Comment processing method and device, equipment and medium

Abstract

The embodiment of the invention relates to a comment processing method and device, equipment and a medium, and the method comprises the steps: responding to a content comment operation for a target document, and obtaining a target comment text corresponding to the content comment operation; obtaining target associated information corresponding to the target comment text in the target document; based on the target associated information, executing a rewriting operation on the target comment text to obtain a rewritten text corresponding to the target comment text; and obtaining a response text corresponding to the rewritten text, and displaying the response text based on the position of the target comment text. According to the embodiment of the invention, the response text with relatively high quality can be intelligently generated and displayed without waiting for other users to respond to the target comment text, so that the user initiating the comment can quickly know the corresponding response, and the user experience is effectively improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a comment processing method, apparatus, device, and medium. Background Art

[0002] In scenarios such as office work, multiple users may interact with the same document. For example, a user may raise a question about a sentence in a document through a comment and wait for other users to respond to the question through comments. The inventors have found that existing methods for interacting with document content through comments still need to be improved. Summary of the Invention

[0003] In order to solve the above technical problems or at least partially solve the above technical problems, the present disclosure provides a comment processing method, device, equipment and medium.

[0004] An embodiment of the present disclosure provides a comment processing method, which includes: in response to a content comment operation on a target document, obtaining a target comment text corresponding to the content comment operation; obtaining target association information corresponding to the target comment text in the target document; based on the target association information, performing a rewrite operation on the target comment text to obtain a rewritten text corresponding to the target comment text; obtaining a response text corresponding to the rewritten text, and displaying the response text based on the position of the target comment text.

[0005] Optionally, obtaining the target association information corresponding to the target comment text in the target document includes: performing semantic parsing on the target comment text, and identifying the target semantic type to which the target comment text belongs from a preset plurality of semantic types based on the semantic parsing result; and when the target semantic type belongs to a specified type among the plurality of semantic types, obtaining the target association information corresponding to the target comment text in the target document.

[0006] Optionally, the target association information includes one or more of the following: the title of the target document, overview information of the target document, other comment texts related to the target comment text, the commented content corresponding to the target comment text in the target document, and contextual information related to the commented content.

[0007] Optionally, the context information related to the commented content is obtained in the following manner: searching for a target node corresponding to the node to which the commented content belongs from the target document; wherein the target node includes one or more of a parent node, a sibling node, and a child node; and obtaining the context information related to the commented content based on the content corresponding to the target node.

[0008] Optionally, based on the target association information, a rewriting operation is performed on the target comment text to obtain a rewritten text corresponding to the target comment text, including: based on the target association information and the target comment text, generating target prompt information corresponding to a preset generation model; wherein the target prompt information is used to prompt the preset generation model to perform a rewriting operation on the target comment text based on the target association information, and the rewriting operation includes at least a sentence completion operation and / or a word interpretation operation; based on the target prompt information, a rewritten text corresponding to the target comment text is generated by the preset generation model.

[0009] Optionally, obtaining the response text corresponding to the rewritten text includes: scoring the rewritten text based on at least one preset target evaluation dimension to obtain the score corresponding to each target evaluation dimension; obtaining the target score of the rewritten text based on the score corresponding to each target evaluation dimension; and obtaining the response text corresponding to the rewritten text when the target score indicates that the rewritten text meets the response conditions.

[0010] Optionally, the at least one target evaluation dimension includes: a dimension for evaluating whether a sentence is complete, and / or a dimension for evaluating whether specific words exist in a sentence; wherein the specific words include pronouns that do not clearly refer to an object.

[0011] Optionally, the target document is an online collaborative text, and obtaining the response text corresponding to the rewritten text includes: obtaining the information acquisition permission of the target user corresponding to the target comment text; filtering out target data that meets the information acquisition permission from a preset database, and generating a response text corresponding to the rewritten text based on the target data.

[0012] Optionally, displaying the response text based on the position of the target comment text includes: identifying a target response category to which the response text belongs from a preset plurality of response categories based on the rewritten text and the response text; the plurality of response categories include: not answering questions raised by the rewritten text, providing information related to the rewritten text, and answering questions raised by the rewritten text; in a case where the target response category belongs to a specified category among the plurality of response categories, determining that the response text is a text that meets the display conditions, and displaying the response text based on the position of the target comment text.

[0013] The embodiment of the present disclosure also provides a comment processing device, including: a comment text acquisition module, used to respond to a content comment operation on a target document and obtain a target comment text corresponding to the content comment operation; an associated information acquisition module, used to obtain target associated information corresponding to the target comment text in the target document; a rewrite text acquisition module, used to perform a rewrite operation on the target comment text based on the target associated information, and obtain a rewrite text corresponding to the target comment text; a response text display module, used to obtain a response text corresponding to the rewrite text, and display the response text based on the position of the target comment text.

