Text processing method and device, electronic equipment and readable storage medium

Abstract

The invention discloses a text processing method and device, electronic equipment and a readable storage medium, and the method comprises the steps: obtaining a to-be-processed text, and determining the number of second texts with the same content as the to-be-processed text in a first database, the first database comprises a plurality of second texts, a processing state corresponding to each second text, and a processing result corresponding to the processed second text in the processing state; in response to the fact that the number is larger than or equal to a preset number threshold value and the processing results corresponding to the multiple second texts with the same content as the to-be-processed text are the same, the processing results serve as the processing results of the to-be-processed text. According to the method, the proportion of manual auditing is reduced, and the labor division and opportunity of machine auditing and manual auditing are clearer.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and specifically to a text processing method, device, electronic device and readable storage medium. Background Art

[0002] The current methods for reviewing texts generated in games and social platforms include: machine review, human review, and a combination of the two.

[0003] Among them, machine review and manual review are combined to give full play to their respective advantages and make up for their respective disadvantages. However, the traditional combination of machine review and manual review is simply a combination of the two. Since the update of machine review rules often lags behind the emergence of new illegal content, and the iterative update speed of algorithms and models often lags behind the emergence of new illegal content, the machine may not be able to review these texts containing new illegal content. Texts that cannot be reviewed by machines are transferred to manual processing, making manual review account for a higher proportion of the review. Therefore, the review path in the existing combined machine review and manual review is unreasonable, making the combined machine review and manual review method more costly. Summary of the invention

[0004] The present application provides a text processing method, device, electronic device and readable storage medium, which solves the problem that the division of labor and timing of machine review and manual review are unclear due to the unreasonable review path in the existing review that combines machine review and manual review.

[0005] In a first aspect, an embodiment of the present application provides a text processing method, the method comprising:

[0006] Acquire a text to be processed, and determine the number of second texts in a first database that have the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is that the text has been processed;

[0007] In response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to the plurality of second texts having the same content as the text to be processed are the same, the processing result is used as the processing result of the text to be processed.

[0008] In a second aspect, an embodiment of the present application further provides a text processing device, the device comprising:

[0009] A text acquisition module, used to acquire a text to be processed, and determine the number of second texts in a first database having the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is processed;

[0010] The processing module is used for responding to the fact that the number is greater than or equal to a preset number threshold and the processing results corresponding to the plurality of second texts having the same content as the text to be processed are the same, and taking the processing result as the processing result of the text to be processed.

[0011] In a third aspect, an embodiment of the present application further provides an electronic device, including:

[0012] Processor; and

[0013] The memory is used to store a data processing program. After the electronic device is powered on and the program is run by the processor, the method described in the first aspect is executed.

[0014] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium storing a data processing program, which is executed by a processor to perform the method described in the first aspect.

[0015] Compared with the prior art, this application has the following advantages:

[0016] A text processing method provided in an embodiment of the present application includes: obtaining a text to be processed, determining the number of second texts in a first database that have the same content as the text to be processed, wherein the first database includes multiple second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is processed; in response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to multiple second texts with the same content as the text to be processed being the same, using the processing result as the processing result of the text to be processed. This method reduces the proportion of manual review, makes the division of labor and timing of machine review and manual review clearer, and reduces the cost of human-computer collaboration. This method can fully utilize the efficiency and consistency of machine review, and the accuracy of manual review. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of a text processing method provided in an embodiment of the present application;

[0018] Figure 2 It is a schematic diagram of obtaining mkey corresponding to a text provided in an embodiment of the present application;

[0019] Figure 3It is a flowchart of another text processing method provided in an embodiment of the present application;

[0020] Figure 4 This is an example diagram of obtaining the review result of a text to be reviewed provided in an embodiment of the present application;

[0021] Figure 5 It is a schematic diagram of a processing process of a text to be reviewed provided in an embodiment of the present application;

[0022] Figure 6 It is a structural block diagram of a text processing device provided in an embodiment of the present application;

[0023] Figure 7 A structural block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0025] It should be noted that the terms "first", "second", "third", etc. in the claims, description and drawings of the present application are used to distinguish similar objects and are not used to describe a specific order or sequence. The data used in this way are interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including", "having" and their variants 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.

[0026] With the rapid development of the Internet and digital technology, the ways people communicate and interact on games and social platforms have become more diverse, generating a massive amount of text. These texts include not only normal communication information, but also insults, abuse or hate speech between users. Therefore, it is necessary to review the texts generated in games and social platforms in real time.

[0027] The current methods for reviewing texts generated in games and social platforms include: machine review, human review, and a combination of the two.

[0028] Among them, manual review uses people's language comprehension ability and background knowledge to deeply understand the semantics and metaphors of the text, and can more accurately identify illegal content. It can also identify some implicit illegal content well. For some new types of illegal content, flexible judgments can be made according to the specific circumstances. However, manual review requires a lot of time and human resources, and the cost is high. The ability and efficiency of manual review are limited by human resources, the review speed is slow, and it is impossible to efficiently process large amounts of text, making it difficult to cope with large-scale review needs. The judgment criteria of different reviewers may differ, and the subjective judgment and personal bias of the reviewers may affect the review results, which may lead to certain unfairness and misjudgment risks.

[0029] Machine auditing can process a large amount of text quickly and automatically based on preset rules and algorithms, greatly improving auditing efficiency. In addition, the auditing standards of machine auditing are consistent, avoiding the influence of human subjective factors. However, machine auditing is limited by algorithms and training data, lacks flexible judgment and processing capabilities for complex situations, and has limited accuracy. In addition, machine auditing is difficult to fully understand the semantics and metaphors of the text, and has a weak ability to identify some implicit or new types of illegal content.

