Information processing method, electronic equipment, storage medium and product

Abstract

The invention relates to an information processing method, electronic equipment, a storage medium and a product, and relates to the technical field of computers. The information processing method comprises the steps of generating target content based on a first instruction of a user; determining first body information according to one or more dimensions of information carried by a first instruction of a user; determining second figure information according to one or more features extracted from the target content; determining a typesetting template corresponding to the target content according to the first body information and the second body information; and outputting the typeset target content based on the typesetting template.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to an information processing method, electronic device, storage medium and product. Background Art

[0002] With the development of artificial intelligence technology, users can communicate with intelligent agents to obtain information efficiently. Intelligent agents can rely on machine learning models to process the information sent by users and reply to the information sent by users based on the processing results. Some machine learning models have powerful natural language processing capabilities, which can meet various user needs. For example, more and more users use intelligent agents to create content, which improves the efficiency of user creation. Summary of the invention

[0003] According to some embodiments of the present disclosure, an information processing method is provided, including: generating target content based on a first instruction of a user; determining first genre information based on information of one or more dimensions carried by the first instruction of the user; determining second genre information based on one or more features extracted from the target content; determining a typesetting template corresponding to the target content based on the first genre information and the second genre information; and outputting the typeset target content based on the typesetting template.

[0004] According to some embodiments of the present disclosure, an electronic device is provided, comprising: a memory; and a processor coupled to the memory, wherein the processor is configured to execute a method of any embodiment described in the present disclosure based on instructions stored in the memory.

[0005] According to some embodiments of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the processor executes the method of any embodiment described in the present disclosure.

[0006] According to some embodiments of the present disclosure, a computer program product is provided. When the computer program product is executed on a computer, the computer is enabled to implement the method of any embodiment described in the present disclosure.

[0007] Other features, aspects and advantages of the present disclosure will become apparent from the following detailed description of exemplary embodiments of the present disclosure with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] The following is an explanation of the embodiments of the present disclosure with reference to the accompanying drawings. It should be understood that the drawings described below only relate to some embodiments of the present disclosure and do not constitute a limitation to the present disclosure. In the accompanying drawings:

[0009] Figure 1 A flowchart of an information processing method according to some embodiments of the present disclosure is shown.

[0010] Figure 2 A schematic flow chart of a method for determining first genre information according to some embodiments of the present disclosure is shown.

[0011] Figure 3 A schematic flow chart of a method for determining second genre information according to some embodiments of the present disclosure is shown.

[0012] Figure 4 A schematic flow chart of a method for typesetting adjustment according to some embodiments of the present disclosure is shown.

[0013] Figure 5 A flowchart of an information processing method according to some other embodiments of the present disclosure is shown.

[0014] Figure 6 A schematic diagram of an interactive interface between a user and an agent according to some embodiments of the present disclosure is shown.

[0015] Figure 7 A schematic diagram of an interactive interface between a user and an agent according to some embodiments of the present disclosure is shown.

[0016] Figure 8 A schematic diagram of the structure of an information processing device according to some embodiments of the present disclosure is shown.

[0017] Fig. 9 A block diagram of an electronic device according to some embodiments of the present disclosure is shown.

[0018] Fig.10 A block diagram of an electronic device according to some other embodiments of the present disclosure is shown.

[0019] It should be understood that, for ease of description, the sizes of the various parts shown in the drawings are not necessarily drawn according to the actual proportional relationship. The same or similar reference numerals are used in the various drawings to represent the same or similar parts. Therefore, once an item is defined in one drawing, it may not be further discussed in subsequent drawings. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present disclosure will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present disclosure. It should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein.

[0021] It should be understood that the various steps described in the method embodiments of the present disclosure can be performed in different orders and / or performed in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect. Unless otherwise specifically stated, the relative arrangement and numerical values ​​of the components and steps described in these embodiments should be interpreted as being merely exemplary and not limiting the scope of the present disclosure.

[0022] The term “including” and its variations used in the present disclosure are intended to be open terms that include at least the following elements / features but do not exclude other elements / features, that is, “including but not limited to.” The term “based on” means “at least partly based on.”

[0023] It should be noted that the concepts of "first", "second", etc. mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units. Unless otherwise specified, the concepts of "first", "second", etc. are not intended to imply that the objects described in this way must be in a given order in time, space, ranking, or any other manner.

