, the greater the historical translation value;

[0106] Step 5: Repeat the process of step 2 to step 4 for each cultural data strip in the order of the time of its corresponding timestamp, so as to complete the allocation of the review task of each cultural data strip;

[0107] By using machine translation systems such as Google Translate and Deep L to quickly generate preliminary translations, the translation speed is improved, especially when processing large amounts of data, preliminary translation results can be generated quickly; according to the translation field and keywords, the review task of the preliminary translation is accurately assigned to the most suitable translator to ensure that each translator can give full play to his or her professional expertise and guarantee the correctness of the preliminary translation review results, thereby improving the overall translation efficiency and quality; at the same time, the review efficiency and quality of each translation end are comprehensively evaluated through the historical translation records of the translation end, and the historical translation value of the translation end is dynamically updated based on multiple indicators such as the calculation of the review time, review speed, and number of modifications, to ensure the continuous improvement of translation quality.

[0108] The translation service module provides translation query services to the user terminal based on the user's input information, or generates new translations according to the user's needs. The user terminal can provide feedback based on the translation, and trigger manual correction based on the user's feedback. The specific process is as follows:

[0109] The input information of the user end is obtained, and the specific input information includes text, picture or audio. Natural language processing (NLP) or image recognition or speech recognition is used to obtain the input information and extract keywords. For example, for a picture of dragon boat racing, the extracted keywords may be Dragon Boat Festival, dragon boat racing, Qu Yuan, zongzi, etc.; the keywords of the input information are compared with the keywords of the cultural data strips in the database to obtain intersection keywords, and the number of keywords in the input information and the number of keywords in the cultural data strips are counted respectively, and they are recorded as R1 and R2 respectively; the TF-IDF algorithm is used to mine the importance scores of each intersection keyword in the input information and the cultural data strip respectively, and they are recorded as Hi and Ti respectively, where i=1,2,3...I, and I is a positive integer, I represents the total number of intersection keywords existing in the keywords of the input information and the keywords of the cultural data strips in the database, and i represents the number of any intersection keyword;

[0110] The number of keywords in the input information R1, the number of keywords in the cultural data strip R2, the importance score Hi of the intersection keywords in the input information, and the importance score Ti of the intersection keywords in the cultural data strip are normalized and their values ​​are taken. The values ​​are analyzed and calculated in a formula to obtain the correlation value RH. The specific calculation formula is:

[0111]

[0112] Among them, λ1 and λ2 are respectively set weight constants. It can be seen from the formula that when the proportion of intersection keywords in the keywords of the input information is greater, and the proportion of intersection keywords in the keywords of the cultural data strip is greater, it means that the overlap degree of the keywords of the input information and the cultural data strip is higher, and the correlation value is greater; when the importance score of the intersection keywords in the input information and the cultural data strip is greater, it means that the content overlap degree of the input information and the cultural data strip is higher, and the correlation value is greater; thus, the correlation value between the input information and each cultural data strip can be obtained, and it is compared and analyzed with the set correlation threshold. If the correlation value is greater than the set correlation threshold, it means that the cultural data strip has a higher degree of correlation with the input information, and the cultural data strip corresponding to the correlation value is regarded as the relevant information strip of the input information;

[0113] Counting the number of relevant data items, if the number of relevant data items is greater than zero, arranging the relevant data items and their final translations in descending order according to their corresponding relevant values, and sending them to the user end as the search result of the input information;

[0114] If the number of relevant data strips is zero, the input information is integrated into a cultural data strip without translation, and sent to the intelligent collection module for translation task assignment to obtain its corresponding final translation;

[0115] When the user receives the search results of the input information, he can give feedback on the search results and upload the feedback information. The specific feedback information is the part of the translation or the remarks or annotations that the user thinks has problems;

