Reading interaction method and apparatus, device, and medium

Abstract

A reading interaction method and apparatus, a device, and a medium. The method comprises: receiving a dialogue message sent by a second user device corresponding to a second user (S101); generating a reply message by a first user on the basis of the dialogue message, wherein the first user is a virtual author corresponding to an author (S102); and sending the reply message to the second user device (S103). Readers can interact with authors during book reading, so as to fulfill interaction needs therebetween, thereby improving the interactivity of readers during book reading and the book reading effect of the readers.

Description

Reading interaction method, device, equipment and medium

[0001] This application claims priority to the Chinese invention patent application entitled “Reading interaction method, device, equipment and medium” and application number 202410309882.3, filed on March 18, 2024. The entire contents of this application are incorporated by reference into this application. Technical Field

[0002] The embodiments of the present application relate to the field of computer technology, and in particular to a reading interaction method, apparatus, device, and medium. Background Art

[0003] Currently, more and more users like to read books through electronic devices (such as mobile phones), which allows users to use their spare time to read and study. However, during the reading process, users can only simply read the content of the book, which makes the reading process relatively boring and lacks interactivity, thereby reducing the user's reading effect.

[0004] Summary of the Invention

[0005] In a first aspect, an embodiment of the present application provides a reading interaction method, which is executed by a first user device, including: receiving a conversation message sent by a second user device corresponding to a second user; generating a reply message by the first user based on the conversation message, wherein the first user is a virtual author corresponding to the author; and sending the reply message to the second user device.

[0006] In the second aspect, an embodiment of the present application provides a reading interaction device, which is configured on a first user device, and includes: a receiving module for receiving a conversation message sent by a second user device corresponding to a second user; a generating module for generating a reply message based on the conversation message by the first user, where the first user is a virtual author corresponding to the author; and a sending module for sending the reply message to the second user device.

[0007] In a third aspect, an embodiment of the present application provides an electronic device comprising: a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the reading interaction method as described in the embodiment of the first aspect.

[0008] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium for storing a computer program, wherein the computer program enables a computer to execute the reading interaction method as described in the embodiment of the first aspect.

[0009] In a fifth aspect, an embodiment of the present application provides a computer program product comprising program instructions. When the program instructions are executed on an electronic device, the electronic device executes the reading interaction method as described in the embodiment of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0011] FIG1 is a schematic diagram of an application scenario of an embodiment of the present application;

[0012] FIG2 is a flow chart of a reading interaction method provided in an embodiment of the present application;

[0013] FIG3a is a schematic diagram of switching a current display interface to a dialogue interface according to an embodiment of the present application;

[0014] FIG3 b is a schematic diagram of another method of switching the current display interface to a dialogue interface according to an embodiment of the present application;

[0015] FIG4 is a schematic diagram of an interaction process between a first user and a second user provided in an embodiment of the present application;

[0016] FIG5 is a flowchart of generating a reply message provided by an embodiment of the present application;

[0017] FIG6 is a flowchart of creating a first user according to an embodiment of the present application;

[0018] FIG7 is a schematic block diagram of a reading interaction device provided in an embodiment of the present application; and

[0019] FIG8 is a schematic block diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments of this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or server 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.

[0022] In the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or solution described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or solutions. Specifically, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0023] In the description of the embodiments of the present application, unless otherwise specified, "a plurality of" refers to two or more than two, that is, at least two. "At least one" refers to one or more than one.

[0024] In order to solve the problem that readers can only read the content of books when reading books through electronic devices (such as mobile phones), which makes the book reading process boring, lacks interactivity, and reduces the readers' book reading effect, the invention concept of this application is to receive a conversation message sent by a second user device corresponding to a second user through a first user device corresponding to a first user, so that the first user generates a corresponding reply message according to the conversation message received by the first user device, and the first user sends the reply message to the second user device through the first user device to achieve interaction between the first user and the second user, so that the reader can interact with the author during the book reading process, thereby meeting the interaction needs between the reader and the author, improving the interactivity of the reader's book reading and improving the reader's book reading effect, and through the interaction between the reader and the author, the stickiness between the reader and the author can also be improved, thereby improving the reader's retention rate.