[0014] An embodiment of the present disclosure also provides an electronic device, which includes: a processor; a memory for storing instructions executable by the processor; the processor is used to read the executable instructions from the memory and execute the instructions to implement the comment processing method provided by the embodiment of the present disclosure.

[0015] The embodiment of the present disclosure further provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program is used to execute the comment processing method provided by the embodiment of the present disclosure.

[0016] The embodiments of the present disclosure further provide a computer program product, including a computer program, which, when executed by a processor, implements the comment processing method provided in the embodiments of the present disclosure.

[0017] The above technical solution provided by the embodiment of the present disclosure can respond to the content comment operation on the target document, obtain the corresponding target comment text, and additionally obtain the target association information corresponding to the target comment text in the target document, and further perform a rewrite operation on the target comment text based on the target association information, which helps to obtain a clearer and more specific rewritten text. On the basis of the rewritten text, it can not only effectively improve the success probability of obtaining the response text, but also help to improve the accuracy and reliability of the obtained response text. After that, the response text can be displayed based on the position of the target comment text. The above method does not need to wait for other users to respond to the target comment text, but can intelligently generate and display higher-quality response text, so that the user who initiated the comment can quickly know the corresponding reply, effectively improving the user experience.

[0018] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0020] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0021] Figure 1 A flowchart of a comment processing method provided in an embodiment of the present disclosure;

[0022] Figure 2 A flowchart of a comment processing method provided in an embodiment of the present disclosure;

[0023] Figure 3 A schematic diagram of a comment processing flow provided in an embodiment of the present disclosure;

[0024] Figure 4 A schematic diagram of the structure of a comment processing device provided in an embodiment of the present disclosure;

[0025] Figure 5 A schematic structural diagram of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0026] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features therein can be combined with each other in the absence of conflict.

[0027] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0028] Figure 1 This is a flow chart of a comment processing method provided by an embodiment of the present disclosure. The method can be executed by a comment processing device, wherein the device can be implemented using software and / or hardware and can generally be integrated into an electronic device. Figure 1 As shown, the method mainly includes the following steps S102 to S108:

[0029] Step S102 : In response to a content review operation on a target document, a target review text corresponding to the content review operation is obtained.

[0030] The embodiments of the present disclosure do not limit the format and content of the target document. In some implementation examples, the target document can be an online collaborative document that can be edited by different users, facilitating information interaction between different users based on the document. The target document can provide users with content comment functions, such as allowing users to comment on words and other content in the target document. For example, the user can select the content to be commented on in the target document, trigger the comment control to create a comment box corresponding to the content to be commented on, and enter the comment content in the comment box. The comment content is the target comment text.

[0031] Step S104: Obtain target association information corresponding to the target comment text in the target document.

[0032] Target association information is information related to the target comment text. It can also be understood that the user may determine the target comment text based on the target association information. Exemplarily, the target association information includes one or more of the following: the title of the target document, the overview information of the target document, other comment texts related to the target comment text, the commented content corresponding to the target comment text in the target document, and contextual information related to the commented content. Among them, the title of the target document can be obtained directly through title recognition, the overview information of the target document can be obtained through the abstract of the target document, or it can be obtained by providing the target document to a language model for document summarization, and other comment texts related to the target comment text may be comment texts related to the commented content corresponding to the target comment text. In other words, the commented content corresponding to other comment texts is related to the commented text corresponding to the target comment text, such as comment texts corresponding to multiple rounds of comments on the same commented content. The original text cited in the content comment operation in the target document is the commented content corresponding to the target comment text in the target document. In addition, contextual information related to the commented content can be identified based on specific strategies, such as information within a specific range before or after the commented content can be used as contextual information. The above is only an example, and the embodiments of the present disclosure do not limit the method of obtaining contextual information.

[0033] Step S106: Based on the target association information, a rewriting operation is performed on the target comment text to obtain a rewritten text corresponding to the target comment text.

[0034] It is understandable that since users comment on the content of the target document, such as entering the target comment text in the comment box by quoting the commented content, the individual target comment texts generally have problems such as missing subjects and unclear references. Most target comment texts are not complete in sentence structure and the content is not clearly expressed. In order to achieve accurate and reliable intelligent responses in the future, the disclosed embodiment will rewrite the target comment text in combination with the target-related information. The rewriting operation is at least used to supplement the complete sentence structure of the target comment text, or to clarify specific words in the target comment text for explanation, such as clarifying the specific referents of words such as pronouns, or explaining specific professional words or words that are prone to ambiguity, etc. Rewriting helps to convert the target comment text into a rewritten text with complete sentence structure and clear semantics.

[0035] Step S108: Obtain the response text corresponding to the rewritten text, and display the response text based on the position of the target comment text.