[0116] When receiving feedback information from the user end, the matching values ​​of other translation ends except those participating in the review of this cultural data strip are retrieved, and the translation end with the largest matching value is selected as the target translation end, and the cultural data strip, the final translation and the feedback information are sent to the target translation end. If the translator of the target translation end finds that the feedback information is not valid after verification, there is no need to modify it, and the final translation is used as the review content of the target translator; if the translator of the target translation end verifies the feedback information, the final translation of the cultural data strip is modified according to the feedback information to form a new final translation, and the new final translation is used as the review content of the target translator's review this time; the modification record of the cultural data strip is increased once, thereby obtaining the modification records of each cultural data strip, and integrating them into the translation ends participating in the review, thereby obtaining several modification records corresponding to each translation end, and sending them to the database. The specific integration process is as follows:

[0117] A modified cultural data strip is randomly selected, and the translation terminals that participated in the review of the cultural data strip before the revision are recorded as relevant translation terminals. The review content of each relevant translation terminal on the cultural data strip is extracted. If the review content is different from the final translation after the revision, the modification record of the cultural data strip by the relevant translation terminal is accumulated as one modification. Thus, the modification of each modified cultural data strip is accumulated into the modification record of each relevant translation terminal. Specifically, (1) there is a cultural data strip J2, and the translation terminals that participated in the review of J2 are D1, D2 and D3. The review content of D1 on the cultural data strip J2 is C1, the review content of D2 on the cultural data strip J2 is C2, and the review content of D3 on the cultural data strip J2 is C3. Then the translation terminal D1 The revision record of cultural data item J2 is: the number of revisions is 2 times; the revision record of cultural data item J2 by ​​translation end D2 is: the number of revisions is 1 time; (2) There is a cultural data item J2, and the translation ends D1, D2 and D3 are involved in the review of J2; the review content of cultural data item J2 by ​​D1 is C1, the review content of cultural data item J2 by ​​D2 is C1 (the translator of translation end D2 feels that there is no problem with the final translation of translation end D1 after review, so it is not revised), and the review content of cultural data item J2 by ​​D3 is C2; then the revision record of cultural data item J2 by ​​translation end D1 is: the number of revisions is 1 time; the revision record of cultural data item J2 by ​​translation end D2 is: the number of revisions is 1 time;

[0118] By using NLP, image recognition or speech recognition technology to extract keywords from input information (text, pictures or audio), and calculating the correlation value between the input information and the cultural data items in the database, it is possible to quickly find the translation result that best matches the user's needs, thereby improving the accuracy and efficiency of retrieval; if there is no cultural data item matching the user's input in the database, the input information can be integrated into a new cultural data item, and the final translation can be generated through the intelligent collection module, thus achieving high flexibility of the translation service; at the same time, users provide feedback on the search results and point out problems in the translation. The platform will reallocate translation tasks based on the feedback information to ensure the accuracy of the final translation, and accumulate the updated translation through modification records.

[0119] The database integrates the audit records and modification records of each translation end received into historical translation records and stores them;

[0120] According to an embodiment of the present invention, Figure 2 As shown, a local cultural translation service method is also provided, which includes the following steps:

[0121] Q1: By using crawler technology to collect local cultural and custom information on open network platforms, the information is cut into several cultural data strips, each of which corresponds to a timestamp; according to the cultural data strips, the cultural data strips are translated and collected to obtain corresponding translations, and the cultural data strips and their corresponding translations are collected and saved; the specific process of translation and collection is as follows:

[0122] 5-1: For each cultural data item, use machine translation to translate it to generate a preliminary translation, so that each cultural data item corresponds to a preliminary translation; assume that there are several translators, each translation field corresponds to several translation terminals, each translation terminal corresponds to one translator, and the translators in each field correspond to several keywords;

[0123] 5-2: Use natural language processing technology to identify the field to which the cultural data bar belongs and extract the keywords therein, and use the TF-IDF algorithm to mine the importance score of each keyword in the text of the cultural data bar; retrieve several translators who are in the same field as the cultural data bar and record them as preliminary translators; analyze the degree of overlap between the cultural data bar and each preliminary translator to obtain the overlap value;

[0124] 5-3: Set each translator to correspond to a historical translation value, and record it as η; obtain the number of translations in the translation list of the preliminary translator in real time, and record it as N; normalize the overlap value η, the historical translation value Fα and the number of translations N and take their values, and perform formulaic calculation on the values ​​to obtain the matching value NF between the cultural data strip and each preliminary translator; the specific calculation formula is:

