Automatic detection method for installation and setting of multilingual input method for translation man-machine conversation examination

Abstract

The invention relates to the technical field of automatic detection of input methods, in particular to an automatic detection method for installation setting of a multilingual input method for a translation man-machine conversation test, which comprises the following steps of: automatically identifying and installing a language input method according to examinee information, and automatically detecting the installation setting of the multilingual input method according to the installed language input method and historical test information of a basic test process; calculating a weight coefficient of a test item in the basic test process; the method comprises the following steps: analyzing historical abnormal information of a language input method, selecting a feature test item, automatically detecting the installed language input method based on a basic test process and the feature test item of the language input method, and calculating a test overall score of the language input method by combining a score after test with a corresponding weight coefficient. According to the method, the test can be more targeted, and meanwhile, the language input method needing to be installed is automatically identified and tested, so that the test process of the language input method is unified, and the influence of the language input method on examinees is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of automatic detection of input methods, and particularly to an automatic detection method for installation and setting of multi-language input methods in translation human-computer dialogue examinations. Background Art

[0002] The translation professional qualification (level) examination adopts a computer-based examination form. This examination involves many minority languages and requires supporting Chinese and foreign language input methods.

[0003] The prior art CN105653056A discloses a test method and device for an input method, including: calculating the coordinates of each key in the input method interface according to a preset algorithm, and writing the coordinates of each key into a configuration file; starting a test program, loading and parsing the configuration file to obtain the coordinates of at least one key corresponding to a test string; according to the coordinates of at least one key corresponding to the test string, simulating a click operation at the corresponding position in the input method interface; obtaining candidate word content and checking whether the candidate word content is correct.

[0004] However, during the translation professional qualification examination, if these input methods are manually installed and tested, not only is the workload huge, but also because of the strong professionalism, general testers do not have relevant test knowledge, and manual testing will result in inconsistent test standards for input method testing, thereby affecting the examination status of candidates. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems in the background art, and propose an automatic detection method for installation and setting of multi-language input methods in translation human-computer dialogue examinations.

[0006] In order to achieve the above purpose, the present invention adopts the following technical solutions: An automatic detection method for installation and setting of multi-language input methods in translation human-computer dialogue examinations, the method specifically includes the following steps: Step 1: Based on the computer-based examination database, obtain candidate information, match the candidate information in the computer-based examination database to obtain the translation examination language of the corresponding candidate, and at the same time, based on the translation examination language, match the corresponding language input method in the input database and perform automatic installation; Step 2: According to the installed language input method, obtain the basic test process of the corresponding language input method and the corresponding standard response value, and calculate the weight coefficient of the test items in the basic test process based on the historical test information of the basic test process; Step 3: Obtain the abnormal information of the installed language input method history, classify the abnormal information according to the abnormal causes of the abnormal information, obtain the data set of each fault type in the abnormal information, process the number of fault occurrences and the fault delay time for each time period in the fault occurrence data based on the fault occurrence data in the data set, obtain the fault fluctuation value and the average delay time, and then calculate the comprehensive influence value of the fault type from the fault fluctuation value and the average delay time. Based on the comprehensive influence value, determine the characteristic fault of the corresponding language input method; Step 4: Based on the characteristic fault, determine the characteristic test items, perform installation tests on the installed language data method according to the characteristic test items and the basic test process, and obtain the test scores of the corresponding test items based on the test process. First, obtain the detection qualified signal and the detection abnormal signal according to the test scores, and then based on the detection qualified signal, calculate the overall test score of the corresponding language input method by combining the test scores of the test items in the basic test process with the corresponding weight coefficients and the test scores of the characteristic test items. Finally, obtain the detection completion signal based on the overall test score.

[0007] As a further solution of the present invention, the method for automatically installing a language input method includes: Set up an exam database, where the exam database sets up a candidate database and an input database. The candidate database is used to identify the candidate information in the current exam. The candidate information includes the admission ticket number and the translation exam language, and each admission ticket number corresponds to a translation exam language. The input database is used to store the input method of the translation exam and the corresponding additional plug-ins; Set up an information input unit and an information matching unit. Before the translation exam starts, the candidate inputs the admission ticket number into the information input unit. The information input unit transmits the admission ticket number to the information matching unit, and the information matching unit retrieves the admission ticket number in the candidate database to obtain the matching translation exam language; Based on the matched translation exam language, perform language matching on the translation exam language in the input database to obtain the matched language input method and plug-in, then automatically generate an installation instruction signal, and automatically install the matched language input method based on the installation instruction signal.

