Continuously improved corpus question classification method and system

Abstract

The invention provides a continuously improved corpus problem classification method, and belongs to the technical field of information processing and corpus construction. The method is realized based on a cluster and comprises the following steps: judging whether the electronic equipment generates parallel corpora or not, verifying the electronic equipment by adopting a verification equipment set, performing pre-submission operation on each parallel corpus stored in a pre-submission corpus, and updating a standard corpus according to a feedback result of the pre-submission operation. The updating comprises the steps of carrying out reclassification on parallel corpora of the standard corpus according to source corpora or target corpora and the like. The cluster is a translation server cluster or a question and answer server cluster. According to the technical scheme, the constructed corpus information can be continuously improved and updated, and corpus input errors caused by subjective deviation accumulation of individual users or devices can be effectively avoided in the updating process; in addition, the updating process is based on a distributed transaction processing mechanism, and the consistency of standards is ensured.

Description

Technical Field

[0001] The present invention belongs to the field of information processing and corpus technology, and in particular relates to a sustainably improved corpus question classification method and system, a computer-readable storage medium for implementing the method, a computer program product and an electronic device. Background Art

[0002] A corpus refers to a large-scale electronic database that has been scientifically sampled and processed, and contains language materials that have actually appeared in the actual use of the language. The original corpus only provided a large amount of natural and real language facts and text data for language research and teaching. For example, the corpus of a parallel corpus can form a translation relationship and is mostly used in application fields such as machine translation and bilingual dictionary compilation. The data-driven learning method based on the corpus conforms to the learning theory of schema structure and constructivism, and can help learners improve their language awareness, internalize the rules of language use, and enhance their language communication skills. The corpus has gradually become one of the important auxiliary tools for independent foreign language learning.

[0003] With the development of artificial intelligence technology, the application of corpus in other industries has gradually shown its advantages. Researchers can extract and analyze the language characteristics of various genres based on detailed corpus text evidence, and describe the vocabulary characteristics and text characteristics of specific genres and registers more scientifically and objectively. Combined with deep learning engines, parallel corpora including questions and answers can be constructed to achieve anthropomorphic computer-assisted human-computer interaction, the most common of which are intelligent customer service systems, intelligent robot dialogue systems, etc.

[0004] Regardless of the application scenario, building a suitable corpus and accurately classifying the questions are prerequisites, among which the accuracy of the corpus, the accuracy of annotation and the consistency are important judgment criteria. Chinese invention patent application CN201710020573.4 proposes a method for constructing an event-based Chinese reference corpus, which claims to be based on Chinese syntactic analysis and semantic analysis, in line with the characteristics of Chinese. The method can also perform consistency checks on the annotated corpus to ensure the quality of the corpus annotation.

[0005] However, natural language changes and evolves rapidly, and the corresponding corpus should also be updated in a timely manner. Based on the existing standard corpus, the corpus can be continuously improved through manual assisted updates, automatic machine updates and other methods. In this process, the first thing that needs to be ensured is the consistency of the corpus update standards; secondly, different users and different machines may have different standards for whether the same corpus is suitable for update. Under the clustering effect, how to effectively avoid corpus entry errors caused by the accumulation of subjective biases of individual users or devices is also a technical problem that needs to be solved. Summary of the invention

[0006] In view of the above technical problems, the present invention provides a method and system for classifying corpus problems that can be continuously improved, a computer-readable storage medium for implementing the method, a computer program product, and an electronic device.

[0007] In the first aspect of the present invention, a method for classifying corpus problems that can be continuously improved is provided. The method is implemented based on a cluster, and the cluster includes N electronic devices.

[0008] The method can be automatically executed based on a computer program. The method includes the following computer-executable process steps:

[0009] S110: Let i = 1;

[0010] S120: Determine whether the electronic device generates parallel corpus; the parallel corpus includes source corpus and target corpus;

[0011] If yes, save the parallel corpus to the pre-submission corpus and proceed to step S130;

[0012] If no, add the electronic device as a verification electronic device to the verification device set and proceed to step S130;

[0013] S130: Determine whether i < N holds. If yes, let i = i + 1 and return to step S120;

[0014] If no, proceed to step S140;

[0015] S140: Use the verification electronic devices in the verification device set to perform pre-submission operations on each parallel corpus saved in the pre-submission corpus;

[0016] S150: Update the standard corpus according to the feedback results of the pre-submission operations. The update includes reclassifying the parallel corpus in the standard corpus according to the source corpus or the target corpus.