[0125]

[0126] Among them, β1, β2, and β3 are respectively set weight constants, and their values ​​are set by personnel in this field; the preliminary translator with the largest matching value is selected as the target translator of this cultural data strip, and this cultural data strip and its corresponding preliminary translation are sent to the target translator, and the number of translations in the translation list of the target translator is increased by one; the target translator reviews the corresponding translation according to the cultural data strip to form a final translation, and the data strip and its corresponding final translation are collected and saved; whenever the translator completes the translation review of a cultural data strip, the number of translations in the corresponding translation list is reduced by one, and an audit record is generated; each specific audit record includes the audit start time, audit end time, data volume and audit content;

[0127] 5-4: Retrieve the latest historical translation records of each translator from the database. The specific historical translation records include the number of review records, the review start time, review end time and data volume corresponding to each review record, modification records and the number of modifications corresponding to each modification record;

[0128] If the review record is equal to zero, it means that the translator is reviewing the translation for the first time, and the historical translation value is assigned a value of one;

[0129] If the audit record is not equal to zero, the historical audit is cumulatively analyzed to obtain the historical translation value, and it is updated to 5-3;

[0130] 5-5: Repeat the process of step 5-2 to step 5-4 for each cultural data strip in the order of the time stamps corresponding to them, so as to complete the distribution of the review tasks for each cultural data strip.

[0131] Q2: Provide translation query services based on user input information, or generate new translations based on user needs; the specific translation service process is as follows:

[0132] Obtaining user input information, which specifically includes text, pictures or audio, using natural language processing, image recognition or speech recognition to obtain input information and extract keywords, performing correlation analysis on the keywords of the input information and the keywords of the cultural data items to obtain correlation values ​​between the input information and each cultural data item; performing comparative analysis on the correlation threshold set for the correlation value, and if the correlation value is greater than the set correlation threshold, taking the cultural data item corresponding to the correlation value as the relevant information item of the input information; counting the number of relevant data items, and if the number of relevant data items is greater than zero, arranging the relevant data items and their final translations in descending order according to their corresponding correlation values, and sending them to the user as the search result of the input information;

[0133] If the number of relevant data strips is zero, the input information is integrated into a cultural data strip without translation, and sent to Q1 for translation and collection to obtain its corresponding final translation;

[0134] When receiving feedback from users, re-examine based on the feedback and execute Q3;

[0135] Q3: Retrieve the matching values ​​of other translators who have not participated in the review of this cultural data strip, select the translator with the largest matching value as the target translator, and send the cultural data strip, final translation and feedback information to the target translator. If the target translator finds that the feedback information is not valid after verification, there is no need to modify it, and the final translation will be used as the review content of the target translator; if the target translator finds that the feedback information is valid after verification, the final translation of the cultural data strip is modified according to the feedback information to form a new final translation, and the new final translation will be used as the review content of the target translator's review this time; the modification record of the cultural data strip is increased once, thereby obtaining the modification records of each cultural data strip, and integrating them into the translators involved in the review, thereby obtaining several modification records corresponding to each translator.

[0136] The above are only embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A local cultural translation service platform, comprising a cultural translation service platform, wherein the cultural translation service platform is connected to a user terminal and a translation terminal by communication; characterized in that: The characteristic cultural translation service platform includes: intelligent collection module and translation service module; The intelligent collection module uses crawler technology to collect local cultural and custom information on open network platforms, and cuts the information into several cultural data strips, each of which corresponds to a timestamp; based on the cultural data strips, the cultural data strips are translated and collected to obtain corresponding translations, and the cultural data strips and their corresponding translations are collected and stored in the database; The translation service module provides translation query services to the user based on the input information of the user. The specific translation service process is as follows: The input information of the user end is obtained, and the specific input information includes text, pictures or audio. The input information is obtained and keywords are extracted by natural language processing, image recognition or speech recognition. The keywords of the input information are analyzed for correlation with the keywords of the cultural data items in the database to obtain the correlation value between the input information and each cultural data item; the correlation threshold set by the correlation value is compared and analyzed. If the correlation value is greater than the set correlation threshold, the cultural data item corresponding to the correlation value is used as the relevant information item of the input information; the number of relevant data items is counted. If the number of relevant data items is greater than zero, the relevant data items and their final translations are arranged in descending order according to their corresponding correlation values, and sent to the user end as the search result of the input information; If the number of relevant data strips is zero, the input information is integrated into a cultural data strip without translation, and sent to the intelligent collection module for translation task assignment to obtain its corresponding final translation; When feedback information is received from the user, a re-examination is performed based on the feedback information.