[0024] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0025] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0026] The user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0027] The embodiments of the present disclosure are described in detail below in conjunction with the accompanying drawings, but the present disclosure is not limited to these specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. In addition, in one or more embodiments, specific features, structures or characteristics can be combined in any suitable manner that will be clear from the present disclosure by a person of ordinary skill in the art.

[0028] In the process of interaction between the user and the agent, the agent usually replies to the user by sending a message in response to the user's instruction to instruct the agent to generate content. The message is usually presented in text form.

[0029] This method can help users obtain the created content itself. However, as user needs become more diverse and refined, this method also has some problems. At present, the content of the agent's reply is in a unified format. For example, whether the user instructs the agent to create a paper or a speech, the agent uses the same format (such as font, font size, layout, etc.) to display the creation results in the agent's reply message. In this way, it is difficult for users to preview the actual display effect of the created content, and they need to manually copy the generated content and then format it themselves. This increases the processing complexity for users.

[0030] The present disclosure provides an information processing method, which automatically determines a layout template based on a user's instructions for generating target content and the target content generated by an agent, and applies the layout template to the target content so that the user can efficiently preview the generated content.

[0031] Figure 1 FIG. 1 is a flow chart showing an information processing method according to some embodiments of the present disclosure. Figure 1 As shown, the information processing method of this embodiment includes steps S11 to S15. The method of this embodiment can be executed on the user device, or can be partially executed on the user device and partially executed on the server. The user device can run an application for interacting with the agent, and the application can execute at least some of the steps of this embodiment.

[0032] In step S11 , target content is generated based on a first instruction from a user.

[0033] The first instruction is used to instruct the generation of the target content. For example, the first instruction may be sent by a user to the agent, thereby triggering the generation process of the target content. For example, the first instruction may be processed using a machine learning model, and the machine learning model may perform semantic understanding on the first instruction and generate the target content corresponding to the first instruction.

[0034] In some embodiments, the machine learning model is a document generation model. The document generation model is used to create content according to the first instruction using a pre-trained language model, such as a model based on a Transformer architecture, combined with a large amount of text data. The model generates content according to the requirements of the first instruction. The generated content can be subjected to preliminary grammar and logic checks before subsequent processing.

[0035] The user may send the first instruction by means of a dialogue with the agent. For example, the first instruction may be a message sent by the user to the agent. Alternatively, the user may trigger the sending of the instruction through the skills or functions of the agent. For example, the user may trigger the "generate article" control in the dialogue interface and input some requirements for the content to be generated through the input interface of the skill or function. In response to the user's confirmation, the first instruction may be generated according to the content input by the user.

[0036] The generated target content includes text. In addition, it can also include other information such as images and music on the basis of the text.

[0037] After the target content is generated, it can be displayed first through the message sent by the intelligent agent, or it can be temporarily not displayed to the user.

[0038] In step S12, first genre information is determined according to information of one or more dimensions carried by the user's first instruction.

[0039] The first instruction includes a user's description or requirement of the target content to be generated, and the description or requirement may include one or more dimensions. The information of the one or more dimensions may include, for example, one or more of the following: theme, length, language style, genre, etc. The first genre information may be regarded as a genre determined based on the user's first instruction.

[0040] For example, one or more keywords can be extracted from the first instruction as information of one or more dimensions by keyword matching based on vocabulary or part of speech. For another example, one or more dimensions can be pre-specified, and then information corresponding to the dimensions can be extracted from the first instruction. For another example, the first instruction and processing prompts can be input into a large language model, and the model is instructed to extract information of one or more dimensions from it.

[0041] When determining the first genre information, one or more features may be determined based on the information of one or more dimensions, and the one or more features are used to characterize the first instruction, and the one or more features may correspond one to one with the one or more dimensions. Then, the representation including the one or more features may be processed using a classification model to obtain a classification result, and the category corresponding to the classification result corresponds to the first genre information.

[0042] When determining the first genre information, the large language model may also be used to process the information of the one or more dimensions to obtain the first genre information output by the large language model.

[0043] In step S13, second genre information is determined according to one or more features extracted from the target content.

[0044] One or more features extracted from the target content are used to reflect the characteristics of the target content from one or more perspectives. The one or more features may include keywords, language style, theme, structural features, etc. Structural features include chapter and paragraph information, etc. A natural language processing model may be used to extract one or more features from the target content, and then the extracted features may be processed using a classification model or a large language model to determine the second genre information.