[0017] In a specific application, the cluster is a translation server cluster, and the electronic device is configured with at least one translation engine; the parallel corpus is translation corpus, the source corpus is the corpus to be translated, the target corpus is the translated corpus, and the corpus to be translated and the translated corpus are in different languages.

[0018] At this time, step S140 specifically includes:

[0019] Use the corpus to be translated corresponding to each translation corpus as the input of the translation engine of each verification electronic device, and determine the feedback result of the pre-submission operation based on the comparison result between the output result of the translation engine and the translated corpus corresponding to each translation corpus.

[0020] Different verification electronic devices call different translation engines each time.

[0021] In another specific application, the cluster is a question-answering server cluster; the parallel corpus is a question-answering corpus, the source corpus is a question corpus, the target corpus is an answer corpus, and the question corpus and the answer corpus are in the same language.

[0022] At this time, each of the electronic devices is configured with an intelligent question and answer engine; different verification electronic devices call different intelligent question and answer engines each time.

[0023] The step S140 specifically includes:

[0024] The question corpus corresponding to each question and answer corpus is used as the input of the intelligent question and answer engine of each verification electronic device, and the feedback result of the pre-submission operation is determined based on the comparison result between the output result of the intelligent question and answer engine and the answer corpus corresponding to each question and answer corpus.

[0025] In a second aspect of the present invention, in order to implement the sustainable improvement corpus question classification method described in the first aspect, a sustainable improvement corpus question classification system is proposed, the system includes a cluster server and a standard corpus, the cluster server communicates wirelessly with N electronic devices, and the system also includes:

[0026] A corpus detection unit, used to detect whether the electronic device generates parallel corpus;

[0027] A pre-submission unit, used for saving the parallel corpus generated by the electronic device to a pre-submission corpus;

[0028] A verification device designation unit, used to designate the electronic device as a verification electronic device when the electronic device does not generate parallel corpus;

[0029] A pre-operation execution unit, configured to use a verification electronic device to perform a pre-submission operation on each parallel corpus stored in the pre-submission corpus;

[0030] An updating unit, configured to update the standard corpus according to the feedback result of the pre-submission operation, wherein the updating includes reclassifying the parallel corpus of the standard corpus according to the source corpus or the target corpus;

[0031] Among them, the cluster server is a translation server and / or a question and answer server; the electronic device is configured with a translation engine or an intelligent question and answer engine; and N is an integer greater than 2.

[0032] The verification device designation unit configures all verification electronic devices to form a distributed transaction execution unit; the pre-commit operation performed by the distributed transaction execution unit on each parallel corpus stored in the pre-commit corpus constitutes a distributed atomic transaction operation.

[0033] The above-mentioned sustainably improved corpus question classification method can be automatically implemented through various forms of electronic devices through computer program instructions; the computer program instructions can be stored in different forms of storage media and loaded into computer electronic devices for execution.

[0034] Therefore, in the third aspect of the present invention, a computer-readable storage medium is also provided for storing computer instructions. When the computer instructions are executed on an electronic device, the electronic device executes all or part of the steps of the above-mentioned sustainably improved corpus question classification method.

[0035] In a fourth aspect of the present invention, a computer device is also proposed, comprising a processor and a memory, wherein the memory is used to store instructions, and the processor is used to call the instructions in the memory, so that the computer device executes the above-mentioned sustainably improved corpus question classification method.

[0036] In a fifth aspect of the present invention, a computer program product is also proposed, the product comprising a computer program, and when the computer program is executed, all or part of the steps of the above-mentioned sustainably improved corpus question classification method are implemented.

[0037] The technical solution of the present invention enables the constructed corpus information to be continuously improved and updated, and the update process can effectively avoid corpus entry errors caused by the accumulation of subjective deviations of individual users or devices; in addition, the update process is based on a distributed transaction processing mechanism to ensure the consistency of standards.