2. A local cultural translation service platform according to claim 1, characterized in that: The specific process of correlation analysis is as follows: Compare the keywords of the input information with the keywords of the cultural data strips in the database to obtain the intersection keywords, count the number of keywords in the input information and the number of keywords in the cultural data strips, and use the TF-IDF algorithm to mine the importance scores of each intersection keyword in the input information and the cultural data strips; The number of keywords in the input information, the number of keywords in the cultural data strip, the importance score of the intersection keywords in the input information and the importance score of the intersection keywords in the cultural data strip are normalized and their values ​​are taken, and the values ​​are analyzed by formula calculation to obtain the relevant value.

3. A local cultural translation service platform according to claim 2, characterized in that: The specific process of re-examination based on feedback information is as follows: Retrieve the matching values ​​of other translation terminals excluding those involved in the review of this cultural data strip, select the translation terminal with the largest matching value as the target translation terminal, and send the cultural data strip, the final translation and feedback information to the target translation terminal. If the translator of the target translation terminal finds that the feedback information is not valid after verification, there is no need to modify it, and the final translation is used as the review content of the target translation terminal; If the feedback information provided by the translator at the target translation end is confirmed after verification, the final translation of the cultural data strip will be modified based on the feedback information to form a new final translation, and the new final translation will be used as the audit content of this audit by the target translation end; the modification record of the cultural data strip will be increased once, thereby obtaining the modification records of each cultural data strip, and integrating them into the translation ends participating in the audit, thereby obtaining several modification records corresponding to each translation end and sending them to the database.

4. A local cultural translation service platform according to claim 3, characterized in that: Integrate the modification records of each cultural data item to the translation end involved in the review. The specific integration process is as follows: A modified cultural data strip is randomly selected, and the translation terminals that participated in the review of the cultural data strip before the revision are recorded as relevant translation terminals. The review content of the cultural data strip by each relevant translation terminal is extracted. If the review content is different from the final translation after the revision, the modification record of the cultural data strip by the relevant translation terminal is accumulated as one modification; thereby, the modifications of each modified cultural data strip are accumulated in the modification record of each relevant translation terminal.

5. A local cultural translation service platform according to claim 1, characterized in that: The specific process of translating and collecting each cultural data item is as follows: 5-1: For each cultural data item, use machine translation to translate it to generate a preliminary translation, so that each cultural data item corresponds to a preliminary translation; A characteristic cultural translation service platform is set up to communicate with several translators, each of whom belongs to a translation field; each translation field corresponds to several translation terminals, each translation terminal corresponds to a translator, and the translation terminal of each field corresponds to several keywords; 5-2: Use natural language processing technology to identify the field to which the cultural data bar belongs and extract the keywords therein, and use the TF-IDF algorithm to mine the importance score of each keyword in the text of the cultural data bar; retrieve several translation ends with the same field as the cultural data bar, and record them as preliminary translation ends; analyze the degree of overlap between the cultural data bar and each preliminary translation end to obtain the overlap value; 5-3: Set each translation end to correspond to a historical translation value, obtain the number of translations in the translation list of the preliminary translation end in real time, normalize the overlap value, historical translation value and translation number and take their values, and perform formula calculation on the values ​​to obtain the matching value between the cultural data strip and each preliminary translation end; The preliminary translation end with the largest matching value is selected as the target translation end of the cultural data strip, and the cultural data strip and its corresponding preliminary translation are sent to the target translation end, and the number of translations in the translation list of the target translation end is increased by one; the target translation end reviews the corresponding translation according to the cultural data strip to form a final translation, and the data strip and its corresponding final translation are recorded in the database for storage; whenever the translation end completes the translation review of a cultural data strip, the number of translations in the corresponding translation list is reduced by one, and an audit record is generated and sent to the database for storage; each specific audit record includes the audit start time, audit end time, data volume and audit content; 5-4: Retrieve the latest historical translation records of each translation end from the database. The specific historical translation records include the number of review records, the review start time, review end time and data volume corresponding to each review record, modification records and the number of modifications corresponding to each modification record; If the review record is equal to zero, it means that the translator is reviewing the translation for the first time, and the historical translation value is assigned a value of one; If the audit record is not equal to zero, the historical audit is cumulatively analyzed to obtain the historical translation value, and it is updated to 5-3; 5-5: Repeat the process of step 5-2 to step 5-4 for each cultural data strip in the chronological order of its corresponding timestamp, so as to complete the allocation of the review task of each cultural data strip.