[0025] It should be understood that the technical solution of this application can be applied to but not limited to the following scenarios:

[0026] As shown in Figure 1, the application scenario may include: a first user device 110 and a second user device 120. The first user device 110 and the second user device 120 are connected to each other via a network. In this application, the network may be a wired network or a wireless network.

[0027] In some optional implementations, the first user device 110 is a terminal device used by an author, and the second user device 120 is a terminal device used by a reader.

[0028] The first user device 110 or the second user device 120 may be, but is not limited to, a smartphone, a tablet computer, a laptop computer, a personal computer, a smartwatch, or other devices.

[0029] Furthermore, various book reading clients may be installed in the first user device 110 and / or the second user device 120, where the book reading client should be understood as a book reading application software (Application, APP). In this application, the first user device 110 and the second user device 120 are both installed with the same book reading client, for example, the AA book reading client.

[0030] In the above application scenario, the second user can, at any time while using the book reading client, send a conversation message to the first user's corresponding first user device 110. The first user, based on the conversation message received by the first user device 110, generates a corresponding reply message and sends the reply message via the first user device 110 to the second user's corresponding second user device 120, thereby achieving interaction between the author and the reader. The interaction between the author and the reader may include, but is not limited to, discussing the subsequent plot of a book being updated, or exchanging information about a piece of data or characters in multiple books by the same author.

[0031] It should be noted that the first user equipment 110 and the second user equipment 120 shown in FIG1 are merely schematic. The number and device types of the first user equipment 110 and the second user equipment 120 may be adjusted according to actual usage and are not limited to those shown in FIG1 .

[0032] After introducing the application scenarios of the embodiments of the present application, the technical solution of the present application is described in detail below.

[0033] Figure 2 is a flow chart of a reading interaction method provided in an embodiment of the present application. The reading interaction method provided in an embodiment of the present application can be executed by a reading interaction device, which can be composed of hardware and / or software and can be integrated into the first user device shown in Figure 1.

[0034] As shown in FIG2 , the method may include the following steps:

[0035] S101: Receive a conversation message sent by a second user device corresponding to a second user.

[0036] S102: A first user generates a reply message according to the conversation message, where the first user is a virtual author corresponding to the author.

[0037] S103: Send a reply message to the second user equipment.

[0038] In this application, the first user is a virtual author corresponding to any book author in the book reading client. The virtual author can be a virtual user created by the author based on the user creation function provided by the book reading client or the first user device to meet personal interaction needs.

[0039] Furthermore, the virtual author may be selected as a model with semantic understanding capabilities.

[0040] Exemplarily, the model with semantic understanding capability may be selected as but not limited to a content generation module (AI-Generated Content, AIGC), a chat generative pre-trained transformer model (Chat Generative Pre-trained Transformer, ChatGPT), etc.

[0041] That is, this application allows the author to create a virtual author corresponding to himself, so that when the author interacts with the reader, he can interact with the reader through the virtual author, thereby reducing the time cost of interaction between the author and the reader, reserving more time for the author to create books, and also improving the stickiness between users and authors and improving user retention rate.

[0042] The second user mentioned above is a reader in the book reading client.

[0043] In some optional embodiments, when a second user wants to interact with the author of any book, such as a reader wanting to discuss a character in a book with the author, or to communicate about the subsequent plot of a book that is being updated, the second user can click on an interactive control provided by the book reading client to send a dialogue interface startup instruction to the book reading client or the second user device, so that the book reading client or the second user device switches the current display interface to the dialogue interface according to the received dialogue interface startup instruction, as shown in Figures 3a and 3b. The dialogue interface shown in Figures 3a and 3b includes: input controls, where the input controls may include but are not limited to an input box and an audio acquisition module.

[0044] In an embodiment of the present application, the above-mentioned current display interface may be a book reading interface or an author's personal information interface.

[0045] In addition, the above-mentioned interactive controls can be set in any area of ​​the book reading interface or the author's personal information interface, for example, they can be set in the lower left corner, upper left corner or lower right corner of the book reading interface, etc., as long as it does not affect the book content in the book reading interface and the display of personal information in the author's personal information interface. This application does not impose any restrictions on the setting location of the interactive controls.