[0038] Further advantages of the present invention will be further reflected in detail in the specific embodiments section in conjunction with the drawings of the specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 This is a schematic diagram of the main process of a sustainable improved corpus question classification method according to an embodiment of the present invention.

[0041] Figure 2 Yes Figure 1 Schematic diagram of a preferred embodiment where the method is specifically applied to the update of a translation corpus

[0042] Figure 3 Yes Figure 1 Schematic diagram of a preferred embodiment where the method is specifically applied to the update of a question - answering corpus

[0043] Figure 4 Yes Figure 1 Schematic diagram of the hardware / software architecture of a specific implementation scenario of the method

[0044] Figure 5 Schematic diagram showing the composition of the functional module units of a sustainable improvement corpus question classification system according to an embodiment of the present invention Detailed implementation manner

[0045] First, it should be pointed out that the embodiments of the sustainable improvement corpus question classification method mentioned in this part can be implemented through a computer program on an electronic device or system configured with a memory and a processor. The forms of the electronic device and system can be a physical machine, a virtual machine, a server, a cluster, or any combination thereof.

[0046] Preferably, the specific form of the electronic device can also be a human - machine interaction terminal, and the human - machine interaction terminal can be a desktop terminal with a human - machine interaction interface, a smart handheld terminal, a mobile terminal, etc.

[0047] First, refer to Figure 1 , Figure 1 which shows the main process schematic diagram of a sustainable improvement corpus question classification method according to an embodiment of the present invention.

[0048] Figure 1 The method is implemented based on a cluster. The cluster includes N electronic devices, N > 1. The computer implementation process of the method is as follows (for ease of understanding, step numbers are omitted in all the figures):

[0049] S110: Let i = 1;

[0050] S120: Determine whether the electronic device generates parallel corpus; the parallel corpus includes source corpus and target corpus;

[0051] If yes, save the parallel corpus to the pre - submission corpus and go to step S130;

[0052] If no, add the electronic device as a verification electronic device to the verification device set and go to step S130;

[0053] S130: Determine whether i < N holds. If yes, let i = i + 1 and return to step S120;

[0054] If not, proceed to step S140;

[0055] S140: using a verification electronic device of the verification device set to perform a pre-submission operation on each parallel corpus stored in the pre-submission corpus;

[0056] S150: updating the standard corpus according to the feedback result of the pre-submission operation, wherein the updating includes reclassifying the parallel corpus of the standard corpus according to the source corpus or the target corpus.

[0057] Figure 1 The method generally summarizes the main process of the sustainably improved corpus question classification method described in the present invention. The process can be applied to all scenario applications that require corpus support, including human-computer interactive translation, human-computer interactive dialogue, intelligent robot customer service, etc. In these scenarios, a standard corpus is constructed in advance. Based on the standard corpus, the corresponding human-computer interaction / dialogue process is executed through the corresponding artificial intelligence engine (translation engine, intelligent customer service engine, human-computer dialogue engine, etc.).

[0058] Due to the continuous updating of application scenarios and natural languages, especially in recent years under the wave of big data and Internet+, the corpus is changing with each passing day. The original standard corpus also needs to be constantly updated and supplemented, otherwise it will not be able to keep up with the latest interaction trends and the user experience will be reduced.

[0059] This process of continuous supplementation and updating is the process of continuous improvement of corpus problems and their classification. This process is usually maintained by a dedicated person or automatically collected, maintained and updated by a machine. However, the number of single dedicated maintenance personnel is limited after all. Therefore, many application scenarios have developed user active participation or task crowdsourcing / subcontracting models; the former encourages a large number of users in the corresponding scenarios to actively report new words, new explanations, new translations, new answers / new questions, and then choose to enter the corpus; or it also adopts a crowdsourcing incentive model, subcontracting the corpus update task to crowdsourcing users, and giving certain rewards to users who participate in actively reporting corpus updates.

[0060] Regardless of which mode is used, since multiple users / user terminal devices are involved, there may be multiple different results to be submitted for the same corpus, which will lead to the problem of consistency in the standard of corpus update. Because different users and different machines may have different judgment criteria when facing the same corpus to be updated, how to effectively avoid corpus entry errors caused by the accumulation of subjective deviations of individual users or devices under the cluster effect is a technical problem that needs to be solved.