[0030] Combine machine review and manual review to give full play to their respective advantages and make up for their respective disadvantages. However, the traditional combination of machine review and manual review is simply a combination of the two. Since the update of machine review rules often lags behind the emergence of new illegal content, and the iteration and update speed of algorithms and models often lags behind the emergence of new illegal content, the machine may not be able to review these texts containing new illegal content. Texts that cannot be reviewed by the machine are transferred to manual processing, making manual review account for a higher proportion of the review. Therefore, the review path in the existing combined machine review and manual review is unreasonable, making the combined machine review and manual review method more costly.

[0031] The embodiment of the present application provides a text processing method, which solves the problem that the audit path in the existing combined machine audit and manual audit is unreasonable, making the audit method of combined machine audit and manual audit more costly. The execution subject of the scheme provided by the present application can be an electronic device, which can be a desktop computer, a laptop computer, a mobile device, a smart watch, a smart TV, a tablet computer, a server, etc., or other devices with data processing and display functions.

[0032] like Figure 1FIG. 1 is a flow chart of a text processing method provided in an embodiment of the present application, wherein the method comprises:

[0033] Step S101, obtaining a text to be processed, and determining the number of second texts in a first database having the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to a second text whose processing status is that the text has been processed;

[0034] In the embodiment of the present application, the text processing method is explained by taking the text audit as an example. Obtaining the text to be processed may be obtaining the text to be audited generated by each social portal in the game. Obtaining the text to be audited generated by each social portal in the game may be to integrate a log system on the game server or client, and use the log system to capture the text of each social portal in the game in real time. Obtaining the text to be audited generated by each social portal in the game may also be to develop or use an existing API interface to extract text from each social portal in the game, store the acquired text in a database, and synchronize the data in the database to a data analysis server or a cloud storage service regularly or in real time. Among them, the structure of the database can be constructed in advance according to the structure of the text, and the amount of text, query efficiency, and security, etc., can be considered when constructing the database. The text collected in real time can also be transmitted from the game server to the data analysis server or cloud storage service.

[0035] Among them, the social entrances within the game are various interfaces or functional points in the game that allow users to interact socially, usually in the game's main interface, menu, toolbar or specific scenes.

[0036] The text to be reviewed includes, but is not limited to: user speech data, user nickname data, and user-defined text. As shown in Table 1, examples of the text to be reviewed provided in the embodiment of the present application are:

[0037] Table 1 Examples of texts to be reviewed

[0038]

[0039] Exemplarily, the user's speech is the recorded communication content of the user during the game, such as text conversations in different chat interfaces such as public channels, private messages, team chats, and guild chats in the game; it can also be voice communication (such as team tactical discussions, small talk, etc.) conducted by users through built-in or third-party voice software in multiplayer online games; it can also be posts, comments, and feedback posted by users in game-related forums, social media groups, official or unofficial communities; it can also be technical support requests, complaints or suggestions submitted by users to game companies, and communication records with customer service personnel; it can also be real-time voice and barrage comments contained in live or recorded game videos. The user's nickname is the user's name or username in the game, which is the unique identifier of the user in the game. User-defined text is a variety of personalized information and dynamics that users can display and share on their personal homepages or similar social spaces, such as personal avatars, cover photos, personal signatures, or status information set by users. User-defined text can also be game experiences, insights, strategies, or interesting things recorded and shared by users.

[0040] In some embodiments, after the step of obtaining the text to be processed, the method further includes:

[0041] The text to be processed is preprocessed so that the text to be processed is in a preset text format.

[0042] In the embodiment of the present application, according to the preset text format, the text information in the text to be reviewed is extracted, and the text information is recombined according to the preset text format to obtain the preprocessed text corresponding to the text to be reviewed. The text information is the specific content of the text.

[0043] It should be noted that the formats of the obtained texts to be reviewed may be different, and the data formats may be predefined. For the texts, the text information corresponding to each structure is extracted according to the predefined text format.

[0044] For example, according to the specific structure of the predefined text format, the text information corresponding to each structure can be extracted from the text to be reviewed, and the extracted text information can be recombined according to the predefined text format to obtain the preprocessed text that conforms to the predefined text format corresponding to the text to be reviewed. All texts are organized into a unified text format to facilitate the system to carry out the next step of processing.

[0045] For example, as shown in Table 2, a partial example of a text to be reviewed and a corresponding preprocessed text after preprocessing is provided in an embodiment of the present application:

[0046] Table 2 The text to be reviewed and the corresponding preprocessed text after preprocessing

[0047]

[0048] That is, through preprocessing, the text to be reviewed is converted into a predefined text format, wherein the predefined text format may be a JSON format.

[0049] In some embodiments, preprocessing the acquired text to be processed further includes:

[0050] The text that is not dynamically generated by the user or will not cause violations dynamically by the user is removed from the preprocessed text.

[0051] It should be noted that text that is not dynamically generated by users or text that will not cause violations by users is text that does not include text generated by users. The above data can be messages automatically sent by the game system, such as welcome messages, task instructions, game tutorials, system announcements, etc. The above data can also be pre-written scripts in the game. The above data can also be descriptions of in-game items, equipment, skills, etc. The above data can also be button labels, menu options, status bar information, etc. in the game interface. Filtering text that is not dynamically generated by users or text that will not cause violations by users can reduce the number of texts to be processed.

[0052] The filtering process may also be to remove irrelevant information in the preprocessed text, for example, to remove HTML tags, special characters, etc. in the preprocessed text.

[0053] In the embodiment of the present application, the number of second texts in the first database having the same content as the text to be processed is determined, and the mkey corresponding to the text to be reviewed can be generated according to a preset mkey generation rule.

[0054] Generate the mkey corresponding to the text to be reviewed according to the preset mkey generation rules, including:

[0055] The mkey of the text to be reviewed is calculated based on the user identification and the text information.

[0056] It should be noted that mkey is used to identify text, and mkey is used to determine whether multiple texts are the same text. If the mkeys of multiple texts are the same, it means that the multiple texts are the same text; if the mkeys of multiple texts are different, it means that the multiple texts are different texts.