[0046] That is, the second user sending a conversation message to the first user via the second user device can be sent at any time during the reading process of a book, such as after reading a book for a period of time, by clicking the interactive control provided in the reading interface to open the conversation interface and send the conversation message in the conversation interface; or, it can be sent when not reading a book (such as before or after reading a book), for example, after reading a book, the second user clicks the identification information of the author corresponding to the book to enter the author's personal information interface. Furthermore, by clicking the interactive control in the author's personal information interface, the conversation interface is opened and the conversation message is sent in the conversation interface, etc. This application does not impose any restrictions on this.

[0047] The author identification information includes at least the author's name information, wherein the author's name information may be a default setting of the book reading client or a custom setting of the author, and there is no limitation on it here.

[0048] Furthermore, after opening the conversation interface, the second user can input a conversation message through the touch input box, or collect the second user's audio data through the audio acquisition module, convert the second user's audio data into text information, and use the text information as a conversation message. When the second user's corresponding second user device or book reading client obtains a conversation message sent by the second user to any author, the second user device can send the conversation message to the corresponding first user device via the network, so that the first user corresponding to the first user device generates a reply message based on the conversation message sent by the second user. Then, the reply message is sent to the second user device through the first user device, thus realizing the interaction between the reader and the author.

[0049] In some optional embodiments, in this application, the first user may generate a reply message based on the conversation message sent by the second user in any of the following ways:

[0050] The first method is to consider that the first user is a model with semantic understanding capabilities. Therefore, when the first user in this application generates a reply message based on the conversation message sent by the second user, the conversation message can be input as input information into the model with semantic understanding capabilities, so that the model can automatically generate a reply message corresponding to the conversation message based on the conversation message.

[0051] The second approach, considering the large number of books and multiple authors in the book reading client, is to ensure that the first user can accurately identify which books each author has created and which books are associated with which authors. This application can create a target conversation database for each first user corresponding to each author. This target conversation database can be a general database and stores data such as which books each author has created and which books are associated with which authors in the book reading client. Furthermore, when the first user generates a reply message based on a conversation message sent by the second user, they can retrieve the corresponding reply message from the target conversation database using a partial search based on the conversation message.

[0052] As an optional implementation method, the present application can input the conversation message into the first user so that the first user can semantically understand the conversation message to determine the interaction intention of the second user, and then obtain the corresponding reply message in the target conversation database based on the interaction intention using a partial search method.

[0053] It should be noted that the above-mentioned implementation method of the first user generating a reply message based on the conversation message is only an example and does not constitute a specific limitation to this application.

[0054] For example, the interaction process between the first user and the second user in this application can be shown in Figure 4. In the example, the first user (i.e., the virtual author) and the second user have a conversation in the conversation interface, and the first user can reply to the conversation message sent by the first user.

[0055] The technical solution disclosed in the embodiment of the present application receives a conversation message sent by a second user device corresponding to a second user, generates a reply message based on the conversation message by the first user corresponding to the first user device, and sends the reply message generated by the first user to the second user device via the first user device. As a result, readers can interact with authors while reading books, thereby meeting the interaction needs between readers and authors, improving the interactivity of readers' book reading and improving readers' book reading results. In addition, the interaction between readers and authors can also improve the stickiness between readers and authors, thereby improving reader retention.

[0056] Based on the above embodiment, the following further explains the process of generating a reply message by a first user based on a conversation message sent by a second user in this application in conjunction with FIG5 . As shown in FIG5 , S102 in FIG2 may include the following steps:

[0057] S102-1, the first user determines a plurality of candidate reply messages in a target conversation database based on the conversation message.

[0058] In some optional embodiments, the first user performs semantic recognition and understanding on the conversation message sent by the second user to determine the second user's conversation intent. Furthermore, the first user retrieves at least one reply message related to the conversation intent from the target conversation database and identifies the at least one reply message as a candidate reply message.