[0008] As a further solution of the present invention, the calculation method of the weight coefficient of the test items in the basic test process includes: S1: Based on the exam requirement information of the language input method, set up multiple basic test processes, where the basic test processes include input function test, performance test, shortcut key test, and word library test; Set up a standard test library, where the standard test library is used to test the test items in the basic test process, and based on the test results, obtain the standard response value of each test item; S2: Obtain the historical test information of the language input method, divide the test items in the historical test information according to the test items in the basic test process, then randomly select a test item in the basic test process and mark it as the target item. Taking the target item as an example, obtain the test data corresponding to the target item in the historical test information and mark the test data as CSi, where i = 1, 2, ……, I, indicating that there are I test data in total for the target item; Use the formula to convert the test data of the target item into standardized data , where j represents the test item, and j = 1, 2, ……, J, indicating that there are j test items in total in the basic test process, and XBj represents the standard response value of test item j; S3: Based on the formula calculate the information entropy Hj of the target item j, where, , represents the probability distribution value of test data i in test item j. Further, ; S4: Obtain the difference coefficient dj of test item j based on the formula dj = 1 - Hj, and then use the formula to obtain the weight coefficient of test item j.

[0009] As a further solution of the present invention, the standard response values of the test items include: The basic test process includes input function test, performance test, shortcut key test, and thesaurus test. The standard response value of the function test refers to the accuracy rate of the input method response. The standard response value of the performance test refers to the real-time response time of the input method. The standard response value of the shortcut key test refers to the accuracy rate of the shortcut key conversion. The standard response value of the thesaurus test refers to the integrity of the thesaurus.

[0010] As a further solution of the present invention, the method for determining the fault fluctuation value includes: SS1: Based on the installed language input method, obtain the abnormal information that occurs during the translation exam of this language input method, and at the same time obtain the corresponding reasons for the abnormal information, and classify the abnormal information according to the reasons for the abnormal information to obtain a data set of abnormal information, where each abnormal information corresponds to a data set; Randomly select a fault type in the abnormal information and mark it as the target fault. Taking the target fault as an example, obtain the data set of the target fault, where the data set contains the historical fault occurrence data corresponding to the fault type; Divide the data in the data set according to the exam time, count the number of fault occurrences of the target fault within the same exam time, and mark it as Fm, where m = 1, 2, ……, M, indicating that there are M exams in total; SS2: Obtain the total number of installations YSm of the corresponding language input method in each time period, and then use the formula Fm÷YSm = GLm to obtain the failure occurrence probability GLm of the target failure; Perform mean processing on the failure occurrence probabilities in different time periods to obtain the failure occurrence mean GLp, and then based on the formula Determine the failure fluctuation value Bn of the target failure, where n represents the target failure.

[0011] As a further solution of the present invention, the method for determining the comprehensive influence value includes: Based on the occurrence data of the target failure, obtain the failure delay time of each failure. At the same time, comprehensively process and perform mean processing on the failure delay times of all failures in the data set to obtain the average delay time tp of the target failure; Use the formula Calculate the comprehensive influence value YZn of the target failure, where and are both proportionality factors, and a1 and a2 are the base coefficients respectively.

[0012] As a further solution of the present invention, the method for determining the characteristic failure includes: Obtain the influence threshold Yy, compare the comprehensive influence value YZn of each failure type with the influence threshold Yy respectively, and select the failure types with the comprehensive influence value YZn greater than or equal to the influence threshold Yy and mark them as characteristic failures.

[0013] As a further solution of the present invention, the method for determining the detection qualified signal includes: Take the characteristic failure as the characteristic test item of the corresponding language input method in this translation test. Based on the basic test process and the characteristic test item, perform installation tests on the language input method after automatic installation. Further, a test score is set in each test program, and the test score is used to evaluate the score of each test item. Among them, the higher the score of the test item, the higher the running stability of the corresponding test item. On the contrary, the lower the score of the test item, the lower the running stability of the corresponding test item; After the installation test is completed, obtain the test scores of all test items. At the same time, mark the test score of the basic test process as PCj, and mark the test score of the characteristic test item as PTe, where e = 1, 2,..., E, indicating that there are E characteristic test items in total for the corresponding language input method in this exam; Compare the test scores of all test items with the test passing scores of the corresponding test items respectively. If the test scores of all test items are greater than or equal to the test passing scores of the corresponding test items, then generate a detection qualified signal.