[0036] In actual applications, the response information required for rewriting the text can be found through database retrieval and other methods, and can be further organized into response text. The response text is displayed based on the position of the target comment text. The display position of the response text is related to the display position of the target comment text, such as displaying the response text below the target comment text. The response text can be regarded as the text obtained in response to the target comment text.

[0037] The above method provided by the embodiment of the present disclosure can respond to the content comment operation on the target document, obtain the corresponding target comment text, and additionally obtain the target association information corresponding to the target comment text in the target document, and further perform a rewrite operation on the target comment text based on the target association information, which helps to obtain a clearer and more specific rewritten text. On the basis of the rewritten text, it can not only effectively improve the success probability of obtaining the response text, but also help to improve the accuracy and reliability of the obtained response text. After that, the response text can be displayed based on the position of the target comment text. The above method does not need to wait for other users to respond to the target comment text, but can intelligently generate and display higher-quality response text, so that the user who initiated the comment can quickly know the corresponding reply, effectively improving the user experience.

[0038] In some implementation examples, the above step S104, i.e., the step of obtaining the target association information corresponding to the target comment text in the target document, can be performed with reference to the following steps (1) and (2):

[0039] Step (1) performs semantic parsing on the target review text, and based on the semantic parsing result, identifies the target semantic type to which the target review text belongs from a plurality of preset semantic types.

[0040] In actual applications, multiple semantic types can be pre-set, including but not limited to question semantics, statement semantics, etc. Question semantics can be further divided into multiple categories, such as question semantics involving knowledge / facts, such as "What does XXX mean?", search question semantics involving specific objects such as documents and groups, such as "Are there any videos or materials shared this time?", and question semantics for specific content or incomplete context, such as "Did you lower the capacity of something?" In addition, questions in non-work scenarios that require the personal wishes of other users to be sought, such as "Do you want to do XXX together?", etc. The disclosed embodiment can pre-set multiple possible semantic types and identify the target semantic type to which the target comment text belongs.

[0041] Step (2): when the target semantic type belongs to a specified type among multiple semantic types, obtaining target association information corresponding to the target comment text in the target document.

[0042] That is, the embodiment of the present disclosure can pre-determine whether the target semantic type is a specified type. Only when it is a specified type, it means that it is more suitable for the intelligent response scenario, and it will trigger subsequent operations such as target related information acquisition. If the target semantic type does not belong to the specified type, the subsequent process will no longer be triggered, and non-question comments that are not suitable for intelligent response can be effectively filtered out, thereby ensuring the reliability of intelligent response. For example, the disclosed embodiments can be divided into multiple categories based on the common expressions used by most users in document comments. For example, the first category is the aforementioned question semantics involving knowledge / facts; the second category is the aforementioned search question semantics involving specific objects such as documents and groups; the third category is the aforementioned question semantics targeting specific content or incomplete context; the fourth category is non-questions; the fifth category is questions in non-work scenarios, such as "Do you want to do XXX together?"; the sixth category is semantics such as requests / instructions / suggestions to users, such as "@someone, start applying"; the seventh category is semantics such as inquiries for opinions / decisions / suggestions / experience judgments / task division / question clarification, such as "@someone, see if anyone else needs to be invited?"; the eighth category is semantics for confirming the real-time status of specific objects such as users, services, code, and tasks, such as "Is anyone working on this issue?"; the ninth category is semantics containing rich text, such as "Which picture is accurate?"; the tenth category is semantics related to progress / planning / scheduling, such as "Ask if there are any next steps." The eleventh category is semantics for searching for specific users, such as "I want to ask who is the contact person for the current resource." The above are all examples. In actual applications, they can also be divided into other semantic types, which are not limited here. In some examples, the semantic type that can be clearly answered through retrieval is the specified type, such as the first three types of semantics mentioned above are specified types. Only when the target semantic type is a specified type that can be clearly answered through retrieval, the embodiment of the present disclosure will further trigger the acquisition of target-related information, so as to subsequently execute the process of intelligent answering, which helps to provide users with accurate and reliable answer text.

[0043] In some implementation examples, the target-related information includes: the title of the target document, the overview information of the target document, other comment texts related to the target comment text, the commented content corresponding to the target comment text in the target document, and the contextual information related to the commented content. The above five types of information can comprehensively guarantee the comprehensiveness of the target-related information. The aforementioned method of obtaining the title of the target document, the overview information of the target document, other comment texts related to the target comment text, and the commented content corresponding to the target comment text in the target document can refer to the aforementioned related content. This article aims to provide a more convenient and reliable method for obtaining contextual information related to the commented content. For example, the target node corresponding to the node to which the commented content belongs can be searched from the target document; wherein the target node includes one or more of a parent node, a sibling node, and a child node; and then the contextual information related to the commented content is obtained based on the content corresponding to the target node. The above-mentioned parent node and child node can be one-level or multi-level, and can be flexibly set, which is not limited here. In practical applications, the content type of the commented content in the target document can be obtained, and then a node search method can be determined based on the content type. For example, content types can be divided according to whether there is a hierarchical relationship within them. Content types with hierarchical relationships can be further divided into subtypes based on ordered lists, unordered lists, tables, columns, etc. In practical applications, node search methods corresponding to different content types can be pre-set, and based on the content type, one or more corresponding parent nodes, sibling nodes, and child nodes can be quickly searched, thereby efficiently finding the corresponding context information.