[0059] For example, suppose the second user sends a conversation message that reads, "Assume characters A and B are in a relationship. This simulates an intimate conversation between two people, reflecting their respective personalities. Character A is gentle and shy, while character B is gloomy and arrogant. This conversation takes place just after character B has occupied location XX." After the first user performs semantic recognition and understanding on the conversation message sent by the second user, they can determine that the second user's conversational intent is to engage in a subsequent plot exchange with the first user. Based on this conversational intent, the first user then retrieves at least one reply message related to this subsequent plot exchange from the target conversation database as a candidate reply message.

[0060] S102-2, determining the matching degree between the conversation message and each candidate reply message through the first user.

[0061] It should be understood that determining the degree of match between the conversation message and each candidate reply message can be achieved by calculating the similarity between the conversation intent corresponding to the conversation message and each candidate reply message. A higher similarity indicates a closer match between the candidate reply message and the conversation message, and vice versa.

[0062] In some optional embodiments, calculating the similarity between the conversation intent corresponding to the conversation message and each candidate reply message can be achieved by utilizing a similarity calculation method, such as a cosine distance calculation method, a Euclidean distance calculation method, or other calculation methods. Specifically, the calculation process may include converting the conversation intent into a first vector and converting each candidate reply message into a second vector, and then utilizing the similarity calculation method to calculate the degree of match between the conversation message and each candidate reply message based on the first vector and each second vector.

[0063] Among them, converting the conversation intention into the first vector, converting each candidate reply message into the second vector, and calculating the similarity process are conventional technologies and will not be described in detail here.

[0064] S102-3, the first user determines a target reply message from all candidate reply messages based on the matching degree, and uses the target reply message as the reply message corresponding to the conversation message.

[0065] In some optional embodiments, the first user may sort the matching degree between the conversation message and each candidate reply message in descending order, obtain the maximum matching degree from the sorting results, and determine the candidate reply message corresponding to the maximum matching degree as the reply message corresponding to the conversation message.

[0066] Considering that there may be multiple maximum matching degrees, this application may use a random selection method to determine the candidate reply message corresponding to any maximum matching degree as the reply message corresponding to the conversation message, and there is no restriction on this here.

[0067] In some optional embodiments, the target conversation database is a pre-built database containing multiple candidate question and answer data based on data such as the content of each book, the corresponding author, and refined conversation data in the book reading client. Furthermore, the candidate question and answer data in the target conversation database can be updated based on historical conversation data between the first user and the second user to continuously expand and refine the candidate question and answer data in the target conversation database.

[0068] Furthermore, the construction of the target dialogue database in this application may include the following steps:

[0069] Step 1: Obtain identification information of multiple authors, identification information of books created by each author, and book content corresponding to each book identification information.

[0070] The term "multiple authors" may be understood as all authors in the book reading client. Furthermore, the author's identification information may be understood as information that can uniquely identify the author, such as the author's name.

[0071] Furthermore, the book identification information can be understood as information that can uniquely identify the book, such as the book name (ie, title).

[0072] In addition, the book content corresponding to each book identification information can be understood as the original text content of the book corresponding to the book identification information.

[0073] In some optional embodiments, considering that all book information in the book reading client and the author information of each book can be stored in the database of the book reading client, or in the database of the server corresponding to the book reading client, the present application can obtain the identification information of multiple authors, as well as the book identification information created by each author and the book content corresponding to each book identification information from the database of the book reading client or the server based on the identification information of the book reading client.

[0074] Step 2: Acquire multiple precisely labeled conversation data based on the identification information of each author, the identification information of the book created by each author, and the book content corresponding to each book identification information, wherein the precisely labeled conversation data is the labeled conversation data.

[0075] In some optional embodiments, conversation content extraction can first be performed on the acquired identification information of each author, the identification information of each author's books, and the book content corresponding to each book identification information to obtain multiple conversation data. Next, partial or all of the constructed conversation data can be annotated according to preset annotation rules through manual annotation or automatic annotation (such as semi-supervised automatic annotation), and the annotated conversation data can be understood as refined conversation data. The data annotation process can be referred to in the prior art and will not be described in detail here.

[0076] It should be noted that the dialogue data to be marked in this application can be flexibly determined according to actual needs, that is, different dialogue data to be marked can be set according to application requirements.