[0014] As a further solution of the present invention, when the test score of a test item is less than the passing score of the test, a detection abnormal signal is generated. At the same time, the corresponding test item and the detection abnormal signal are transmitted to the display terminal, and an audible and visual reminder signal is generated to remind the corresponding responsible person. The responsible person performs fault maintenance on the corresponding test item. After the fault maintenance is completed, the corresponding test item is reinstalled and tested until a detection qualified signal is generated.

[0015] As a further solution of the present invention, the method for generating a detection completion signal includes: When a detection qualified signal is generated, using the formula calculate to obtain the overall test score Zf, represents the weight coefficient of test item j in the basic test process; Compare the overall test score Zf with the overall test threshold. If the overall test score Zf is greater than or equal to the overall test threshold, a detection completion signal is generated and transmitted to the terminal device. Otherwise, if the overall test score Zf is less than the overall test threshold, an input lag signal is generated and transmitted to the display terminal. Then, when the corresponding responsible person obtains the input lag signal, the corresponding language input method is adjusted.

[0016] Compared with the existing technology, the advantages of the present invention are: The present invention automatically identifies the language input method to be installed according to the candidate information, and analyzes the weights of the test items in the basic test processes of different input methods based on the basic test processes of the installed language input methods, so that when different input methods perform the basic test processes, they can accurately reflect the problems that occur in the basic test processes of the input methods. At the same time, analyze the historical abnormal information of the installed language input method, select the corresponding characteristic faults, and determine the characteristic test items based on the characteristic faults, making the test more targeted and able to focus on the key parts that are prone to problems in the actual use of the input method, improving the test efficiency and accuracy; The present invention automatically identifies the language input method to be installed according to the candidate information, and automatically installs, tests and scores the installed language input method. For the same language input method, the test methods are the same and the scoring criteria are the same, making the test process of the language input method unified, thereby reducing the impact of the language input method on the candidate. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic structural diagram of the method flow of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0019] Refer to Figure 1 , a method for automatically detecting the installation and setting of a multi - language input method for a translation human - machine dialogue examination. The method specifically includes the following steps: Step 1: Set up an examination database for machine - based examinations. Among them, in the examination database for machine - based examinations, a candidate database and an input database are set up. Further, the candidate database is used to identify the candidate information in the current examination. The candidate information includes the admission ticket number and the translation examination language, and each admission ticket number corresponds to a translation examination language. The input database is used to store the input methods for the translation examination and the corresponding additional plugins. Set up an information input unit and an information matching unit. Before the translation examination starts, the candidate inputs the admission ticket number into the information input unit. The information input unit transmits the admission ticket number to the information matching unit, and the information matching unit retrieves the admission ticket number in the candidate database to obtain the matching translation examination language. Based on the matched translation examination language, perform language matching in the input database to obtain the matching language input method and plugin, and then automatically generate an installation instruction signal, and automatically install the matched language input method based on the installation instruction signal. Step 2: After the automatic installation of the language input method is completed, collect the examination requirement information for this language based on the translation examination language corresponding to the language input method. Then, based on the examination requirement information, set weight factors for the basic test process. The specific setting method includes: S1: Based on the examination requirement information of the language input method, set multiple basic test processes. Among them, the basic test processes include input function test, performance test, shortcut key test, and thesaurus test. Then set up a standard test library. Among them, the standard test library is used to test the test items in the basic test process, and based on the test results, obtain the standard response value of each test item. Among them, the standard response value of the function test refers to the accuracy rate of the input method response, the standard response value of the performance test refers to the real - time response time of the input method, the standard response value of the shortcut key test refers to the accuracy rate of the shortcut key conversion, and the standard response value of the thesaurus test refers to the integrity of the thesaurus. S2: Obtain the historical test information of the language input method, divide the test items in the historical test information according to the test items in the basic test process, then arbitrarily select a test item in the basic test process and mark it as the target item. Taking the target item as an example, obtain the test data corresponding to the target item in the historical test information and mark the test data as CSi, where i = 1, 2, ……, I, indicating that there are I test data in total for the target item; Use the formula to convert the test data of the target item into standardized data , where j represents the test item, and j = 1, 2, ……, J, indicating that there are j test items in total in the basic test process, and XBj represents the standard response