[0061] Figure 1The embodiment of is proposed to solve the above problem. Among them, N electronic devices {Dev1, Dev2, ..., Dev N} corresponds to the multiple users / user terminal devices mentioned above, that is, crowdsourcing users / devices.

[0062] Take a translation APP as an example. This translation APP implements computer-assisted translation based on the support of a cluster translation server. The cluster translation server is connected to a standard translation corpus. The standard translation corpus stores parallel corpora, which include source corpora and target corpora. The source corpora and target corpora are in a corresponding relationship. The corresponding relationship includes one-to-one, one-to-many, many-to-many or many-to-one. That is, one source corpus may correspond to one or more target corpora, and multiple source corpora may also correspond to one or more target corpora.

[0063] In this scenario, the parallel corpus is specifically a translation corpus, the source corpus is the corpus to be translated, the target corpus is the translated corpus, and the corpus to be translated and the translated corpus are in different languages.

[0064] When the user inputs the word to be translated W1 and the sentence to be translated S1 (the corpus to be translated) in the first language, the translation APP calls the standard translation corpus through the cluster translator server to execute the translation engine and outputs the corresponding translation results as the translated word W2 and the translated sentence S2 (the target corpus) in the second language.

[0065] Take a second APP application that supports intelligent human-machine customer service as an example. The second APP can provide intelligent human-machine customer service. When users have questions related to the use of the APP and need to consult, there is no need for online manual customer service. The second APP can call the question-and-answer server cluster based on the standard question-and-answer corpus to autonomously respond to the questions entered by the user. Each parallel corpus contained in the parallel corpus of the standard question-and-answer corpus is also composed of a source corpus and a target corpus. The parallel corpus is a question-and-answer corpus, the source corpus is a question corpus, and the target corpus is an answer corpus. The question corpus and the answer corpus are in the same language. The source corpus and the target corpus are in a corresponding relationship; the corresponding relationship can also include one-to-one, one-to-many, many-to-many or many-to-one. That is, one source corpus can correspond to one or more target corpora, and multiple source corpora may also correspond to one or more target corpora.

[0066] Whether it is a standard translation corpus or a standard question-and-answer corpus, the stored corpus and questions need to be updated in a timely manner, and the questions (source corpus) or answers (target corpus) also need to be reclassified according to the situation, so as to follow the latest interaction trends and update the user experience.

[0067] Figure 1In general, a corpus problem classification method with sustainable improvement is provided, in which N users using a certain standard corpus correspond to N electronic devices; these N users are encouraged to actively submit new corpora to update the standard corpus. In the prior art, different users submit corpora randomly, the values ​​of the submitted corpora are different, and the judgment criteria for submission are also different.

[0068] To change this situation, Figure 1 The method has been improved in at least three aspects:

[0069] (1) The user (device / terminal) submitting the corpus is separated from the user (device / terminal) verifying the corpus: When a user (device / terminal) submits the corpus, the user (device / terminal) cannot be used as a verification device to achieve cross-verification. Step S120 corresponds to the corresponding technical means.

[0070] (2) Before the corpus is formally submitted, it is necessary to perform a pre-submission operation to ensure standard consistency; step S140 corresponds to the corresponding technical means.

[0071] (3) According to the feedback result of the pre-submission operation, the standard corpus is updated to reflect the distributed atomic transaction operation. Step S150 corresponds to the corresponding technical means.

[0072] As a further preference, in order to improve the execution efficiency of the method, the N electronic devices are currently active users of the application corresponding to the current standard corpus, that is, the N electronic devices are currently using the application (the aforementioned translation APP or the second application APP), thereby ensuring that the data interaction process is active.

[0073] Figure 1 The embodiment of the present invention is to embody the improved concept of the present invention as a whole. Figure 2 and Figure 3 , respectively give Figure 1 Schematic diagram of a preferred embodiment of the method specifically applied to updating a translation corpus and Figure 1 A schematic diagram of a preferred embodiment of the method being specifically applied to updating a question-answer corpus.