[0077] Step 3: construct a target conversation database based on the preset search platform, the book content corresponding to each book identification information, and multiple precise conversation data.

[0078] Among them, the preset search platform can be any search tool that can support search operations, such as a search engine or a search encyclopedia, and this application does not impose any restrictions on this.

[0079] In some optional embodiments, the conversation messages input by different second users in the preset search platform, the reply messages corresponding to the conversation messages, the book content corresponding to each book identification information, and multiple precise conversation data can be stored as question and answer candidate data to obtain a target conversation database.

[0080] In some optional implementation scenarios, considering that the book content corresponding to each book identification information is too long and contains a lot of content, it is possible to understand the content of each book and extract each book content into a question-and-answer format, that is, convert the book content into conversation data, so that the first user can quickly generate a reply message based on the conversation message of the second user.

[0081] Optionally, after step 2, the following steps may be performed:

[0082] Step 4: input the book content corresponding to each book identifier into the base model, and extract the content of each book through the base model to obtain a plurality of common conversation data.

[0083] In this application, the base model is a content understanding model, that is, as long as it can understand the content of the book and extract the book content into data in the form of dialogue, this application does not impose any specific restrictions on the content understanding model.

[0084] Correspondingly, step 3 is to construct a target conversation database according to the preset search platform, the book content corresponding to each book identification information, a plurality of precise conversation data and a plurality of general conversation data.

[0085] This application sets a target conversation database for the first user, so that the first user can determine candidate reply messages from multiple candidate question and answer data stored in the target conversation database based on the conversation message sent by the second user, and determine the target reply message from the candidate reply messages, and send the target reply message to the second user device corresponding to the second user, thereby ensuring that the reply message sent by the first user to the second user is more consistent and accurate with the conversation message, and can also improve the reply efficiency of the interaction between the author and the user.

[0086] The following is a detailed description of the process of creating a virtual author (i.e., the first user) corresponding to the author in the above embodiment with reference to FIG6. As shown in FIG6, the process of creating the first user may include the following steps:

[0087] S201, obtaining user attribute information input by the author.

[0088] In this application, the author is the creator of one or more books in the book reading client.

[0089] In some optional embodiments, the author clicks on a user creation function provided by the book reading client to enter a user creation interface. The author can then enter user attribute information in the user creation interface, so that the first user device creates the first user based on the user attribute information entered by the author.

[0090] The user attribute information may include character setting information, character personality information, character gender information, and reply tone information.

[0091] The above character setting information can be understood as character appearance, body proportions, service style, etc.

[0092] The above reply tone information can be understood as a specific tone or manner that should be followed during the conversation, such as using a friendly tone during the conversation.

[0093] S202 : Inputting user attribute information into the target large language model to encapsulate the target large language model into a virtual author corresponding to the author.

[0094] The target large language model is the large language model obtained after training. It should be noted that the large language model is also called a large language model (LLM), which can be used for content generation and dialogue interaction.

[0095] In some optional embodiments, the large language model can be trained by obtaining normal conversation data corresponding to each book content in the book reading client, as well as real conversation data from different real conversation scenarios. Furthermore, based on the normal conversation data and the real conversation data, the corresponding reply message for each conversation data is determined. The normal conversation data and the real conversation data are then input into the large language model to be trained, thereby training the large language model to obtain a target large language model. The training data also includes the reply message corresponding to each conversation data.

[0096] The large language model to be trained can be understood as a large language model whose parameters have not been optimized. In other words, it is a large language model in its initial state.

[0097] In addition, the above-mentioned acquisition of the common dialogue data of each book content dialogue can be referred to step 4 of the construction process of the aforementioned target dialogue database, which will not be described in detail here.

[0098] The real conversation data in the above-mentioned different real conversation scenarios can be obtained based on different types of multimedia content. The different types of multiple media content can be comics, videos, or audios.

[0099] In this application, based on different types of multimedia content, real conversation data in different real conversation scenarios are obtained. Optionally, different types of multimedia content can be understood and extracted through content understanding algorithms or models to obtain real conversation data in different real conversation scenarios.