value of test item j; S3: Based on the formula calculate the information entropy Hj of the target item j, where, , represents the probability distribution value of test data i in test item j. Further, ; After that, sequentially use the remaining test items in the basic test process as the target item and process them according to the methods in steps S2 and S3 above to obtain the information entropy Hj of each test item j; S4: Obtain the difference coefficient dj of test item j based on the formula dj = 1 - Hj, and then use the formula to obtain the weight coefficient of test item j; Step Three: Based on the installed language input method, obtain the abnormal information that occurs during the translation exam of this language input method, and at the same time obtain the corresponding reasons for the abnormal information, and classify the abnormal information according to the reasons for the abnormal information to obtain the data set of abnormal information. Among them, each abnormal information corresponds to a data set. After that, analyze the data set of abnormal information to determine the characteristic faults. Further, the specific determination method of the characteristic faults includes: SS1: Arbitrarily select a fault type in the abnormal information and mark it as the target fault. Taking the target fault as an example, obtain the data set of the target fault, where the data set contains the historical fault occurrence data corresponding to the fault type; Divide the data in the data set according to the exam time, count the number of occurrences of the target fault within the same exam time and mark it as Fm, m = 1, 2, ……, M, indicating that there are M exams in total; SS2: Obtain the total number of installations YSm of the corresponding language input method in each time period, and then use the formula Fm ÷ YSm = GLm to obtain the fault occurrence probability GLm of the target fault; The failure occurrence probabilities in different time periods are averaged to obtain the failure occurrence mean value GLp, and then based on the formula the failure fluctuation value Bn of the target failure is determined, where n represents the target failure; Based on the occurrence data of the target failure, the failure delay time of each failure is obtained. At the same time, the failure delay times of all failures in the data set are integrated and averaged to obtain the average delay time tp of the target failure; SS3: Use the formula to calculate the comprehensive influence value YZn of the target failure, where and are both proportionality factors, a1 and a2 are the base coefficients respectively, and the specific 、 、a1 and a2 values are obtained by those skilled in the art through big data operations; It should be further noted that when the comprehensive influence value of the target failure is larger, it indicates that the target failure has a greater impact on the examinee. On the contrary, when the comprehensive influence value of the target failure is smaller, it indicates that the target failure has a smaller impact on the examinee; After that, the remaining failure types in the abnormal information are successively used as the target failures, and processed according to the methods in the above steps SS1 to SS3 to obtain the comprehensive influence values YZn of each failure type; SS4: Obtain the influence threshold Yy, compare the comprehensive influence value YZn of each failure type with the influence threshold Yy respectively, and select the failure types with the comprehensive influence value YZn greater than or equal to the influence threshold Yy and mark them as characteristic failures, where the specific value of the influence threshold Yy is obtained by those skilled in the art through big data operations; Step Four: Use the characteristic failures as the characteristic test items for the corresponding language input method in this translation exam. Based on the basic test process and the characteristic test items, install and test the automatically installed language input method. Further, a test score is set in each test program, and the test score is used to evaluate the score of each test item. Among them, the higher the score of the test item, the higher the running stability of the corresponding test item. On the contrary, the lower the score of the test item, the lower the running stability of the corresponding test item. Specifically, setting the test score in the test program is set by those skilled in the art according to big data experience; After the installation test is completed, obtain the test scores of all test items. At the same time, mark the test score of the basic test process as PCj, and mark the test score of the characteristic test item as PTe, where e = 1, 2, ……, E, indicating that there are E characteristic test items for the corresponding language input method in this exam; First, compare the test scores of all test items with the passing scores of the corresponding test items respectively. If the test scores of all test items are greater than or equal to the passing scores of the corresponding test items, a detection qualified signal is generated. On the contrary, if there is a test item whose test score is less than the passing score of the test item, a detection abnormal signal is generated. At the same time, the corresponding test item and the detection abnormal signal are transmitted to the display terminal, and an audible and visual reminder signal is generated to remind the corresponding responsible person. The responsible person performs fault maintenance on the corresponding test item. After the fault maintenance is completed, the corresponding test item is reinstalled and tested until a detection qualified signal is generated during the installation test; When a detection qualified signal is generated, use the formula to calculate the overall test score Zf. Then compare the overall test score Zf with the overall test threshold. If the overall test score Zf is greater than or equal to the overall test threshold, a detection completion signal is generated and transmitted to the terminal device. On the contrary, if the overall test score Zf is less than the overall test threshold, an input lag signal is generated and transmitted to the display terminal. Then, when the corresponding responsible person obtains the input lag signal, the corresponding language input method is adjusted to reduce the impact of the language input method on the examinee; It should be further noted that the specific values of the passing scores of the test items and the overall test threshold are set by those skilled in the art according to big data experience respectively.