[0057] Texts with the same mkey can be processed in the same way, such as texts with the same mkey can be reviewed in the same way.

[0058] For example, as shown in Table 3, an example of a text to be processed and its corresponding mkey is provided in an embodiment of the present application:

[0059] Table 3 Text and its corresponding mkey examples

[0060]

[0061] For example, Figure 2 As shown, it is a schematic diagram of obtaining the mkey corresponding to the text provided by an embodiment of the present application. After reading the text A, text B and text C in the above Table 3, the mkey corresponding to the text A, text B and text C is generated, and the mkey1, mkey2 and mkey3 corresponding to the text A, text B and text C are obtained. Assuming that mkey1 and mkey3 are the same, it can be determined that text A and text C are the same text, and text A and text C can use the same audit method.

[0062] It should be noted that the mkey of the generated text can be obtained by performing the md5 algorithm on the text, such as performing the md5 algorithm on the text to be reviewed to obtain the mkey of the text, that is, mkey=md5(id, content).

[0063] In an embodiment of the present application, three types of data pools are pre-constructed, namely, a report order pool (ROP), an arbitration order pool (AOP), and a cache order pool (COP). Among them, the report order pool ROP stores historical texts, the audit status of each historical text, and the audit results of historical texts in an audited state; the audit status of a historical text refers to the different stages in the process of checking and verifying the historical text, including pending and audited. The audit result refers to the conclusion or status drawn after checking and verifying the historical text. The arbitration pool AOP is used to store the texts to be audited and the audit results of the texts to be audited when the number of texts to be audited in the report order pool is greater than or equal to the quantity threshold, but the audit results of these texts are inconsistent. The cache pool COP is used to temporarily store texts to be audited whose audit flow is unclear.

[0064] Step S102, in response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to a plurality of second texts having the same content as the text to be processed are the same, taking the processing result as the processing result of the text to be processed.

[0065] In some embodiments, after the step of responding that the quantity is greater than or equal to the quantity threshold, the method further includes:

[0066] In response to the processing status of at least one of the second texts having the same content as the text to be processed being unprocessed, the text to be processed is cached in a third database; the third database is used to store the text to be processed whose processing flow is unclear.

[0067] In the embodiments of the present application, Figure 3 As shown, it is a flowchart of another text processing method provided by an embodiment of the present application. Determine whether there is text with the same mkey as the text to be reviewed in the work order pool ROP and the number of texts. If there is text with the same mkey as the text to be reviewed in the work order pool ROP and the number of texts reaches a preset number threshold, then execute scheme 2 (for scheme 2, detailed introduction is given in the following embodiment).

[0068] When the number is greater than or equal to the preset number threshold, the processing result of the text to be processed is determined according to the processing status result corresponding to each second text. If all the second texts are in the processed state and the processing results of all the second texts are the same, the processing results of all the second texts are used as the processing results of the text to be processed. If all the second texts are in the processed state, but there is at least one second text processing result that is different from the processing results of other second texts among the processing results of all the second texts, the text to be processed is temporarily cached in the cache pool and waits for subsequent processing.

[0069] In some embodiments, in response to the number being less than a number threshold, the text to be processed is input into a pre-trained classification model to obtain a label and label credibility for the text to be processed output by the pre-trained classification model;

[0070] Based on the label of the text to be processed and the label credibility, a processing result of the text to be processed is obtained.

[0071] In some embodiments, obtaining a processing result of the text to be processed based on the label and label credibility of the text to be processed includes:

[0072] In response to the label credibility being greater than or equal to a preset credibility threshold, taking the label output by the pre-trained classification model as a processing result of the text to be processed;

[0073] In response to the label credibility being less than the preset credibility threshold, the text to be processed is assigned to manual processing, and the manual annotation label of the text to be processed is used as the processing result of the text to be processed.

[0074] In the embodiments of the present application, Figure 3As shown, it is determined whether there is a text with the same mkey as the text to be reviewed in the work order pool ROP and the number of texts. If there is no text with the same mkey as the text to be reviewed in the work order pool ROP, or the number of texts does not reach the quantity threshold, then scheme one is executed (scheme one is described in detail in the following embodiment). Among them, there is no text with the same mkey as the text to be reviewed in the work order pool ROP, which means that the text to be reviewed has never been reviewed.

[0075] Exemplarily, read text A, text B and text C, and compare the mkey1, mkey2 and mkey3 corresponding to text A, text B and text C with the mkey of each text in the work order pool ROP, and determine whether there is the same mkey1, the same mkey2 and the same mkey3 in the work order pool ROP. Assuming that there is no same mkey3 in the work order pool ROP or the number of texts corresponding to the same mkey3 does not reach the quantity threshold, the audit result of the text to be audited is determined by scheme one.

[0076] Option 1:

[0077] Step A: inputting the text to be reviewed into a pre-trained classification model, and the pre-trained classification model outputs the label of the text to be reviewed and the label credibility;

[0078] Step B: comparing the tag credibility with a preset credibility threshold corresponding to the tag;

[0079] Step C: In response to the label credibility being greater than or equal to the preset credibility threshold, the text to be reviewed is reviewed by a machine, and the label output by the pre-trained classification model is used as the review result of the text to be reviewed;

[0080] Step D: In response to the label credibility being less than the preset credibility threshold, the text to be reviewed is assigned to manual processing, and the manual annotated label for the text to be reviewed is used as the review result of the text to be reviewed.