[0045] Since the generation of the target content is also somewhat random and may contain some information that is not explicitly specified by the user, the target content may be embodied in genre information that is not completely the same as the first instruction. That is, the second genre information may not be completely the same as the first genre information.

[0046] In step S14, a layout template corresponding to the target content is determined according to the first genre information and the second genre information.

[0047] The first genre information is used to determine the genre from the user's perspective, and the second genre information is used to determine the genre from the perspective of the generated content. By considering both at the same time, it is possible to more accurately determine which template the generated content is suitable for. For example, the target genre information can be determined by the first genre information and the second genre information, and then the typesetting template corresponding to the target genre information is determined as the typesetting template of the target content.

[0048] The layout template is used to set the overall display style of the target content, or the overall and local display style. The display style includes font, font size, color, background, underline, bold, line spacing, reference format, etc. The local display style can be the display style of prominent objects such as titles and charts.

[0049] In step S15, the typeset target content is output based on the typeset template.

[0050] For example, the formatted target content may be displayed on a display interface, or a document in a specified format may be sent to the user, the document including the formatted target content.

[0051] The above-mentioned embodiments can respectively determine genre information based on the user's instructions for generating target content and the target content generated by the intelligent agent, and automatically determine the typesetting template for the target content based on the genre information determined from the above two perspectives, so that the user can efficiently preview the generated content, saving the user's additional typesetting operations and improving the user's content creation efficiency and user experience.

[0052] The first genre information, the second genre information, or the target genre information is used to indicate the genre, which may be a genre name or information represented by letters, numbers, etc. Any of the above-mentioned genre information may be divided by content type, such as poetry, prose, script, essay, and other literary genres; or, it may be divided by the type of the publishing platform of the content, such as blogs, online news publishing platforms, social networks, etc. Different platforms may have different typesetting requirements or typesetting conventions. The following is an exemplary description of the determination methods of these genre information.

[0053] Figure 2 FIG. 1 is a flow chart showing a method for determining first genre information according to some embodiments of the present disclosure. Figure 2 As shown, the determination method of this embodiment includes steps S121 to S123.

[0054] In step S121, semantic understanding is performed on the user's first instruction to determine one or more dimensions of information carried by the first instruction.

[0055] For example, a large language model may be used to perform semantic understanding on the first instruction, and the large language model may be instructed to output information of one or more dimensions carried by the first instruction.

[0056] In step S122 , for each dimension of the one or more dimensions, candidate genre information corresponding to the information of the dimension is determined.

[0057] Each dimension has multiple values, and each value can have one or more alternative genres. Taking the subject dimension as an example, it can include technology, entertainment, life, education, etc. It can be preset that technology corresponds to thesis genre and blog genre, and life corresponds to prose genre and social network content genre, etc. The alternative genre corresponding to each information of each dimension can be determined by pre-statistics or after processing by a large language model.

[0058] In step S123, first genre information is determined according to candidate genre information corresponding to information of each dimension.

[0059] There may be multiple candidate genres corresponding to the information of one or more dimensions of the first instruction, and the first genre information may be determined based on some or all of these candidate genres.

[0060] The user may have explicitly indicated the genre in the first instruction, that is, the dimensions of the first instruction include the “genre” dimension, or the user may not have explicitly indicated the genre.

[0061] For example, the first instruction sent by the user is "Use easy-to-understand language to explain the basic knowledge of computer graphics." Through "Explain the basic knowledge of computer graphics," it can be determined that the theme is "technology" and the language style is "popular." The alternative genres corresponding to the theme of "technology" include, for example, popular science articles and papers; the alternative genres corresponding to the language style of "popular" include, for example, blog articles, social network articles, essays, etc. Accordingly, the first genre information can be determined based on popular science articles, papers, blog articles, social network articles, and essays.

[0062] For example, the first instruction sent by the user is "write a blog to explain the basic knowledge of computer graphics in easy-to-understand language". The genre dimension information carried in the first instruction is "blog", and the information of other dimensions refers to the above examples. Since the user has clearly indicated the genre in the first instruction, it is possible to consider filtering out genres that are inconsistent with it and are determined by information in other dimensions, such as "social network articles" and "essays". However, "popular science articles" and "papers" can be considered to be retained to further refine the genres explicitly indicated by the user.