[0044] To improve rewriting efficiency and ensure rewriting reliability, in some implementation examples, step S106, i.e., performing a rewriting operation on the target comment text based on the target association information to obtain a rewritten text corresponding to the target comment text, can be performed with reference to steps a and b below:

[0045] Step a: Generate target prompt information corresponding to the preset generation model based on the target association information and the target comment text; wherein the target prompt information is used to prompt the preset generation model to perform a rewriting operation on the target comment text based on the target association information, and the rewriting operation at least includes a sentence completion operation and / or a word interpretation operation.

[0046] The disclosed embodiment can rewrite the target comment text with the help of the strong language processing capabilities of the generation model such as the large language model. In order to ensure the model rewriting effect, the disclosed embodiment can first generate target prompt information corresponding to the preset generation model based on the target association information and the target comment text, and the target prompt information can be a specified format that the model can understand. The target prompt information is used to prompt the generation model to rewrite the target comment text through operations such as sentence completion and word interpretation.

[0047] Step b: Based on the target prompt information, a rewritten text corresponding to the target comment text is generated through a preset generation model.

[0048] The target prompt information is input into the preset generation model, and the rewritten text output by the preset generation model is obtained, so that when the target comment text is incomplete and unclear, the original meaning of the comment can be restored with the help of the generation model. It is understandable that the target comment text and the corresponding target association information input in the document comment operation may be long texts, and there may be more noise information in the text. It is also easy to have complex hierarchical structures such as multi-level titles, ordered / unordered symbols, and table hierarchical structures. There may also be information of different content types such as tables and column contents. Moreover, since the target comment text corresponds to a quotation (that is, the commented content), it is easy to have problems such as missing subjects or a large number of pronouns. The embodiment of the present disclosure can first integrate the target association information and the target comment text into target prompt information that can be understood by the model, and then use the powerful information processing capabilities of the generation model to analyze and process the target prompt information, thereby rewriting the target comment text to obtain a rewritten text. Compared with other rewriting methods, the above method can effectively improve the rewriting efficiency and the reliability of the rewriting results.

[0049] To further ensure the reliability of the response, in some implementation examples, the step of obtaining the response text corresponding to the rewritten text in step S108 can be performed with reference to the following steps A to C:

[0050] Step A: scoring the rewritten text based on at least one preset target evaluation dimension to obtain scores corresponding to each target evaluation dimension.

[0051] The disclosed embodiments may set one or more target evaluation dimensions to measure whether the rewritten text is clear from different angles. In some specific implementation examples, at least one target evaluation dimension includes: a dimension for evaluating whether the sentence is complete, and / or a dimension for evaluating whether there are specific words in the sentence; wherein the specific words include pronouns that do not clearly refer to objects. Statements such as "What is this" that do not clearly refer to objects can be regarded as unclear statements. In actual applications, neural network models such as classification models can be used to score the rewritten text to obtain scores for each target evaluation dimension.

[0052] Step B, based on the scores corresponding to each target evaluation dimension, obtain the target score of the rewritten text. In practical applications, the scores corresponding to each target evaluation dimension can be integrated to obtain the target score of the rewritten text. The embodiment of the present disclosure does not limit the integration method. For example, the classification model can be directly used to integrate the scores corresponding to the rewritten text in multiple dimensions, and the probability value of whether the rewritten text is clear can be output. The probability value can be used as the target score. In some specific examples, the weights corresponding to each target evaluation dimension can be set according to needs. In addition, if there is only one target evaluation dimension, the weights corresponding to the other target evaluation dimensions can be regarded as zero, or the scores corresponding to the target evaluation dimension can be directly used as the target score of the rewritten text.

[0053] Step C: If the target score indicates that the rewritten text meets the response condition, obtain the response text corresponding to the rewritten text. For example, if the target score is higher than a preset score threshold, the target review text is considered to meet the response condition.

[0054] Through the above method, the corresponding response text can be obtained only for the rewritten text that meets the response conditions. For the rewritten text that does not meet the response conditions, no additional resources are wasted to obtain the response text. It can also effectively avoid the generation of unreliable response text due to the rewritten text not meeting the response conditions. The above method can also be understood as judging and intercepting unclear issues such as unclear reference, incomplete expression, or missing context that still exist after rewriting in the intelligent response process, so as to prevent comments that cannot be effectively answered from entering the subsequent response processing chain.