[0081] For example, Figure 4 As shown in FIG. 1 , an example diagram of obtaining the audit result of a text to be audited is provided in an embodiment of the present application. Figure 4In the example, assuming that text 1, text 2, and text 3 do not have the same mkey in the work order pool ROP, or that text 1, text 2, and text 3 have the same mkey in the work order pool ROP, but the number of texts corresponding to the mkey does not reach the quantity threshold, then text 1, text 2, and text 3 are input into the pre-trained classification model respectively. The pre-trained classification model outputs label 1 and label credibility 1 corresponding to text 1, label 2 and label credibility 2 corresponding to text 2, and label 3 and label credibility 3 corresponding to text 3. If label credibility 1 is 0.99, label credibility 2 is 0.72, label credibility 3 is 0.99, and the preset credibility thresholds corresponding to label 1, label 2, and label 3 are all 0.9, then label credibility 1, label credibility 2, and label credibility 3 are compared with the preset credibility thresholds respectively. If label credibility 1 and label credibility 3 are respectively greater than the preset credibility thresholds, the machine directly outputs the label of the text to be reviewed 1 output by the pre-trained classification model as the review result of the text to be reviewed 1, and the machine directly outputs the label of the text to be reviewed 3 output by the pre-trained classification model as the review result of the text to be reviewed 3, and then determines whether the text to be reviewed 1 and the text to be reviewed 3 are in violation according to the review results of the text to be reviewed 1 and the text to be reviewed 3. If label credibility 2 is less than the preset credibility threshold, the text to be reviewed is sent to manual review, and the label of the text to be reviewed 2 is manually annotated, and the manually labeled label is used as the review result of the text to be reviewed 2.

[0082] It should be noted that the pre-trained classification model (SDK) can label text and give label credibility. It is a pre-trained algorithm or model. Among them, obtaining a pre-trained classification model includes:

[0083] A training set is collected, wherein the training set includes multiple texts and labels corresponding to the texts and label credibility; wherein the labels of the texts are set according to the needs of analyzing the texts, such as spam, sensitive words, emotional tendencies, positive, negative, and neutral, etc.

[0084] The parameters of the initial classification model are adjusted through the training set until the output of the classification model converges to obtain a pre-trained classification model.

[0085] It should be noted that label confidence refers to a quantitative indicator of the confidence level of the label or classification result of a given data point in data annotation or machine learning classification tasks.

[0086] In data labeling projects, label credibility reflects the confidence of the labeler in a sample label. For example, when labeling the object category in a picture, the labeler may give a higher credibility score; on the contrary, if it is difficult to distinguish the object category, the labeler may give a lower credibility score.

[0087] In machine learning, especially supervised learning and semi-supervised learning, label credibility is often associated with the confidence of the model's prediction. When the model classifies new data, it outputs a probability distribution, indicating the likelihood that the data belongs to each category. The category corresponding to the highest probability is usually regarded as the prediction result, and the probability value itself can be regarded as the model's confidence in the classification result.

[0088] It should be noted that different tag types correspond to different preset credibility thresholds, and a preset credibility threshold matching each tag category can be pre-set. As shown in Table 4, examples of preset credibility thresholds matching some tag categories with each tag category provided in the embodiment of the present application are as follows:

[0089] Table 4 Examples of some label categories and the preset confidence thresholds for matching each label category

[0090] Tag Category Preset confidence threshold for matching Tag 1 0.9 Tag 2 0.85 …… ……

[0091] After obtaining the label of each text, the preset confidence threshold matching the label can be obtained.

[0092] In some embodiments, there are multiple texts to be processed, and the method also includes: in response to the content of a first text to be processed and a second text to be processed being the same, using the processing result of the first text to be processed as the processing result of the second text to be processed, the first text to be processed is a processed text to be processed, and the first text to be processed and the second text to be processed are any two of the texts to be processed.

[0093] In the embodiment of the present application, when there are multiple texts to be processed, the same audit texts can be used for the texts to be reviewed with the same content. If the text to be reviewed 1 and the text to be reviewed 2 are texts to be reviewed with the same content, and the text to be reviewed 1 obtains the corresponding audit result through the above method, then the audit result of the text to be reviewed 1 can be directly used as the audit result of the text to be reviewed 2.

[0094] For example, Figure 5 As shown, it is a schematic diagram of a processing process of a text to be reviewed provided by an embodiment of the present application. During the processing of the text to be reviewed, multiple dynamic adjustments are performed, as follows:

[0095] like Figure 5 As shown in (a)-(d), the audit path of the text to be audited is dynamically adjusted according to whether there is a text with the same mkey as the text to be audited in the work order pool. Figure 5As shown in (a), if there is no text with the same mkey as the text to be reviewed in the work order pool, or the number of texts with the same mkey as the text to be reviewed does not reach the preset number threshold, the text to be reviewed is input into the pre-trained classification model, and the pre-trained classification model outputs the label of the text to be reviewed and the label credibility; according to the label credibility and the preset credibility threshold corresponding to the label, the audit path of the text to be reviewed is dynamically adjusted. That is, the label credibility is compared with the label threshold corresponding to the label. If the label credibility is greater than or equal to the preset credibility threshold corresponding to the label, the label of the text to be reviewed output by the pre-trained classification model is used as the audit result of the text to be reviewed by the machine. If the label credibility is less than the preset credibility threshold corresponding to the label, the text to be reviewed is manually reviewed to obtain the audit result of the audit text.

[0096] In some embodiments, the method further includes: in response to the processing results of the second text having the same content as the text to be processed being at least partially different, determining whether there is a third text having the same content as the text to be processed in the second database; wherein the second database is used to store text to be processed when the number of second texts having the same content as the text to be processed in the first database is greater than or equal to a preset number threshold, and when a plurality of the second texts are all processed but the processing results corresponding to the plurality of the second texts are at least partially different, the processing status corresponding to the text to be processed, and the processing result whose processing status is processed;

[0097] In response to the absence of a third text having the same content as the text to be processed in the second database, assigning the text to be processed to manual processing, and using the manual annotation label of the text to be processed as the processing result of the text to be processed;

[0098] In response to the presence of a third text identical to the text to be processed in the second database and the third text having been processed, the processing result of the third text is used as the processing result of the text to be processed.