[0074] Figure 2 A method for classifying corpus questions that can be continuously improved is proposed. The method is implemented based on a translation server cluster, wherein the cluster includes N electronic devices {Dev1, Dev2, ..., Dev N}, N>1; each of the electronic devices is configured with at least one translation engine;

[0075] The method comprises:

[0076] S210: let i=1;

[0077] S220: Determine whether the electronic device Dev i generates translation corpus L i ; The translation corpus includes source corpus S i and target corpus O i ;

[0078] If yes, save the translation corpus L i to the pre-submission corpus and enter step S230;

[0079] If no, add the electronic device Dev i to the set of verification devices as a verification electronic device and enter step S230;

[0080] S230: Determine whether i < N holds. If yes, set i = i + 1 and return to step S220;

[0081] If no, enter step S240;

[0082] S240: Use the verification electronic devices in the set of verification devices to perform pre-submission operations on each translation corpus saved in the pre-submission corpus;

[0083] S250: Update the translation standard corpus according to the feedback results of the pre-submission operations. The update includes re-classifying and aligning or updating the alignment of the translation corpus in the translation standard corpus according to the source corpus and the target corpus.

[0084] Preferably, in the above Figure 2 embodiment, different verification electronic devices call different translation engines.

[0085] Next, a specific adaptation example is used to describe the implementation process of the above Figure 2 embodiment.

[0086] Assume that in the current translation scenario, there are 5 active user devices A1, A2, A3, A4, A5; among them, user device A2 is generating a translation corpus "magazine - magazine clip" to be submitted to the corpus to obtain a reward, and other users (A1, A3, A4, A5) are using the translation APP but not generating new corpus.

[0087] It can be understood that the conventional translated corpus corresponding to the source corpus "magazine" is usually "magazine" (which has been saved in the translation standard corpus). However, in the mechanical field, especially in the field of special equipment, another translation of the source corpus "magazine" also includes forms such as "clip, magazine, cartridge chamber", etc. Therefore, the translation corpus "magazine - cartridge magazine" generated by the user device A2 can be considered as an update to the standard translation corpus because the existing translation standard corpus did not recommend this result first when the user used it.

[0088] At this time, save the translation corpus "magazine - cartridge magazine" to the pre - submission corpus first;

[0089] After 5 cycles of iteration, assume that the only translation corpus saved in the pre - submission corpus is "magazine - cartridge magazine"; the verification device set is {A1, A3, A4, A5};

[0090] For this translation corpus, use the verification electronic devices in the verification device set to perform pre - submission operations on each translation corpus saved in the pre - submission corpus, specifically including:

[0091] Take the source corpus corresponding to each translation corpus as the input of the translation engine of each verification electronic device, and determine the feedback result of the pre - submission operation based on the comparison result between the output result of the translation engine and the translated corpus corresponding to each translation corpus.

[0092] Specifically, the translation engines used by other user devices (A1, A3, A4, A5) are different, or, corresponding to different translation modes (translation fields). In the case of multiple other users, it is very likely that the translation results output by at least one user device can be the same or similar to the translated corpus corresponding to the translation corpus (clip, magazine, cartridge chamber).

[0093] When the comparison result between the output result of the translation engine of at least one verification electronic device and the translated corpus corresponding to the translation corpus is a match (that is, the similarity is greater than the preset threshold), determine that the feedback result of the pre - submission operation is a positive feedback.

[0094] At this time, according to the positive feedback result of the pre - submission operation, update the translation standard corpus, that is, correspond "magazine" in the standard corpus to "clip, magazine, cartridge chamber", and classify "magazine" as "Mechanical - Special Equipment - Firearms".

[0095] Preferably, for this translation corpus, use the verification electronic devices in the verification device set to perform pre - submission operations on each translation corpus saved in the pre - submission corpus, specifically including:

[0096] Use the translated corpus corresponding to each translation corpus as the input of the translation engine of each verification electronic device, and determine the feedback result of the pre-submission operation based on the comparison result between the output result of the translation engine and the corpus to be translated corresponding to each translation corpus.

[0097] Continuing the above example, the translated corpora (clip, magazine, cartridge chamber) corresponding to the translation corpus "magazine - cartridge magazine" can be used as the input of the translation engines of the translation APPs respectively used by other user devices (A1, A3, A4, A5), and the reverse translation results (magazine) can be obtained;

[0098] At this time, only when the translation results output by the translation engines of the translation APPs respectively used by all other user devices (A1, A3, A4, A5) all contain the corpus to be translated corresponding to the translation corpus, it is determined that the feedback result of the pre-submission operation is a positive feedback; as long as the translation result output by the translation engine of the translation APP used by one user device does not contain the corpus to be translated corresponding to the translation corpus at all, there is no need to submit, that is, the feedback result is a negative feedback, and the operation of updating the standard corpus is not performed.