[0055] In some specific examples, the target document is an online collaborative text. The step of obtaining a response text corresponding to the rewritten text includes: obtaining the information access permissions of the target user corresponding to the target comment text; filtering target data that meets the information access permissions from a preset database, and generating a response text corresponding to the rewritten text based on the target data. The target document can be viewed and edited by multiple people. To ensure information security, the disclosed embodiments can first obtain the information access permissions of the target user. The target user can be the initiator of the target comment text or a user who can view the response results of the target comment text. The preset database can be, for example, an enterprise database. The disclosed embodiments can filter target data that meets the target user's information access permissions, retrieve the answer corresponding to the rewritten text from the target data, and generate a response text based on the retrieval results. For example, the retrieval results can be directly used as the response text, or the retrieval results can be further edited and adjusted to generate a more readable response text. For example, in the comment scenario, the response text can be made as concise as possible and conclusions can be pre-placed for closed questions that require clear answers such as "yes" or "no". This approach can provide responses based solely on the target user's available information, fully ensuring information security.

[0056] In some embodiments, the step of displaying the response text based on the position of the target comment text in step S108 can be performed with reference to the following steps 1 and 2:

[0057] Step 1, based on the rewritten text and the answer text, identify the target answer category to which the answer text belongs from the preset multiple answer categories; the multiple answer categories include: not answering the questions raised by the rewritten text, providing information related to the rewritten text, and answering the questions raised by the rewritten text. For example, when answering through the model, the following three situations may occur: the first situation is that the comment question is not answered at all, and the comment is just repeated, and the answer is brief. The second situation is that it contains text that refuses to answer the question directly, but provides some information related to the question topic, and typical sentence patterns are such as "no..., but...", "not found..., but...". The third situation is that the sentence pattern does not contain relevant text that refuses to answer, and the comment question is answered positively in content, and further relevant information can be provided. In addition, there may be other situations, which will not be described one by one here.

[0058] Step 2: When the target response category belongs to a specified category among multiple response categories, the response text is determined to be text that meets the display conditions, and the response text is displayed based on the position of the target comment text.

[0059] Exemplarily, the designated category may be a category for answering questions raised in response to the rewritten text. In this case, it is indicated that the response text meets the requirements and will be displayed accordingly. In actual applications, for ease of understanding, the above-mentioned response categories may correspond to corresponding scores, such as the first case corresponds to a score of 1, the second case corresponds to a score of 2, and the third case corresponds to a score of 3. Only a score of 3 is considered an acceptable answer. The above method can be understood as performing quality inspection on the response text generated by the model, and only displaying answers that have passed the quality inspection; when the target response category belongs to a designated category among multiple response categories, the response text can be regarded as having passed the quality inspection. In this way, the reliability of the intelligent response is further guaranteed, as well as the reliability of the answer ultimately provided to the user.

[0060] Based on the above, the present disclosure provides the following Figure 2 The flowchart of a comment processing method shown in FIG. 1 mainly includes the following steps S202 to S220:

[0061] Step S202 : In response to a content review operation on a target document, a target review text corresponding to the content review operation is obtained.

[0062] Step S204 , performing semantic parsing on the target comment text, and identifying the target semantic type to which the target comment text belongs from a plurality of preset semantic types based on the semantic parsing result.

[0063] Step S206: If the target semantic type belongs to a specified type among multiple semantic types, target association information corresponding to the target comment text in the target document is obtained. The target association information includes: the title of the target document, overview information of the target document, other comments related to the target comment text, the commented content corresponding to the target comment text in the target document, and contextual information related to the commented content.

[0064] Step S208, based on the target association information and the target comment text, generate target prompt information corresponding to the preset generation model; wherein the target prompt information is used to prompt the preset generation model to perform a rewriting operation on the target comment text based on the target association information, and the rewriting operation at least includes a sentence completion operation and / or a word interpretation operation.

[0065] Step S210: Based on the target prompt information, a rewritten text corresponding to the target comment text is generated through a preset generation model.

[0066] Step S212: Score the rewritten text based on at least one preset target evaluation dimension to obtain a score corresponding to each target evaluation dimension.

[0067] Step S214: Obtain a target score for the rewritten text based on the scores corresponding to each target evaluation dimension.

[0068] Step S216, when the target score indicates that the rewritten text meets the response conditions, obtain the information acquisition rights of the target user corresponding to the target comment text; filter out target data that meets the information acquisition rights from the preset database, and generate a response text corresponding to the rewritten text based on the target data.

[0069] Step S218, based on the rewritten text and the response text, identify the target response category to which the response text belongs from the preset multiple response categories; the multiple response categories include: not answering questions raised by the rewritten text, providing information related to the rewritten text, and answering questions raised by the rewritten text.