[0099] In some embodiments, in response to the absence of a third text identical to the text to be processed in the second database, after the step of assigning the text to be processed to manual processing and using the manual annotation label of the text to be processed as the processing result of the text to be processed, the method further includes:

[0100] Updating the processing status of the text to be processed;

[0101] The to-be-processed text, the processing status corresponding to the to-be-processed text, and the processing result of the to-be-processed text whose processing status is processed are stored in the second database.

[0102] In some embodiments, after the step of responding to the presence of a third text identical to the text to be processed in the second database, the method further comprises:

[0103] In response to the processing result of the third text being unprocessed, the to-be-processed text is cached in a third database.

[0104] In some embodiments, after the step of caching the to-be-processed text in a third database, the method further comprises:

[0105] In response to at least one of the to-be-processed texts being stored in the third database, acquiring a storage duration of the to-be-processed text;

[0106] In response to the storage time being greater than or equal to a time threshold, the text to be processed is assigned to manual processing, and a manual label for the text to be processed is used as a processing result of the text to be processed;

[0107] In response to the storage duration being less than the duration threshold, the step of responding to the number being greater than or equal to the number threshold is continued to obtain a processing result of the text to be processed.

[0108] In the embodiments of the present application, Figure 3 and Figure 5 As shown in (b)-(d), if the number of texts with the same mkey as the text to be reviewed in the work order pool is greater than or equal to the preset quantity threshold, the review path of the text to be reviewed is dynamically adjusted according to the review status of all texts with the same mkey as the text to be reviewed. If the review status of all texts with the same mkey as the text to be reviewed is not all reviewed, the text to be reviewed is cached in the cache pool.

[0109] If the mkey of all the texts is the same as the text to be reviewed, the review status is all reviewed, such as Figure 5 As shown in (d)-(e), the audit path of the text to be reviewed is dynamically adjusted according to the audit results of all texts with the same mkey as the text to be reviewed, including: if the audit results of all texts with the same content as the text to be reviewed are consistent, then according to the audit results of all texts with the same mkey as the text to be reviewed, the audit results of all texts with the same content as the text to be reviewed are used as the audit results of the text to be reviewed by the machine, and the text to be reviewed and the audit results of the machine audit are stored in the work order pool. And the audit status of the text to be reviewed is set to reviewed.

[0110] If the audit results of all texts are inconsistent, check whether there is any text with the same mkey as the text to be audited in the arbitration pool. If there is no text with the same mkey as the text to be audited in the arbitration pool, put the text to be audited into the arbitration pool, manually annotate the text to be audited, and store the text to be audited in the arbitration pool with the manually annotated label, and set the audit status of the text to be audited to audited.

[0111] If there is a text with the same mkey as the text to be reviewed in the arbitration pool, determine whether there is an audit result for the text with the same mkey as the text to be reviewed in the arbitration pool. If there is no audit result, put the text to be reviewed into the cache pool and wait for the next processing. If there is an audit result, the machine marks the audit result of the text with the same mkey as the text to be reviewed as the audit result of the text to be reviewed, and stores the text to be reviewed and the audit result of the machine audit in the work order pool, and sets the audit status of the text to be reviewed to reviewed.

[0112] In an embodiment of the present application, for the texts stored in the buffer pool, the above-mentioned dynamic adjustment is repeated until all the texts to be reviewed are reviewed.

[0113] Exemplarily, read text A, text B and text C, and compare the mkey1, mkey2 and mkey3 corresponding to text A, text B and text C with the mkey of each text in the work order pool ROP, and determine whether there is the same mkey1, the same mkey2 and the same mkey3 in the work order pool ROP. Assuming that there are the same mkey1 and the same mkey2 in the work order pool ROP, and the number of texts with the same mkey1 and the same mkey2 is greater than or equal to the quantity threshold, then text A and text B determine the audit method by executing scheme 2.

[0114] Solution 2 is as follows:

[0115] Step 1: Determine whether all texts with the same mkey as the text to be reviewed have been reviewed;

[0116] If all are reviewed, proceed to step 2;

[0117] Otherwise, go to step 3;

[0118] Step 2: Determine whether the audit results of the text with the same mkey as the text to be audited are consistent;

[0119] If they are consistent, go to step 4;

[0120] If not, proceed to step 5;

[0121] Step 3: Put the text to be reviewed into the cache pool and wait for the next processing;

[0122] Step 4: The machine reviews the text to be reviewed according to the review result of the text with the same mkey as the text to be reviewed, and obtains the review result; and stores the text to be reviewed and the review result of the machine review in the work order pool, and sets the review status of the text to be reviewed to reviewed;

[0123] Step 5: Check whether there is text with the same mkey in the arbitration pool;

[0124] If not, go to step 6;

[0125] If yes, go to step 7;

[0126] Step 6: The text to be reviewed is put into the arbitration pool, and the review result is obtained by manually reviewing the text to be reviewed; the text to be reviewed and the review result of the manual review are stored in the arbitration pool, and the review status of the text to be reviewed is set to reviewed;

[0127] Step 7: Determine whether the text with the same mkey as the text to be reviewed has an audit result;

[0128] If not, proceed to step 8;

[0129] If yes, go to step 9:

[0130] Step 8: Put the text to be reviewed into the cache pool and wait for the next processing;

[0131] Step 9: The machine reviews the text to be reviewed according to the review result of the text with the same mkey as the text to be reviewed, and obtains the review result; and stores the text to be reviewed and the review result of the machine review in the work order pool, and sets the review status of the text to be reviewed to reviewed;

[0132] Step 10: If there is text stored in the cache pool, pull the text in the cache pool at a preset time interval; determine whether the cache time of each text exceeds the preset time;

[0133] If it exceeds, go to step 11;

[0134] If not, proceed to step 12:

[0135] Step 11: The text is manually reviewed to obtain the review result; the text to be reviewed and the review result of the machine review are stored in the work order pool, and the review status of the text to be reviewed is set to reviewed;

[0136] Step 12: Repeat steps 1-11 above until all the texts to be reviewed have been reviewed.