[0100] In some optional embodiments, common conversation data and real conversation data are used as training data and input into a large language model to be trained. The large language model to be trained processes the common conversation data and real conversation data and outputs corresponding predicted reply messages. The predicted reply messages are then compared with the reply messages in the training data to determine whether the predicted reply messages output by the trained large language model are accurate. If, after comparison, the error between the predicted reply messages and the reply messages in the training data is determined to be greater than or equal to a preset value, the large language model is not trained. Based on the error between the predicted reply messages and the reply messages in the training data, the large language model is then reversely trained, and the error between the newly predicted reply messages output by the large language model and the reply messages in the training images is again determined. If the difference between the newly predicted reply messages and the reply messages in the training data is still greater than the preset value, the reverse training and result comparison operations are repeated until a stopping condition is met. If the error between the latest predicted reply message output by the large language model and the reply messages in the training data is less than the preset value, the large language model is trained, and the trained large language model is determined as the target large language model.

[0101] In some optional embodiments, determining the error between the predicted reply message and the reply message in the training data can be achieved by utilizing a backpropagation algorithm, a cross-entropy loss function, a gradient descent method, a cost function or an error function. As long as the error between the predicted reply message and the reply message in the training data can be determined, this application does not impose any restrictions on the specific determination method.

[0102] In some optional embodiments, the large language model is reversely trained based on the error between the predicted reply message and the reply message in the training data, specifically adjusting or optimizing the parameters in the large language model.

[0103] The above-mentioned stopping condition may be that the error between the predicted reply message and the reply message in the training data is less than a preset value, or the number of training times reaches a preset number.

[0104] In this application, the preset value and the preset number of times are both adjustable parameters. Optionally, if you want to obtain a reply message with higher accuracy, you can set the preset value to a smaller value or set the number of training times to a larger value. Conversely, you can set the preset value to a relatively larger value or set the number of training times to a relatively smaller value. These parameters can be flexibly adjusted based on the actual accuracy requirements of the reply message, and this application does not impose any restrictions on this.

[0105] After training the target large language model, the present application can input the acquired user attribute information into the target large language model to encapsulate the target large language model into a virtual author corresponding to the author, so that when the first user responds based on the dialogue message sent by the second user, a reply message can be sent to the second user in the reply tone set by the author. In this way, personalized interaction with readers can be achieved in the personalized reply tone set by the author, thereby increasing the frequency of interaction between readers and authors and further improving the stickiness between authors and readers.

[0106] The technical solution disclosed in the embodiment of the present application receives a conversation message sent by a second user device corresponding to a second user, generates a reply message based on the conversation message by the first user corresponding to the first user device, and sends the reply message generated by the first user to the second user device via the first user device. As a result, readers can interact with authors while reading books, thereby meeting the interaction needs between readers and authors, improving the interactivity of readers' book reading and improving readers' book reading results. In addition, the interaction between readers and authors can also improve the stickiness between readers and authors, thereby improving reader retention.

[0107] A reading interaction device according to an embodiment of the present application will be described below with reference to FIG7 . FIG7 is a schematic block diagram of a reading interaction device according to an embodiment of the present application. The reading interaction device in the present application is configured in the first user device of FIG1 .

[0108] As shown in FIG. 7 , the reading interaction device 500 includes a receiving module 510 , a generating module 520 and a sending module 530 .

[0109] The receiving module 510 is configured to receive a conversation message sent by a second user device corresponding to a second user;

[0110] A generating module 520 is configured to generate a reply message according to the conversation message by a first user, where the first user is a virtual author corresponding to the author;

[0111] The sending module 530 is configured to send the reply message to the second user equipment.

[0112] In an optional implementation method of an embodiment of the present application, the generation module 520 is specifically used to: determine a plurality of candidate reply messages in a target conversation database according to the conversation message by the first user; determine a matching degree between the conversation message and each of the candidate reply messages by the first user; determine a target reply message from all candidate reply messages according to the matching degree by the first user, and use the target reply message as the reply message corresponding to the conversation message.