[0070] Step S220, when the target response category belongs to a specified category among multiple response categories, determines that the response text is a text that meets the display conditions, and displays the response text based on the position of the target comment text; the specified category includes a category that answers questions raised by the rewritten text.

[0071] The specific implementation of the above steps can refer to the above-mentioned relevant content and will not be repeated here. In the above method, after the user comments on the content of the document, there is no need to wait for other users to reply to the comments, and the target comment text can be automatically semantically parsed. When it is identified as belonging to the target semantic type, it is confirmed that subsequent intelligent responses can be made, and the subsequent intelligent response process will be executed. Specifically, target-related information such as context information will be obtained, and the model will be rewritten based on the target-related information and the target comment text. The target comment text will be rewritten with the help of the model's powerful information processing capabilities. After obtaining the rewritten text, it will be further scored. After confirming that the rewritten text meets the response conditions, the response text will be further generated based on the user information acquisition authority, and the response text can be further evaluated. It will only be displayed if it meets the display conditions. Through the above-mentioned layered processing method, the reliability of the response text finally displayed to the user is fully guaranteed.

[0072] For ease of understanding, Figure 2 Based on this, see also Figure 3 The schematic diagram of a comment processing process shown in the figure illustrates the process of intelligent answering. First, the operation of judging whether to trigger intelligent answering can be performed based on the comment text. If not, the process ends. If so, a rewriting operation is performed based on the target-related information such as the context passed in, and further an operation of judging whether the rewritten text has unclear problems is performed. If so, the process ends. If not, a response text is generated. After that, an operation of judging whether the response text has passed the quality inspection can be further performed. If not, the process ends. If so, the response text is displayed. Among them, the above-mentioned operation of judging whether to trigger intelligent answering can refer to the aforementioned steps S204 to S206, the above-mentioned rewriting operation can refer to the aforementioned steps S208 to S210, the above-mentioned operation of judging whether the rewritten text has unclear problems and the step of generating the response text can refer to the aforementioned steps S212 to S216, and the above-mentioned operation of judging whether the response text has passed the quality inspection and the step of displaying the response text can refer to the aforementioned steps S218 to S220. The specific implementation method and the achievable effect can be referred to the aforementioned related content, which will not be repeated here.

[0073] To sum up, the comment processing method provided by the embodiment of the present disclosure does not need to wait for other users to respond to the target comment text, but can intelligently generate and display higher-quality response text, ensuring the accuracy of the displayed response text, so that the user who initiated the comment can quickly obtain the corresponding reply, effectively improving the user experience, and also enhancing the interactivity between the user and the document.

[0074] Corresponding to the aforementioned comment processing method, the embodiment of the present disclosure further provides a comment processing device, Figure 4This is a structural diagram of a comment processing device provided by an embodiment of the present disclosure. The device can be implemented by software and / or hardware and can generally be integrated into an electronic device, such as Figure 4 As shown, the comment processing device includes:

[0075] A comment text acquisition module 402 is configured to acquire a target comment text corresponding to a content comment operation on a target document in response to the content comment operation on the target document;

[0076] The associated information acquisition module 404 is used to acquire the target associated information corresponding to the target comment text in the target document;

[0077] A rewritten text acquisition module 406 is used to perform a rewriting operation on the target comment text based on the target association information to obtain a rewritten text corresponding to the target comment text;

[0078] The response text display module 408 is used to obtain the response text corresponding to the rewritten text and display the response text based on the position of the target comment text.

[0079] The above-mentioned device provided by the embodiment of the present disclosure can respond to the content comment operation on the target document, obtain the corresponding target comment text, and additionally obtain the target association information corresponding to the target comment text in the target document, and further perform a rewrite operation on the target comment text based on the target association information, which helps to obtain a clearer and more specific rewritten text. On the basis of the rewritten text, it can not only effectively improve the success probability of obtaining the response text, but also help to improve the accuracy and reliability of the obtained response text. After that, the response text can be displayed based on the position of the target comment text. The above method does not need to wait for other users to respond to the target comment text, but can intelligently generate and display higher-quality response text, so that the user who initiated the comment can quickly know the corresponding reply, effectively improving the user experience.

[0080] In some embodiments, the associated information acquisition module 404 is specifically used to: perform semantic parsing on the target comment text, and based on the semantic parsing results, identify the target semantic type to which the comment text belongs from a preset plurality of semantic types; and when the target semantic type belongs to a specified type among the plurality of semantic types, obtain the target associated information corresponding to the target comment text in the target document.

[0081] In some embodiments, the target-related information includes one or more of the following: the title of the target document, overview information of the target document, other comment texts related to the target comment text, the commented content corresponding to the target comment text in the target document, and contextual information related to the commented content.