[0063] That is, an example of determining the first genre information is: in response to the user not indicating the genre through the first instruction, the first genre information is determined according to the union of the alternative genre information corresponding to the information of each dimension; or, in response to the user indicating the genre through the first instruction, the information in the alternative genre information that contradicts the genre indicated by the user is filtered out, and the first genre information is determined based on the genre indicated by the user and the remaining information in the alternative genre information. In this way, regardless of whether the user explicitly indicates the genre, the genre information that does not contradict the user's intention can be retained. Therefore, the first genre information can more accurately determine the genre applicable to the target content from the user's perspective.

[0064] Through the above method, each dimension of the first instruction can be used to determine the alternative genre respectively, and then the first genre information can be determined comprehensively based on multiple alternative genres. Therefore, the corresponding first genre information can be determined for any first instruction sent by the user, which improves the flexibility of the user in sending instructions.

[0065] Figure 3 FIG. 2 is a flow chart showing a method for determining second genre information according to some embodiments of the present disclosure. Figure 3 As shown, the determination method of this embodiment includes steps S131 to S133.

[0066] In step S131 , one or more features are extracted from the target content.

[0067] For example, the target content can be input into the large language model, and the large language model can be instructed to understand the semantics of the target content and output one or more features. The output features can be significant features in the target content understood by the model, or can be features of which dimensions the large language model is pre-specified to extract, so that the model outputs feature content corresponding to the pre-specified dimensions.

[0068] In step S132 , a representation of the target content is determined based on one or more features.

[0069] The representation of the target content can be expressed in natural language, in vector form, or in any other way. For example, one or more features can be directly combined into the representation of the target content, or one or more features can be processed (such as natural language to numerical conversion, feature screening, feature fusion, etc.), and then one or more elements obtained after the processing can be combined to obtain the representation of the target content.

[0070] In step S133, second genre information of the target content is determined using the representation of the target content.

[0071] In some embodiments, the representation of the target content is processed using a classification model, and the second genre information of the target content is determined based on the classification result. For example, the representation of the target content can be input into the classification model, and each classification result of the classification model corresponds to each genre. The classification model can be a single classification model or a multi-label classification model. That is, the classification result can include one or more categories, and each category corresponds to a genre.

[0072] The classification model is, for example, a convolutional neural network model or a recurrent neural network model. The model can be trained using documents annotated with genres so that the model learns the features of texts of different genres. The input of the model can be a vector representation of the target content. Through the feature extraction and classification layers of the multi-layer neural network, the probability distribution of the genre to which the target content belongs is output, and the genre with the highest probability is selected as the recognition result. In order to improve the recognition accuracy, an integrated learning method can be used to combine multiple models with different structures for comprehensive judgment.

[0073] In some embodiments, the representation of the target content is matched with one or more preset representations, and the genre information corresponding to the matched preset representation is determined as the second genre information of the target content. For example, the similarity or distance between the representation of the target content and the preset representation can be calculated, and a matching condition can be set, and the preset representation that meets the matching condition can be regarded as matching the representation of the target content. Each preset representation corresponds to one or more genre information, and the corresponding relationship can be preset. For example, a genre representation library can be pre-established, which includes the corresponding relationship between each preset representation and the genre information.

[0074] Of course, the above process of extracting features and then determining the second genre information can also be implicit. For example, the target content is input into the natural language processing model, and the model is directly instructed to extract one or more features from the target content, and then determine the second genre information based on these features. The extracted features may not be output directly, but can be used as a processing basis for further determining the second genre information within the model. That is, the model can be instructed on which information to use to determine the genre, which can make the model's processing more accurate.

[0075] The above method characterizes the target content by using the features extracted from the target content, and can describe the target content by using the features, thereby more accurately determining the genre information adapted to the target content from the perspective of the target content.

[0076] After the first genre information and the second genre information are determined, the target genre may be comprehensively determined.

[0077] In some embodiments, the target genre is determined according to the intersection of the first genre information and the second genre information; and the layout template corresponding to the target genre is determined as the layout template corresponding to the target content. The target genre determined according to the intersection of the first genre information and the second genre information can simultaneously reflect the user's intention and the generated target content, so the layout template determined based on the target genre can be more accurate.