[0137] For example, Figure 3As shown, assuming that the quantity threshold is 2, the mkey of the text to be reviewed is compared with the mkey of all texts in the work order pool; if the number of texts with the same mkey in the work order pool exceeds 2, such as 3, it means that the same text has been reviewed, and the review status of these 3 texts is determined. If the 3 texts are not all in the reviewed state, the text to be reviewed is temporarily stored in the above-mentioned cache pool and waits for the next processing. If the 3 texts are all in the reviewed state, it is determined whether the review results of the 3 texts are consistent. If they are consistent, the machine review is performed according to the review results of the 3 texts, and the text to be reviewed and the machine review results are imported into the work order pool, and the work order status is set to reviewed. If the review results of the 3 texts are inconsistent, first query whether there is a text with the same mkey in the arbitration pool. If the arbitration pool has a text with the same mkey and has been reviewed, the machine review is directly performed according to the review results of the arbitration pool, and the text to be reviewed and the machine review results are imported into the work order pool, and the work order status is set to reviewed. If the arbitration pool has text with the same mkey but has not been reviewed, the text to be reviewed will be stored in the cache pool and wait for the next processing. If the arbitration pool does not have text with the same mkey, the text to be reviewed will be stored in the arbitration pool and manually reviewed, and the text to be reviewed and the review results of the manual review will be stored in the arbitration pool, and the review status of the text to be reviewed will be set to reviewed.

[0138] For example, suppose that 96 game products are connected to the text review system, and there are 140 text entries in the 96 game products. Assume that the review volume of texts generated in the 140 text entries is 20 million+ per month. Through the data processing method in this application, the machine review volume is 17 million+, and the machine review rate reaches 85%, which improves the review efficiency of combining manual review and mechanical review, and reduces the review cost.

[0139] It should be noted that a player's speech may be reported by multiple players at different times, or the same speech may be posted at different times, which will generate multiple texts. For example, text A: {"id":"Id123","time":"2024-07-2512:00:00","conent":"Ni Ma","type":"Speech report"}; text B: {"id":"Id123","time":"2024-07-25 12:10:00","conent":"Ni Ma","type":"Speech report"}; text C: {"id":"Id123","time":"2024-07-25 12:20:00","conent":"Ni Ma","type":"Speech report"}, so there is a lot of duplicate data in the texts to be reviewed.

[0140] In this embodiment, according to the mkey generation rule, the mkeys of text A, text B and text C are obtained. Then text A, text B and text C can adopt the same audit method to obtain the audit results, and no subsequent judgment is required, which saves audit resources for auditing repeated texts.

[0141] It should be noted that before storing the text to be reviewed in the cache pool, it is first determined whether the text to be reviewed already exists in the cache pool. If so, it is not stored in the cache pool. By deduplicating the text stored in the cache pool, it is possible to avoid storing the same text to be reviewed in the cache pool multiple times, thus saving review resources.

[0142] In some embodiments, in response to the number being greater than or equal to the number threshold and the processing results corresponding to the plurality of second texts having the same content as the text to be processed being the same, after the step of using the processing result as the processing result of the text to be processed, the method further includes:

[0143] Updating the processing status of the text to be processed;

[0144] The text to be processed, the processing status corresponding to the text to be processed, and the processing result with the processing status being processed are stored in the first database.

[0145] In the implementation of this application, the audit results of each text to be processed and reviewed, and the audit status of the text to be reviewed are updated and stored in the work order pool, so that the previously processed text to be reviewed, the audit status and the audit results can be used as the audit standards for the audit path of the subsequent text to be reviewed.

[0146] In some embodiments, in response to the number being greater than or equal to the number threshold and the processing results corresponding to the plurality of second texts having the same content as the text to be processed being the same, after the step of using the processing result as the processing result of the text to be processed, the method further includes:

[0147] Based on the processing result of the text to be processed, determining whether the user who generated the text to be processed is punished;

[0148] In response to the user of the to-be-processed text being punished, the user who generated the to-be-processed text is punished.

[0149] In the embodiment of the present application, whether the text to be reviewed is a text that violates the rules is determined based on the review result, and the user who generates the text that violates the rules is punished. If it is determined that there is no violation, no punishment is imposed on the user who generates the text.

[0150] It should be noted that when punishing a user, a gradient penalty can be imposed according to the severity of the violation, wherein the severity of the violation can be determined by the label credibility, and the larger the value of the label credibility, the greater the severity of the text violation.

[0151] A text processing method provided in an embodiment of the present application includes: obtaining a text to be processed, determining the number of second texts in a first database that have the same content as the text to be processed, wherein the first database includes multiple second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is processed; in response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to multiple second texts with the same content as the text to be processed being the same, using the processing result as the processing result of the text to be processed. This method reduces the proportion of manual review, makes the division of labor and timing of machine review and manual review clearer, and reduces the cost of human-computer collaboration. This method can fully utilize the efficiency and consistency of machine review, and the accuracy of manual review.

[0152] Corresponding to the text processing method provided in the embodiment of the present application, the embodiment of the present application also provides a text processing device, such as Figure 6 As shown, the device 600 includes:

[0153] The text acquisition module 601 is used to acquire the text to be processed and determine the number of second texts in the first database that have the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is processed;

[0154] The processing module 602 is used to respond to the number being greater than or equal to a preset number threshold and the processing results corresponding to the plurality of second texts having the same content as the text to be processed being the same, and use the processing result as the processing result of the text to be processed.

[0155] In some embodiments, the apparatus further comprises:

[0156] A first sub-processing module, configured to, in response to the number being less than a number threshold, input the to-be-processed text into a pre-trained classification model, and obtain a label and label credibility for the to-be-processed text output by the pre-trained classification model;

[0157] Based on the label of the text to be processed and the label credibility, a processing result of the text to be processed is obtained.