[0082] In some embodiments, the contextual information related to the commented content is obtained in the following manner: searching for a target node corresponding to the node to which the commented content belongs from the target document; wherein the target node includes one or more of a parent node, a sibling node, and a child node; and obtaining the contextual information related to the commented content based on the content corresponding to the target node.

[0083] In some embodiments, the rewritten text acquisition module 406 is specifically used to: generate target prompt information corresponding to a preset generation model based on the target association information and the target comment text; wherein the target prompt information is used to prompt the preset generation model to perform a rewriting operation on the target comment text based on the target association information, and the rewriting operation includes at least a sentence completion operation and / or a word interpretation operation; based on the target prompt information, generate a rewritten text corresponding to the target comment text through the preset generation model.

[0084] In some embodiments, the response text display module 408 is specifically used to: score the rewritten text based on at least one preset target evaluation dimension to obtain the score corresponding to each target evaluation dimension; obtain the target score of the rewritten text based on the score corresponding to each target evaluation dimension; and obtain the response text corresponding to the rewritten text when the target score indicates that the rewritten text meets the response conditions.

[0085] In some embodiments, the at least one target evaluation dimension includes: a dimension for evaluating whether a sentence is complete, and / or a dimension for evaluating whether a specific word exists in a sentence; wherein the specific word includes a pronoun that does not clearly refer to an object.

[0086] In some embodiments, the target document is an online collaborative text, and the response text display module 408 is specifically used to: obtain the information acquisition authority of the target user corresponding to the target comment text; filter out target data that meets the information acquisition authority from a preset database, and generate a response text corresponding to the rewritten text based on the target data.

[0087] In some embodiments, the response text display module 408 is specifically used to: based on the rewritten text and the response text, identify the target response category to which the response text belongs from a preset plurality of response categories; the plurality of response categories include: not answering the questions raised by the rewritten text, providing information related to the rewritten text, and answering the questions raised by the rewritten text; when the target response category belongs to a specified category among the plurality of response categories, determine that the response text is a text that meets the display conditions, and display the response text based on the position of the target comment text.

[0088] The comment processing device provided in the embodiments of the present disclosure can execute the comment processing method provided in any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0089] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described device embodiment can refer to the corresponding process in the method embodiment, and will not be repeated here.

[0090] An embodiment of the present disclosure provides an electronic device, which includes: a storage device storing a computer program; and a processing device configured to execute the computer program in the storage device to implement the steps of any one of the methods in the present disclosure.

[0091] Reference below Figure 5 , which shows a schematic structural diagram of an electronic device 500 suitable for implementing the embodiments of the present disclosure. The terminal devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Devices), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital televisions and desktop computers. Figure 5 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present disclosure.

[0092] like Figure 5 As shown, the electronic device 500 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage device 508 into a random access memory (RAM) 503. Various programs and data required for the operation of the electronic device 500 are also stored in the RAM 503. The processing device 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0093] Typically, the following devices may be connected to the I / O interface 505: an input device 506 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 507 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 508 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 509. The communication device 509 may allow the electronic device 500 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 5 The electronic device 500 is shown with various devices, but it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed instead.

[0094] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 509, or installed from the storage device 508, or installed from the ROM 502. When the computer program is executed by the processing device 501, the above-mentioned functions defined in the method of the embodiment of the present disclosure are performed.

[0095] In addition to the above-mentioned methods and devices, the embodiments of the present disclosure may also be a computer program product, which includes computer program instructions, which, when executed by a processor, cause the processor to perform the method provided by the embodiments of the present disclosure. The computer program product can be written in any combination of one or more programming languages to write program codes for performing the operations of the embodiments of the present disclosure, and the programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0096] In addition, the embodiment of the present disclosure may also be a computer-readable storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, the processor is caused to execute the comment processing method provided by the embodiment of the present disclosure.

[0097] The computer-readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can, for example, include but is not limited to a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical fiber, a portable compact disk read-only memory, an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0098] The embodiments of the present disclosure also provide a computer program product, including a computer program / instruction, which implements the comment processing method in the embodiments of the present disclosure when executed by a processor.

[0099] It is understandable that before using the technical solutions disclosed in the various embodiments of this disclosure, the type, scope of use, usage scenarios, etc. of the personal information involved in this disclosure should be informed to the user and the user's authorization should be obtained in an appropriate manner in accordance with relevant laws and regulations.

[0100] For example, in response to a user's active request, a prompt message is sent to the user to clearly inform the user that the operation requested will require the acquisition and use of the user's personal information. This allows the user to independently choose whether to provide personal information to the electronic device, application, server, storage medium, or other software or hardware that performs the operations of the disclosed technical solution based on the prompt message.

[0101] As an optional but non-limiting implementation, in response to receiving a user's active request, the prompt information may be sent to the user in the form of a pop-up window, in which the prompt information may be presented in text form. Furthermore, the pop-up window may also contain a selection control for the user to select "agree" or "disagree" to provide personal information to the electronic device.