[0099] At this time, it reflects the distributed atomic transaction operation characteristics of the distributed unit composed of all other user devices (A1, A3, A4, A5), that is, the atomicity of "all execute or all do not execute", to ensure consistency of standards.

[0100] Next, refer to Figure 3 , Figure 3 shows Figure 1 Schematic diagram of a preferred embodiment in which the method is specifically applied to the update of the Q&A corpus.

[0101] Similarly, Figure 3 The execution process of the method is as follows:

[0102] A method for classifying corpus problems with sustainable improvement, the method is implemented based on a Q&A server cluster, the cluster includes N electronic devices {Dev1, Dev2,..., Dev N}, N > 1; each of the electronic devices is configured with at least one intelligent Q&A engine;

[0103] The method includes:

[0104] S310: Let i = 1;

[0105] S320: Determine whether the electronic device Dev i generates a Q&A corpus L i ; the Q&A corpus includes a question corpus S i and an answer corpus O i ;

[0106] If so, save the Q&A corpus L i to the pre-submission corpus and proceed to step S330;

[0107] If not, add the electronic device Dev i as a verification electronic device to the verification device set and proceed to step S330;

[0108] S330: Determine whether i < N holds. If so, set i = i + 1 and return to step S320;

[0109] If not, proceed to step S340;

[0110] S340: Use the verification electronic devices in the verification device set to perform pre-submission operations on each Q&A corpus saved in the pre-submission corpus;

[0111] S350: Update the Q&A standard corpus according to the feedback results of the pre-submission operations. The update includes re-classifying and aligning or updating the alignment of the Q&A corpora in the Q&A standard corpus according to the question corpora and answer corpora.

[0112] Preferably, in the above Figure 2 embodiment, different verification electronic devices call different intelligent Q&A engines.

[0113] In Figure 3 the intelligent customer service Q&A scenario, it is similar to the foregoing Figure 2 . The difference is Figure 3 that the generated Q&A corpora usually appear as "Q1 - W1". After being edited and maintained by user i, it is necessary to determine and verify and then enter the standard Q&A corpus for update. At this time, it needs to be agreed by the other (N - 1) users before it can be entered and updated. Here, the unanimous agreement means that as long as one disagrees, it cannot be entered into the library.

[0114] Specifically, use the question corpus corresponding to each Q&A corpus as the input of the intelligent Q&A engine of each verification electronic device, and determine the feedback result of the pre-submission operation based on the comparison result between the output result of the intelligent Q&A engine and the answer corpus corresponding to each Q&A corpus.

[0115] Continuing the above example, assume that in a certain APP scenario, there are 5 active user devices A1, A2, A3, A4, A5; among them, user device A2 edits and maintains a Q&A corpus of "Question 1 - Answer 1" and is ready to submit it to the corpus to obtain a reward, and the other users (A1, A3, A4, A5) do not generate new corpora

[0116] At this time, first save the Q&A corpus "Question 1 - Answer 1" to the pre-submission corpus;

[0117] After 5 loop iterations, assuming that the question and answer corpus saved in the pre-submission corpus only has "Question 1-Answer 1"; the verification device set is {A1, A3, A4, A5}.

[0118] At this time, only when the other four devices {A1, A3, A4, A5} unanimously agree that the question and answer corpus "Question 1-Answer 1" can be stored, the feedback result of the pre-submission operation is a positive feedback result, and the operation of updating the standard corpus can be performed based on the positive feedback result.

[0119] Therefore, the question corpus corresponding to each question and answer corpus is used as the input of the intelligent question and answer engine of each verification electronic device. When the similarity between the output result of the intelligent question and answer engine of each verification electronic device and the answer corpus corresponding to each question and answer corpus is greater than a preset threshold, the feedback result of the pre-submission operation is determined to be a positive feedback result to reflect atomicity and standard consistency.