[0158] In some embodiments, the sub-processing module is further configured to:

[0159] In response to the label credibility being greater than or equal to a preset credibility threshold, taking the label output by the pre-trained classification model as a processing result of the text to be processed;

[0160] In response to the label credibility being less than the preset credibility threshold, the text to be processed is assigned to manual processing, and the manual annotation label of the text to be processed is used as the processing result of the text to be processed.

[0161] In some embodiments, the to-be-processed text is multiple, and the apparatus further comprises:

[0162] An assignment module is used to respond to the fact that the contents of a first text to be processed and a second text to be processed are the same, and use the processing result of the first text to be processed as the processing result of the second text to be processed, the first text to be processed is a processed text to be processed, and the first text to be processed and the second text to be processed are any two of the texts to be processed.

[0163] In some embodiments, the apparatus further comprises:

[0164] The first cache module is used to cache the to-be-processed text to a third database in response to the processing status of at least one of the second texts having the same content as the to-be-processed text being unprocessed; the third database is used to store the to-be-processed text whose processing flow is unclear.

[0165] In some embodiments, the apparatus further comprises:

[0166] A second sub-processing module is used for judging whether there is a third text with the same content as the text to be processed in the second database in response to the processing results of the second text with the same content as the text to be processed being at least partially different; wherein the second database is used for storing text to be processed, the processing status corresponding to the text to be processed, and the processing result with the processing status being processed when the number of second texts with the same content as the text to be processed in the first database is greater than or equal to a preset number threshold, and a plurality of the second texts are all processed but the processing results corresponding to the plurality of the second texts are at least partially different;

[0167] In response to the absence of a third text having the same content as the text to be processed in the second database, assigning the text to be processed to manual processing, and using the manual annotation label of the text to be processed as the processing result of the text to be processed;

[0168] In response to the presence of a third text identical to the text to be processed in the second database and the third text having been processed, the processing result of the third text is used as the processing result of the text to be processed.

[0169] In some embodiments, the apparatus further comprises:

[0170] A first updating module, used for updating the processing status of the text to be processed;

[0171] The to-be-processed text, the processing status corresponding to the to-be-processed text, and the processing result of the to-be-processed text whose processing status is processed are stored in the second database.

[0172] In some embodiments, the apparatus further comprises:

[0173] The second cache module is used for caching the to-be-processed text into a third database in response to the processing result of the third text being unprocessed.

[0174] In some embodiments, the apparatus further comprises:

[0175] A third sub-processing module, configured to obtain a storage time of the to-be-processed text in response to at least one to-be-processed text being stored in the third database;

[0176] In response to the storage time being greater than or equal to a time threshold, the text to be processed is assigned to manual processing, and a manual label for the text to be processed is used as a processing result of the text to be processed;

[0177] In response to the storage duration being less than the duration threshold, the step of responding to the number being greater than or equal to the number threshold is continued to obtain a processing result of the text to be processed.

[0178] In some embodiments, the apparatus further comprises:

[0179] A second updating module, used for updating the processing status of the text to be processed;

[0180] The text to be processed, the processing status corresponding to the text to be processed, and the processing result with the processing status being processed are stored in the first database.

[0181] In some embodiments, the apparatus further comprises:

[0182] A penalty module, used for determining whether a user who generates the text to be processed is punished based on the processing result of the text to be processed;

[0183] In response to the user of the to-be-processed text being punished, the user who generated the to-be-processed text is punished.

[0184] In some embodiments, the apparatus further comprises:

[0185] The preprocessing module is used to preprocess the text to be processed so as to convert the text to be processed into a preset text format.

[0186] Corresponding to the text processing method provided in the embodiment of the present application, the embodiment of the present application also provides an electronic device. Figure 7 As shown, the electronic device includes: a processor 701; and a memory 702, which is used to store a program of a text processing method. After the electronic device is powered on and runs the program of the text processing method stored by the processor, the following steps are performed:

[0187] Acquire a text to be processed, and determine the number of second texts in a first database that have the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is that the text has been processed;

[0188] In response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to the plurality of second texts having the same content as the text to be processed are the same, the processing result is used as the processing result of the text to be processed.

[0189] Corresponding to the text processing method provided in the embodiment of the present application, the embodiment of the present application further provides a computer-readable storage medium storing a program of the text processing method, the program is executed by a processor to perform the following steps:

[0190] Acquire a text to be processed, and determine the number of second texts in a first database that have the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is that the text has been processed;

[0191] In response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to the plurality of second texts having the same content as the text to be processed are the same, the processing result is used as the processing result of the text to be processed.

[0192] Corresponding to the text processing method provided in the embodiment of the present application, the embodiment of the present application further provides a computer program product, the program product comprising: a computer program, the computer program is stored in a readable storage medium, at least one processor of the electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device performs the following steps:

[0193] Acquire a text to be processed, and determine the number of second texts in a first database that have the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is that the text has been processed;

[0194] In response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to the plurality of second texts having the same content as the text to be processed are the same, the processing result is used as the processing result of the text to be processed.

[0195] It should be noted that for the detailed description of the apparatus, electronic device, computer-readable storage medium and computer program product provided in the embodiments of the present application, reference can be made to the relevant description of the method in the embodiments of the present application, which will not be repeated here.

[0196] Although the present application is disclosed as above in the form of a preferred embodiment, it is not intended to limit the present application. Any technical personnel in this field may make possible changes and modifications without departing from the spirit and scope of the present application. Therefore, the scope of protection of the present application shall be based on the scope defined by the claims of the present application.

[0197] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0198] The above-mentioned integrated module implemented in the form of a software function module can be stored in a computer-readable storage medium. The above-mentioned software function module is stored in a storage medium, including a number of instructions for enabling an electronic device (which can be a personal computer, a server, or a network device, etc.) or a processor (English: processor) to perform some steps of the method described in each embodiment of the present application.

[0199] It should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be implemented by a combination of hardware and software modules in the processor.