[0102] It is understandable that the above notification and user authorization process are merely illustrative and do not limit the implementation of the present disclosure. Other methods that comply with relevant laws and regulations may also be applied to the implementation of the present disclosure.

[0103] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0104] The foregoing description is intended only to provide specific embodiments of the present disclosure, intended to enable those skilled in the art to understand and implement the present disclosure. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present disclosure. Therefore, the present disclosure is not intended to be limited to the embodiments described herein, but rather to be construed in the broadest manner consistent with the principles and novel features disclosed herein.

Claims

1. A comment processing method, characterized in that: include: In response to a content review operation on a target document, obtaining a target review text corresponding to the content review operation; Obtain target association information corresponding to the target comment text in the target document; Based on the target association information, a rewriting operation is performed on the target comment text to obtain a rewritten text corresponding to the target comment text; Obtain a response text corresponding to the rewritten text, and display the response text based on the position of the target comment text.

2. The method according to claim 1, characterized in that The step of obtaining target association information corresponding to the target comment text in the target document includes: Performing semantic parsing on the target review text, and identifying a target semantic type to which the target review text belongs from a plurality of preset semantic types based on the semantic parsing result; In a case where the target semantic type belongs to a specified type among the multiple semantic types, target association information corresponding to the target comment text in the target document is obtained.

3. The method according to claim 1 or 2, characterized in that The target association information includes one or more of the following: the title of the target document, overview information of the target document, other comment texts related to the target comment text, the commented content corresponding to the target comment text in the target document, and context information related to the commented content.

4. The method according to claim 3, characterized in that The context information related to the commented content is obtained in the following manner: Searching the target document for a target node corresponding to the node to which the commented content belongs; wherein the target node includes one or more of a parent node, a sibling node, and a child node; Context information related to the commented content is obtained based on the content corresponding to the target node.

5. The method according to claim 1, wherein The step of performing a rewriting operation on the target comment text based on the target association information to obtain a rewritten text corresponding to the target comment text includes: Based on the target association information and the target review text, generating target prompt information corresponding to a preset generation model; wherein the target prompt information is used to prompt the preset generation model to perform a rewriting operation on the target review text based on the target association information, and the rewriting operation includes at least a sentence completion operation and / or a word interpretation operation; Based on the target prompt information, a rewritten text corresponding to the target comment text is generated through the preset generation model.

6. The method according to claim 1, characterized in that The obtaining of a response text corresponding to the rewritten text includes: Scoring the rewritten text based on at least one preset target evaluation dimension to obtain scores corresponding to each target evaluation dimension; Obtaining a target score for the rewritten text based on the scores corresponding to the target evaluation dimensions; In a case where the target score indicates that the rewritten text meets the response condition, a response text corresponding to the rewritten text is obtained.

7. The method according to claim 6, characterized in that The at least one target evaluation dimension includes: a dimension for evaluating whether a sentence is complete, and / or a dimension for evaluating whether a specific word exists in a sentence; wherein the specific word includes a pronoun that does not clearly refer to an object.

8. The method according to claim 6, characterized in that The target document is an online collaborative text, and obtaining a response text corresponding to the rewritten text includes: Obtain information acquisition permissions of the target user corresponding to the target comment text; Target data that meets the information acquisition authority is screened out from a preset database, and a response text corresponding to the rewritten text is generated based on the target data.

9. The method according to claim 1, characterized in that The displaying of the response text based on the position of the target comment text includes: Based on the rewritten text and the response text, identifying a target response category to which the response text belongs from a plurality of preset response categories; the plurality of response categories include: not answering the question raised by the rewritten text, providing information related to the rewritten text, and answering the question raised by the rewritten text; In a case where the target response category belongs to a designated category among the multiple response categories, the response text is determined to be a text that meets the display conditions, and the response text is displayed based on the position of the target comment text.

10. A comment processing device, characterized in that: include: A comment text acquisition module, configured to, in response to a content comment operation on a target document, acquire a target comment text corresponding to the content comment operation; A related information acquisition module is used to obtain target related information corresponding to the target comment text in the target document; A rewritten text acquisition module, configured to perform a rewriting operation on the target comment text based on the target association information to obtain a rewritten text corresponding to the target comment text; The response text display module is used to obtain the response text corresponding to the rewritten text and display the response text based on the position of the target comment text.

11. An electronic device, characterized in that: The electronic device comprises: a storage device having a computer program stored thereon; A processing device is used to execute the computer program in the storage device to implement the steps of the comment processing method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that The storage medium stores a computer program, and the computer program is used to execute the comment processing method according to any one of claims 1 to 9.

13. A computer program product, characterized in that The invention comprises a computer program, which implements the comment processing method according to any one of claims 1 to 9 when executed by a processor.

Images