[0113] In an optional implementation of the embodiment of the present application, the device 500 also includes: a construction module, which is used to obtain identification information of multiple authors, as well as book identification information created by each of the authors and book content corresponding to each of the book identification information; obtain multiple precise labeled conversation data based on the identification information of each of the authors, the identification information of the books created by each of the authors, and the book content corresponding to each of the book identification information, wherein the precise labeled conversation data is labeled conversation data; and construct the target conversation database based on a preset search platform, the book content corresponding to each of the book identification information, and the multiple precise labeled conversation data.

[0114] In an optional implementation of the embodiment of the present application, the apparatus 500 further includes: a processing module configured to input the book content corresponding to each of the book identifiers into a base model, so as to extract the content of each of the book contents using the base model to obtain a plurality of common conversation data;

[0115] Correspondingly, the construction module is specifically used to: construct the target conversation database according to the preset search platform, the book content corresponding to each book identification information, the multiple precise conversation data and the multiple common conversation data.

[0116] In an optional implementation of the embodiment of the present application, the base model is a content understanding model.

[0117] In an optional implementation of the embodiment of the present application, the acquisition module is further configured to acquire user attribute information input by the author;

[0118] The apparatus 500 further includes: a packaging module, configured to input the user attribute information into a target large language model, so as to package the target large language model into a virtual author corresponding to the author.

[0119] In an optional implementation of an embodiment of the present application, the user attribute information includes: character setting information, character personality information, character gender information, and reply tone information.

[0120] In an optional implementation of the embodiment of the present application, the acquisition module is further configured to acquire common conversation data corresponding to each book content, as well as real conversation data in different real conversation scenarios;

[0121] Correspondingly, the device 500 also includes: a model training module, which is used to input the ordinary conversation data and the real conversation data as training data into the large language model to be trained, train the large language model to be trained, and obtain a target large language model.

[0122] In an optional implementation of the embodiment of the present application, the real conversation data in different real conversation scenarios are obtained based on different types of multimedia content.

[0123] It should be understood that the device embodiment and the aforementioned method embodiment may correspond to each other, and similar descriptions can refer to the method embodiment. To avoid repetition, they are not described here. Specifically, the device 500 shown in FIG7 can execute the method embodiment corresponding to FIG2, and the aforementioned and other operations and / or functions of each module in the device 500 are respectively for implementing the corresponding processes in each method in FIG2. For the sake of brevity, they are not described here.

[0124] The above describes the device 500 of the embodiment of the present application from the perspective of functional modules in conjunction with the accompanying drawings. It should be understood that the functional module can be implemented in hardware form, can be implemented by instructions in software form, and can also be implemented by a combination of hardware and software modules. Specifically, the steps of the first aspect method embodiment in the embodiment of the present application can be completed by the hardware integrated logic circuit and / or software form instructions in the processor, and the steps of the first aspect method disclosed in conjunction with the embodiment of the present application can be directly embodied as being executed by a hardware decoding processor, or can be executed by a combination of hardware and software modules in the decoding processor. Optionally, the software module can be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory and completes the steps in the above-mentioned first aspect method embodiment in conjunction with its hardware.

[0125] Figure 8 is a schematic block diagram of an electronic device provided in accordance with an embodiment of the present application. As shown in Figure 8 , the electronic device 600 may include a memory 610 and a processor 620. The memory 610 is configured to store computer programs and transmit the program code to the processor 620. In other words, the processor 620 may retrieve and execute the computer program from the memory 610 to implement the reading interaction method provided in accordance with an embodiment of the present application.

[0126] For example, the processor 620 may be configured to execute the aforementioned reading interaction method according to instructions in the computer program.

[0127] In some embodiments of the present application, the processor 620 may include but is not limited to:

[0128] General-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc.

[0129] In some embodiments of the present application, the memory 610 includes but is not limited to:

[0130] Volatile memory and / or non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus random access memory (DR RAM).

[0131] In some embodiments of the present application, the computer program may be divided into one or more modules, which are stored in the memory 610 and executed by the processor 620 to implement the reading interaction method provided by the present application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device.

[0132] As shown in FIG. 8 , the electronic device 600 may further include a transceiver 630 , which may be connected to the processor 620 or the memory 610 .