[0200] The memory may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk or an optical disk, etc.

[0201] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of the present application is not limited to only one bus or one type of bus.

[0202] The above storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The storage medium can be any available medium that can be accessed by a general or special purpose computer.

[0203] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0204] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A text processing method, characterized in that: The method comprises: Acquire a text to be processed, and determine the number of second texts in a first database that have the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is that the text has been processed; In response to the number being greater than or equal to a preset number threshold, and the processing results corresponding to the plurality of second texts having the same content as the text to be processed are the same, the processing result is used as the processing result of the text to be processed.

2. The text processing method according to claim 1, characterized in that: The method further comprises: In response to the number being less than a number threshold, inputting the to-be-processed text into a pre-trained classification model to obtain a label and label credibility for the to-be-processed text output by the pre-trained classification model; Based on the label of the text to be processed and the label credibility, a processing result of the text to be processed is obtained.

3. The text processing method according to claim 2, characterized in that: Based on the label of the text to be processed and the label credibility, obtaining the processing result of the text to be processed includes: In response to the label credibility being greater than or equal to a preset credibility threshold, taking the label output by the pre-trained classification model as a processing result of the text to be processed; In response to the label credibility being less than the preset credibility threshold, the text to be processed is assigned to manual processing, and the manual annotation label of the text to be processed is used as the processing result of the text to be processed.

4. The text processing method according to claim 1, characterized in that: There are multiple texts to be processed, and the method further includes: In response to the fact that the contents of a first text to be processed and a second text to be processed are the same, the processing result of the first text to be processed is used as the processing result of the second text to be processed, the first text to be processed is the processed text to be processed, and the first text to be processed and the second text to be processed are any two of the texts to be processed.

5. The text processing method according to claim 1, characterized in that: After the step of responding that the quantity is greater than or equal to the quantity threshold, the method further includes: In response to the processing status of at least one of the second texts having the same content as the text to be processed being unprocessed, the text to be processed is cached in a third database; the third database is used to store the text to be processed whose processing flow is unclear.

6. The text processing method according to claim 5, characterized in that: The method further comprises: In response to the processing results of the second text having the same content as the text to be processed being at least partially different, determining whether there is a third text having the same content as the text to be processed in the second database; wherein the second database is used to store text to be processed when the number of second texts having the same content as the text to be processed in the first database is greater than or equal to a preset number threshold, and when a plurality of the second texts are all processed but the processing results corresponding to the plurality of the second texts are at least partially different, the processing status corresponding to the text to be processed, and the processing result whose processing status is processed; In response to the absence of a third text having the same content as the text to be processed in the second database, assigning the text to be processed to manual processing, and using the manual annotation label of the text to be processed as the processing result of the text to be processed; In response to the presence of a third text identical to the text to be processed in the second database and the third text having been processed, the processing result of the third text is used as the processing result of the text to be processed.

7. The text processing method according to claim 6, characterized in that: In response to the absence of a third text identical to the text to be processed in the second database, after the step of assigning the text to be processed to manual processing and using the manual annotation label of the text to be processed as the processing result of the text to be processed, the method further includes: Updating the processing status of the text to be processed; The to-be-processed text, the processing status corresponding to the to-be-processed text, and the processing result of the to-be-processed text whose processing status is processed are stored in the second database.

8. The text processing method according to claim 6, characterized in that: After the step of responding to the presence of a third text identical to the text to be processed in the second database, the method further comprises: In response to the processing result of the third text being unprocessed, the to-be-processed text is cached in a third database.

9. The text processing method according to claim 8, characterized in that: After the step of caching the to-be-processed text in the third database, the method further comprises: In response to at least one of the to-be-processed texts being stored in the third database, acquiring a storage duration of the to-be-processed text; In response to the storage time being greater than or equal to a time threshold, the text to be processed is assigned to manual processing, and a manual label for the text to be processed is used as a processing result of the text to be processed; In response to the storage duration being less than the duration threshold, the step of responding to the number being greater than or equal to the number threshold is continued to obtain a processing result of the text to be processed.

10. The text processing method according to claim 1, characterized in that: In response to the number being greater than or equal to the number threshold and the processing results corresponding to the plurality of second texts having the same content as the text to be processed being the same, after the step of using the processing result as the processing result of the text to be processed, the method further includes: Updating the processing status of the text to be processed; The text to be processed, the processing status corresponding to the text to be processed, and the processing result with the processing status being processed are stored in the first database.

11. The text processing method according to claim 10, characterized in that: In response to the number being greater than or equal to the number threshold and the processing results corresponding to the plurality of second texts having the same content as the text to be processed being the same, after the step of using the processing result as the processing result of the text to be processed, the method further includes: Based on the processing result of the text to be processed, determining whether the user who generated the text to be processed is punished; In response to the user of the to-be-processed text being punished, the user who generated the to-be-processed text is punished.

12. The text processing method according to claim 1, characterized in that: After the step of obtaining the text to be processed, the method further includes: The text to be processed is preprocessed so as to convert the text to be processed into a preset text format.

13. A text processing device, characterized in that: The device comprises: A text acquisition module, used to acquire a text to be processed, and determine the number of second texts in a first database having the same content as the text to be processed, wherein the first database includes a plurality of second texts, a processing status corresponding to each second text, and a processing result corresponding to the second text whose processing status is processed; The processing module is used for responding to the fact that the number is greater than or equal to a preset number threshold and the processing results corresponding to the plurality of second texts having the same content as the text to be processed are the same, and taking the processing result as the processing result of the text to be processed.

14. An electronic device, characterized in that: include: processor; as well as The memory is used to store a data processing program. After the electronic device is powered on and the program is run by the processor, the method according to any one of claims 1 to 12 is executed.

15. A computer-readable storage medium, characterized in that: A data processing program is stored, and the program is run by a processor to execute the method according to any one of claims 1 to 12.

Images