6. A local cultural translation service platform according to claim 5, characterized in that: The specific process of analyzing the degree of overlap between cultural data and each translation end is as follows: Retrieving the keywords corresponding to the preliminary translation end, and comparing them with the keywords corresponding to the cultural data strip, identifying the intersection keywords and union keywords between the keywords of the preliminary translation end and the keywords of the cultural data strip, and counting the number of intersection keywords and the number of union keywords respectively, thereby obtaining the importance score of each intersection keyword in the text of the cultural data strip; The importance scores corresponding to the intersection keywords are averaged to obtain the mean score, which is then normalized with the number of intersection keywords and the number of union keywords and its value is taken. The value is then analyzed by formulaic calculation to obtain the overlap value between the cultural data strip and the preliminary translation end.

7. A local cultural translation service platform according to claim 6, characterized in that: The specific method of generating the audit content is as follows: Compare the preliminary translation with the final text obtained after review by the translation end. If there is any inconsistency, the final text obtained after review by the translation end will be used as the review content of this review; If they are exactly the same, the preliminary translation will be used as the review content.

8. A local cultural translation service platform according to claim 7, characterized in that: The cumulative analysis process of historical audit is as follows: 8-1: Analyze the audit efficiency of the translation end based on the audit records to obtain the audit efficiency value; 8-2: Analyze the review quality of the translation end based on the modification records to obtain the review quality value; 8-3: Normalize the audit efficiency value and the audit quality value and take their numerical values, and perform weighted calculation on the numerical values ​​to obtain the historical translation value.

9. A local cultural translation service platform according to claim 8, characterized in that: The specific process of analyzing the audit efficiency of the translation end based on the audit records is as follows: The time difference between the audit start time and the audit end time of each audit record is calculated to get the audit duration, and then the data volume is divided by the audit duration to get the audit speed, thus obtaining the audit speed corresponding to each audit record; Compare and analyze the audit speed with the set speed range to divide the audit records corresponding to each audit speed into efficient audit, medium-efficiency audit and inefficient audit, count the cumulative number of efficient audit, medium-efficiency audit and inefficient audit respectively, and calculate the average of the audit speed corresponding to each audit record to obtain the audit mean; The audit efficiency value is obtained by formulating and analyzing the cumulative number of efficient audits, the cumulative number of medium-efficient audits, the cumulative number of inefficient audits and the audit mean; The specific process of analyzing the review quality of the translation end based on the modification records is as follows: Count the total number of audit records, compare and analyze the number of modifications of each modification record with the set number of modifications to divide the modification records corresponding to the number of modifications into high-frequency modifications, medium-frequency modifications, and low-frequency modifications; The total number of high-frequency modifications, medium-frequency modifications, and low-frequency modifications are counted respectively, and the number of modifications corresponding to the high-frequency modifications, the medium-frequency modifications, and the low-frequency modifications are summed up to obtain the high-frequency value, the medium-frequency value, and the low-frequency value; The total number of audit records, the total number of high-frequency modifications, the total number of medium-frequency modifications, the total number of low-frequency modifications, the high-frequency value, the medium-frequency value and the low-frequency value are formulated and analyzed to obtain the audit quality value.