[0020] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution of the present invention and its inventive concept, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. Method for automatic detection of installation and setting of multi - language input method in human - machine dialogue examination, characterized in that, The method specifically includes the following steps: Step 1: Based on the computer-based examination database, obtain candidate information, match the candidate information in the computer-based examination database to obtain the translation examination language of the corresponding candidate. At the same time, based on the translation examination language, match the corresponding language input method in the input database and install it automatically; Step 2: According to the installed language input method, obtain the basic test process of the corresponding language input method and the corresponding standard response value. Based on the historical test information of the basic test process, calculate the weight coefficient of the test items in the basic test process; Step 3: Then obtain the historical abnormal information of the installed language input method, classify the abnormal information according to the cause of the abnormality, obtain the data set of each fault type in the abnormal information, process the number of fault occurrences and the fault delay time of each time period in the fault occurrence data in the data set respectively to obtain the fault fluctuation value and the average delay time. Then, based on the fault fluctuation value and the average delay time, calculate the comprehensive influence value of the fault type, and determine the characteristic fault of the corresponding language input method based on the comprehensive influence value; Step 4: Based on the characteristic fault, determine the characteristic test items, perform installation tests on the installed language data method according to the characteristic test items and the basic test process, and obtain the test scores of the corresponding test items based on the test process. First, obtain the detection qualified signal and the detection abnormal signal according to the test scores, and then based on the detection qualified signal, combine the test scores of the test items in the basic test process with the corresponding weight coefficients and the test scores of the characteristic test items to calculate the overall test score of the corresponding language input method. Finally, obtain the detection completion signal based on the overall test score; 2. The automatic detection method for installation and setting of a multi-language input method in a translation human-machine dialogue examination, characterized in that, The method for automatically installing a language input method includes: Set up a computer-based examination database, where the candidate database and the input database are set in the computer-based examination database. The candidate database is used to identify the candidate information in the current examination. The candidate information includes the admission ticket number and the translation examination language, and each admission ticket number corresponds to a translation examination language. The input database is used to store the input methods and corresponding additional plugins for the translation examination; Set up an information input unit and an information matching unit. Before the translation examination starts, the candidate inputs the admission ticket number into the information input unit. The information input unit transmits the admission ticket number to the information matching unit, and the information matching unit retrieves the admission ticket number in the candidate database to obtain the matching translation examination language; Based on the matched translation examination language, perform language matching on the translation examination language in the input database to obtain the matched language input method and plugin, then automatically generate an installation instruction signal, and automatically install the matched language input method based on the installation instruction signal; 3. The automatic detection method for installation and setting of a multi-language input method for translation human-machine dialogue examination according to claim 1, characterized in that, The calculation method of the weight coefficient of the test items in the basic test process includes: S1: Based on the examination requirement information of the language input method, set multiple basic test processes, where the basic test processes include input function test, performance test, shortcut key test, and thesaurus test; Set up a standard test library, where the standard test library is used to test the test items in the basic test process and obtain the standard response values of each test item based on the test results; S2: Obtain the historical test information of the language input method, divide the test items in the historical test information according to the test items in the basic test process, then randomly select a test item in the basic test process and mark it as the target item. Taking the target item as an example, obtain the test data corresponding to the target item in the historical test information and mark the test data as CSi, where i = 1, 2,..., I, indicating that there are I test data in total for the target item; Using the formula Convert the test data of the target project into standardized data , where j represents the test item, and j = 1, 2,..., J, indicating that there are j test items in the basic test process, and XBj represents the standard response value of test item j; S3: Based on the formula calculate the information entropy Hj of the target item j, where , represents the probability distribution value of the test data i in the test item j. Further, ; S4: Obtain the coefficient of variation dj of test item j based on the formula dj = 1 - Hj, and then use the formula to obtain the weight coefficient of test item j.

4. The automatic detection method for installation and setting of a multi-language input method in a translation human-machine dialogue examination, characterized in that, The standard response values of the test items include: The basic test process includes input function test, performance test, shortcut key test, and thesaurus test. The standard response value of the function test refers to the accuracy of the input method response. The standard response value of the performance test refers to the real-time response time of the input method. The standard response value of the shortcut key test refers to the accuracy of the shortcut key conversion. The standard response value of the thesaurus test refers to the integrity of the thesaurus.