[0120] See next Figure 4 , Figure 4 yes Figure 1 A schematic diagram of the hardware / software architecture of a specific implementation scenario of the method. The scenario includes a standard corpus, a cluster server, and multiple electronic devices, each of which is configured with at least one AI engine, including an intelligent customer service engine, an intelligent question-and-answer engine, an intelligent translation engine, etc. The corresponding standard corpus can be based on Figure 1-Figure 3 The sustainably-improvable corpus question classification method is continuously improved and updated.

[0121] Figure 5 The present invention is a schematic diagram of the functional module unit composition of a continuously improved corpus question classification system according to an embodiment of the present invention.

[0122] The system includes a cluster server and a standard corpus, wherein the cluster server communicates wirelessly with N electronic devices, and the system further includes:

[0123] A corpus detection unit, used to detect whether the electronic device generates parallel corpus;

[0124] A pre-submission unit, used for saving the parallel corpus generated by the electronic device to a pre-submission corpus;

[0125] A verification device designation unit, used to designate the electronic device as a verification electronic device when the electronic device does not generate parallel corpus;

[0126] A pre-operation execution unit, configured to use a verification electronic device to perform a pre-submission operation on each parallel corpus stored in the pre-submission corpus;

[0127] An updating unit, configured to update the standard corpus according to the feedback result of the pre-submission operation, wherein the updating includes reclassifying the parallel corpus of the standard corpus according to the source corpus or the target corpus;

[0128] Among them, the cluster server is a translation server and / or a question and answer server; the electronic device is configured with a translation engine or an intelligent question and answer engine; and N is an integer greater than 2.

[0129] The verification device designation unit configures all verification electronic devices to form a distributed transaction execution unit; the pre-commit operation performed by the distributed transaction execution unit on each parallel corpus stored in the pre-commit corpus constitutes a distributed atomic transaction operation.

[0130] In distributed atomic transaction operations, a transaction is a reliable, independent unit of work consisting of a group of operations, in which all operations either succeed or fail. Transactions have four basic characteristics: atomicity, consistency, isolation, and persistence, also referred to as the ACID characteristics of transactions.

[0131] By comparing the above method embodiments, we can understand this atomicity more clearly, namely:

[0132] Figure 2 In a preferred embodiment, the feedback result of the pre-submission operation is determined to be positive feedback only when the translation results output by the translation engines of the translation APPs used by all other user devices (A1, A3, A4, A5) all contain the corpus to be translated corresponding to the translation corpus, and the feedback result of the pre-submission operation is determined to be positive feedback; as long as the translation result output by the translation engine of the translation APP used by one user device does not contain the corpus to be translated corresponding to the translation corpus at all, there is no need to submit, that is, the feedback result is negative feedback, and the operation of updating the standard corpus is not performed.

[0133] Figure 3 In the embodiment of , the question and answer corpus is usually represented as "Q1-W1". After being edited and maintained by user i, it needs to be updated by entering the standard question and answer corpus after verification. At this time, it needs to be agreed by other (N-1) users before it can be updated. The unanimous agreement here means that if there is only one disagreement, it cannot be entered into the database.

[0134] Therefore, the technical solution of the present invention can enable the constructed corpus information to be continuously improved and updated, and the update process can effectively avoid corpus entry errors caused by the accumulation of subjective deviations of individual users or devices; in addition, the update process is based on a distributed transaction processing mechanism to ensure the consistency of standards, and the advantages include at least:

[0135] (1) The user (device / terminal) who submits the corpus is separated from the user (device / terminal) who verifies the corpus: When a user (device / terminal) submits a corpus, the user (device / terminal) cannot be used as a verification device to achieve cross-verification.

[0136] (2) Before formal submission, the corpus must first be pre-submitted to ensure standard consistency.

[0137] (3) Based on the feedback results of the pre-commit operation, the standard corpus is updated to reflect the distributed atomic transaction operation.

[0138] For other technologies, principles, algorithms or models not elaborated in detail in this application, please refer to the prior art.

[0139] In the aforementioned embodiment section, the present invention provides multiple embodiments, each of which can constitute an independent technical solution and may contribute to the prior art and solve corresponding technical problems.

[0140] However, it should be pointed out that different embodiments can be combined with each other without violating logic; at the same time, each embodiment can solve at least one technical problem, but it is not required that each individual embodiment must solve multiple or all technical problems.