10. A method for providing a local cultural translation service, characterized in that Applied to a local cultural translation service platform as described in any one of claims 1-9, the method comprises the following steps: Q1: Use crawler technology to collect local cultural and custom information on open network platforms, and cut the information into several cultural data strips, each of which corresponds to a timestamp; translate and collect the cultural data strips to obtain corresponding translations, and collect and save the cultural data strips and their corresponding translations; the specific process of translation and collection is as follows: 5-1: For each cultural data item, use machine translation to translate it to generate a preliminary translation, so that each cultural data item corresponds to a preliminary translation; assume that there are several translators, each translation field corresponds to several translation terminals, each translation terminal corresponds to one translator, and the translators in each field correspond to several keywords; 5-2: Use natural language processing technology to identify the field to which the cultural data bar belongs and extract the keywords therein, and use the TF-IDF algorithm to mine the importance score of each keyword in the text of the cultural data bar; retrieve several translators who are in the same field as the cultural data bar and record them as preliminary translators; analyze the degree of overlap between the cultural data bar and each preliminary translator to obtain the overlap value; 5-3: Set a historical translation value for each translator, obtain the number of translations in the translation list of the preliminary translator in real time, normalize the overlap value, historical translation value and translation number and take their values, and calculate the values ​​in a formula to obtain the matching value between the cultural data strip and each preliminary translator; select the preliminary translator with the largest matching value as the target translator of this cultural data strip, and send this cultural data strip and its corresponding preliminary translation to the target translator, and the number of translations in the translation list of the target translator increases by one; the target translator reviews the corresponding translation based on the cultural data strip to form a final translation, and collects and saves the data strip and its corresponding final translation; whenever the translator completes the translation review of a cultural data strip, the number of translations in the corresponding translation list decreases by one, and a review record is generated; each specific review record includes the review start time, review end time, data volume and review content; 5-4: Retrieve the latest historical translation records of each translator from the database. The specific historical translation records include the number of review records, the review start time, review end time and data volume corresponding to each review record, modification records and the number of modifications corresponding to each modification record; If the review record is equal to zero, it means that the translator is reviewing the translation for the first time, and the historical translation value is assigned a value of one; If the audit record is not equal to zero, the historical audit is cumulatively analyzed to obtain the historical translation value, and it is updated to 5-3; 5-5: Repeat the process of step 5-2 to step 5-4 for each cultural data strip in the order of the time stamps corresponding to them, so as to complete the distribution of the review tasks for each cultural data strip. Q2: Provide translation query services based on user input information, or generate new translations based on user needs; the specific translation service process is as follows: Obtaining user input information, which specifically includes text, pictures or audio, using natural language processing, image recognition or speech recognition to obtain input information and extract keywords, performing correlation analysis on the keywords of the input information and the keywords of the cultural data items to obtain correlation values ​​between the input information and each cultural data item; performing comparative analysis on the correlation threshold set for the correlation value, and if the correlation value is greater than the set correlation threshold, taking the cultural data item corresponding to the correlation value as the relevant information item of the input information; counting the number of relevant data items, and if the number of relevant data items is greater than zero, arranging the relevant data items and their final translations in descending order according to their corresponding correlation values, and sending them to the user as the search result of the input information; If the number of relevant data strips is zero, the input information is integrated into a cultural data strip without translation, and sent to Q1 for translation and collection to obtain its corresponding final translation; When receiving feedback from users, re-examine based on the feedback and execute Q3; Q3: Retrieve the matching values ​​of other translators who have not participated in the review of this cultural data strip, select the translator with the largest matching value as the target translator, and send the cultural data strip, final translation and feedback information to the target translator. If the target translator finds that the feedback information is not valid after verification, there is no need to modify it, and the final translation will be used as the review content of the target translator; if the target translator finds that the feedback information is valid after verification, the final translation of the cultural data strip is modified according to the feedback information to form a new final translation, and the new final translation will be used as the review content of the target translator's review this time; the modification record of the cultural data strip is increased once, thereby obtaining the modification records of each cultural data strip, and integrating them into the translators involved in the review, thereby obtaining several modification records corresponding to each translator.

Images