5. The automatic detection method for installation and setting of a multi-language input method for a translation human-machine dialogue examination according to claim 1, characterized in that, The method for determining the fault fluctuation value includes: SS1: Based on the installed language input method, obtain the abnormal information that occurs during the translation exam of this language input method, and at the same time obtain the corresponding reasons for the abnormal information. Classify the abnormal information according to the reasons for the abnormal information to obtain a data set of abnormal information, where each abnormal information corresponds to a data set; Randomly select a fault type in the abnormal information and mark it as the target fault. Taking the target fault as an example, obtain the data set of the target fault, where the data set contains the historical fault occurrence data corresponding to the fault type; Divide the data in the data set according to the exam time, count the number of fault occurrences of the target fault within the same exam time, and mark it as Fm, where m = 1, 2,..., M, indicating that there are M exams in total; SS2: Obtain the total number of installations YSm of the corresponding language input method in each time period, and then use the formula Fm÷YSm = GLm to obtain the fault occurrence probability GLm of the target fault; The failure occurrence probabilities in different time periods are averaged to obtain the average failure occurrence GLp, and then based on the formula the failure fluctuation value Bn of the target failure is determined, where n represents the target failure.

6. The installation and setting automatic detection method of the translation human-machine dialogue examination multi-language input method according to claim 5, characterized in that The method for determining the comprehensive influence value includes: Based on the occurrence data of the target fault, obtain the fault delay time of each fault, and at the same time comprehensively process the fault delay times of all faults in the data set and perform mean processing to obtain the average delay time tp of the target fault; Use the formula to calculate the comprehensive influence value YZn of the target fault, where and are both proportionality factors, and a1 and a2 are the base coefficients respectively.

7. The automatic detection method for installation and setting of a multi-language input method for a translation human-computer dialogue examination according to claim 6, characterized in that, The method for determining the characteristic fault includes: Obtain the influence threshold Yy, compare the comprehensive influence values YZn of each fault type with the influence threshold Yy respectively, and select the fault types with the comprehensive influence value YZn greater than or equal to the influence threshold Yy and mark them as characteristic faults.

8. The automatic detection method for installation and setting of a multi-language input method in a translation human-computer dialogue examination, characterized in that The method for determining the detection qualified signal includes: Take the feature failure as the feature test item of the corresponding language input method in this translation exam. Based on the basic test process and the feature test item, conduct the installation test on the language input method after automatic installation. Further, a test score is set in each test program, and the test score is used to evaluate the score of each test item. Among them, the higher the score of the test item, the higher the running stability of the corresponding test item; on the contrary, the lower the score of the test item, the lower the running stability of the corresponding test item. After the installation test is completed, obtain the test scores of all test items. At the same time, mark the test score of the basic test process as PCj, and mark the test score of the feature test item as PTe, where e = 1, 2,..., E, indicating that there are E feature test items for the corresponding language input method in this exam. Compare the test scores of all test items with the test passing scores of the corresponding test items respectively. If the test scores of all test items are greater than or equal to the test passing scores of the corresponding test items, then generate a detection qualified signal.

9. The automatic detection method for installation and setting of a multi-language input method for a translation human-computer dialogue examination according to claim 8, characterized in that, When the test score of a test item is less than the test passing score, generate a detection abnormal signal, and at the same time transmit the corresponding test item and the detection abnormal signal to the display terminal, and generate an audible and visual reminder signal to remind the corresponding responsible person. The responsible person conducts fault maintenance on the corresponding test item. After the fault maintenance is completed, reinstall the test on the corresponding test item until the installation test generates a detection qualified signal.

10. The automatic detection method for installation and setting of a multi-language input method in a translated human-machine dialogue examination, characterized in that The method for generating the detection completion signal includes: When a qualified detection signal is generated, use the formula to calculate the overall test score Zf, indicating the weight coefficient of test item j in the basic test process; Compare the overall test score Zf with the overall test threshold. If the overall test score Zf is greater than or equal to the overall test threshold, then generate a detection completion signal and transmit it to the terminal device. On the contrary, if the overall test score Zf is less than the overall test threshold, then generate an input lag signal and transmit it to the display terminal. After that, when the corresponding responsible person obtains the input lag signal, adjust the corresponding language input method.

Images