[0141] At the same time, in the specific implementation of this application, if user-related data is involved, when the embodiment of this application is applied to a specific product or technology, the user's permission or consent must be obtained, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0142] The foregoing has shown and described the method embodiments and system of the present invention, but it is understandable to those skilled in the art that various changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for classifying corpus problems with sustainable improvement, the method being implemented based on a cluster, the cluster including N electronic devices, N > 1; It is characterized in that The method includes the following steps: S110: Let i = 1; S120: Determine whether the electronic device generates parallel corpus; the parallel corpus includes source corpus and target corpus; If yes, save the parallel corpus to the pre-submission corpus and proceed to step S130; If no, add the electronic device as a verification electronic device to the verification device set and proceed to step S130; S130: Determine whether i < N holds. If yes, let i = i + 1 and return to step S120; If no, proceed to step S140; S140: Use the verification electronic devices in the verification device set to perform pre-submission operations on each parallel corpus saved in the pre-submission corpus; S150: Update the standard corpus according to the feedback results of the pre-submission operations. The update includes reclassifying the parallel corpus in the standard corpus according to the source corpus or the target corpus.

2. A method for sustainably improving corpus question classification as claimed in claim 1, characterized in that: The cluster is a translation server cluster, and the electronic device is configured with at least one translation engine; the parallel corpus is translation corpus, the source corpus is the corpus to be translated, the target corpus is the translated corpus, and the corpus to be translated and the translated corpus are in different languages.

3. A method for sustainably improving corpus question classification as claimed in claim 1, characterized in that: The cluster is a question-and-answer server cluster; the parallel corpus is question-and-answer corpus, the source corpus is question corpus, the target corpus is answer corpus, and the question corpus and the answer corpus are in the same language.

4. A method for sustainably improving corpus question classification as claimed in claim 2, characterized in that: The specific content of step S140 includes: Use the corpus to be translated corresponding to each translation corpus as the input of the translation engine of each verification electronic device, and determine the feedback result of the pre-submission operation based on the comparison result between the output result of the translation engine and the translated corpus corresponding to each translation corpus.

5. A method for sustainably improving corpus question classification as claimed in claim 3, characterized in that: Each of the electronic devices is configured with an intelligent question-and-answer engine; The specific content of step S140 includes: Use the question corpus corresponding to each question-and-answer corpus as the input of the intelligent question-and-answer engine of each verification electronic device, and determine the feedback result of the pre-submission operation based on the comparison result between the output result of the intelligent question-and-answer engine and the answer corpus corresponding to each question-and-answer corpus.

6. A method for sustainably improving corpus question classification as claimed in claim 2, characterized in that: The intelligent question-and-answer engines called by different verification electronic devices are different.

7. A method for sustainably improving corpus question classification as claimed in claim 2, characterized in that: The translation engines called by different verification electronic devices are different.

8. A continuously improved corpus question classification system, the system comprising a cluster server and a standard corpus, the cluster server wirelessly communicating with N electronic devices, characterized in that: The system further includes: A corpus detection unit for detecting whether the electronic device generates parallel corpus; A pre-submission unit for saving the parallel corpus generated by the electronic device to the pre-submission corpus; A verification device designation unit for designating the electronic device as a verification electronic device when the electronic device does not generate parallel corpus; A pre-operation execution unit for using the verification electronic device to perform pre-submission operations on each parallel corpus saved in the pre-submission corpus; An update unit for updating the standard corpus according to the feedback results of the pre-submission operations. The update includes reclassifying the parallel corpus in the standard corpus according to the source corpus or the target corpus; Among them, the cluster server is a translation server and / or a question and answer server; the electronic device is configured with a translation engine or an intelligent question and answer engine; and N is an integer greater than 2.

9. A continuously improving corpus question classification system as claimed in claim 8, characterized in that: The verification device designation unit configures all verification electronic devices to form a distributed transaction execution unit; the pre-commit operation performed by the distributed transaction execution unit on each parallel corpus stored in the pre-commit corpus constitutes a distributed atomic transaction operation.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are executed by a processor, the steps of a sustainably improved corpus question classification method as described in any one of claims 1 to 7 are implemented.

Images