[0133] The processor 620 may control the transceiver 630 to communicate with other devices. Specifically, the processor 620 may send information or data to other devices or receive information or data sent by other devices. The transceiver 630 may include a transmitter and a receiver. The transceiver 630 may further include one or more antennas.

[0134] It should be understood that the various components in the electronic device are connected via a bus system, wherein the bus system includes not only a data bus but also a power bus, a control bus and a status signal bus.

[0135] The present application also provides a computer storage medium on which a computer program is stored. When the computer program is executed by a computer, the computer is enabled to perform the reading interaction method of the above method embodiment.

[0136] An embodiment of the present application further provides a computer program product comprising program instructions, which, when executed on an electronic device, enables the electronic device to execute the reading interaction method of the above method embodiment.

[0137] When software is used for implementation, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a digital video disc (DVD)), or a semiconductor medium (e.g., a solid state drive (SSD)).

[0138] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are merely schematic. For example, the division of the modules is merely a logical function division. In actual implementation, there may be other division methods, such as multiple modules or components 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 through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0140] Modules described as separate components may or may not be physically separate, and components displayed as modules may or may not be physical modules, i.e., they may be located in one place or distributed across multiple network elements. Some or all of the modules may be selected based on actual needs to achieve the purpose of the present embodiment. For example, the functional modules in the various embodiments of the present application may be integrated into a processing module, or each module may exist physically separately, or two or more modules may be integrated into a single module.

[0141] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program that has a predetermined function and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0142] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A reading interaction method, performed by a first user device, comprising: receiving a conversation message sent by a second user device corresponding to the second user; Generate a reply message according to the conversation message by a first user, where the first user is a virtual author corresponding to the author; Send the reply message to the second user equipment.

2. The method according to claim 1, wherein generating a reply message according to the conversation message by the first user comprises: determining, by the first user according to the conversation message, a plurality of candidate reply messages in a target conversation database; determining, by the first user, a degree of matching between the conversation message and each of the candidate reply messages; The first user determines a target reply message from all candidate reply messages based on the matching degree, and uses the target reply message as the reply message corresponding to the conversation message.

3. The method according to claim 2, wherein the target dialogue database is constructed by the following steps: Obtain identification information of multiple authors, identification information of books created by each author, and book content corresponding to each book identification information; Acquire a plurality of precisely labeled conversation data according to the identification information of each author, the identification information of the book created by each author, and the book content corresponding to each book identification information, wherein the precisely labeled conversation data is the labeled conversation data; The target conversation database is constructed according to a preset search platform, the book content corresponding to each book identification information, and the plurality of precise conversation data.

4. The method according to claim 3, further comprising: Inputting the book content corresponding to each of the book identifiers into a base model, so as to extract the content of each of the book contents through the base model to obtain a plurality of common conversation data; Accordingly, the building of the target dialogue database includes: The target conversation database is constructed according to the preset search platform, the book content corresponding to each book identification information, a plurality of the precise conversation data and a plurality of the general conversation data. The method according to claim 4 , wherein the base model is a content understanding model.

6. The method according to claim 1, further comprising: Get the user attribute information entered by the author; The user attribute information is input into a target large language model to encapsulate the target large language model into a virtual author corresponding to the author.

7. The method according to claim 6, wherein the user attribute information comprises: Character setting information, character personality information, character gender information, and reply tone information.

8. The method according to claim 6, further comprising: Obtain common conversation data corresponding to each book content, as well as real conversation data in different real conversation scenarios; The normal conversation data and the real conversation data are used as training data and input into the large language model to be trained. The large language model to be trained is trained to obtain a target large language model.

9. The method according to claim 8, wherein the real conversation data in different real conversation scenarios are obtained based on different types of multimedia content.

10. A reading interaction device, configured on a first user device, comprising: A receiving module, configured to receive a conversation message sent by a second user device corresponding to a second user; A generating module, configured to generate a reply message according to the conversation message by a first user, where the first user is a virtual author corresponding to the author; A sending module is used to send the reply message to the second user equipment.

11. An electronic device comprising: A processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the reading interaction method according to any one of claims 1 to 9.

12. A computer-readable storage medium for storing a computer program, wherein the computer program enables a computer to execute the reading interaction method according to any one of claims 1 to 9.

Images