[0078] Of course, those skilled in the art may also determine the target genre in other ways. For example, when determining the first genre information and the second genre information, the probability of each genre may be determined separately. Then, the target genre is determined based on the probability. For another example, the target genre includes at least one of the literary genre and the publishing platform genre. The literary genre and the publishing platform genre in the first genre information and the second genre information may be determined, the target literary genre may be determined from one or more literary genres, and the target publishing platform genre may be determined from one or more publishing platform genres. The determination method may refer to the above-mentioned intersection-based or probability-based method. Then, the combination of the target literary genre and the target publishing platform genre is determined as the target genre.

[0079] The above methods can automatically determine the layout template. Furthermore, it is also possible to adjust the display style in more detail for the elements in the target content. That is, for two contents that use the same layout template, the specific display details may be different due to their different contents. Figure 4 FIG. 1 is a flow chart showing a method for typesetting adjustment according to some embodiments of the present disclosure. Figure 4 As shown, the typesetting adjustment method of this embodiment includes steps S46 and S47. These steps can be performed after step S15.

[0080] In step S46, a target element is determined from the target content.

[0081] The target element may be an element of a specified data type or an element with specified semantics, that is, the target element may be determined from the target content according to the data type or semantics.

[0082] If the target content is generated with a style code after the typesetting template is applied to the target content, the target content can be determined based on the style tag. For example, <h1>The content identified by the tag indicates the title. <link> The content identified by the tag represents a hyperlink, etc.

[0083] In step S47, the display style of the target element is determined, wherein the target element is displayed in the display style in the laid out target content.

[0084] A correspondence between elements and display styles may be established in advance, so as to determine the display style of the target element according to the correspondence.

[0085] For example, in response to the target element being a term in a specified field, the display style of the target element is determined to be highlighted. Highlighting means that the content in the neighborhood of the target element is displayed in a different style, such as setting a different background color, using a different font size, adding an underline, using italics, etc., so that the key content in the target content can be highlighted.

[0086] For another example, in response to the target element being a specified data type, the display style pre-configured for the data type is determined as the display style of the target element. For example, for data, a specified number format can be set, and for a data list, a specified alignment can be set; for another example, for a chart, the alignment, size, annotation style, spacing from adjacent text paragraphs, etc. can be determined.

[0087] Determining the display style of the target element can be based on a combination of a rule engine and a machine learning model. For example, the rule engine is used to predefine typesetting rules for different genres and target elements, and the machine learning model is used to optimize and supplement the rules. An example is using a natural language processing tool to perform text analysis and extract target elements. For common target elements, the rule engine directly applies the corresponding rules; for complex or newly emerging elements, the machine learning model learns and predicts based on previous typesetting cases and user feedback to generate a suitable typesetting style. The personalized style rule library is stored using an extensible structure (such as XML or JSON format) for easy updating and management.

[0088] The target element is often an element that the user needs to adjust during the typesetting process. The above method can achieve a better typesetting effect while saving the user's refined typesetting operations by automatically determining the display style of the target element.

[0089] Sometimes, the user is not satisfied with the target content generated by the agent, and adjusts the generated content by sending a second instruction. After receiving the adjustment instruction, the layout template of the adjusted target content can be jointly determined according to the instruction and the information transmitted by the historical instructions. That is, the present disclosure supports multiple rounds of dialogue between the user and the agent for generating content, and determines the layout template based on the information of the multiple rounds of dialogue.

[0090] After each round of conversation, if the target content is generated (including the original content and the content adjusted based on the original content), the typesetting process can be automatically triggered; alternatively, the typesetting trigger can be executed in response to the user's trigger, that is, after the user instructs typesetting, the typesetting template is applied to the target content.

[0091] Figure 5 FIG. 2 is a flow chart showing an information processing method according to other embodiments of the present disclosure. Figure 5 As shown, the information processing method of this embodiment includes steps S51 to S54. In this embodiment, the target content has been generated according to the first instruction, and the first genre information and the second genre information have been determined, and then the second instruction sent by the user is received.

[0092] In step S51, a second instruction from a user is received, the second instruction being used to adjust the target content, such as expanding, abbreviating, rewriting the generated target content, or adjusting some details of the target content.

[0093] The second instruction can be a message sent by the user to the agent, such as a message sent through a dialogue interface, or can be triggered based on the user's use of the agent's skills or functions. In addition, if the content generated based on the first instruction is presented in the display area, and the display area supports the user's editing operation, the second instruction can be determined based on the user's editing operation.

[0094] In step S52 , based on the second instruction, the adjusted target content is generated.

[0095] In step S53, a layout template corresponding to the adjusted target content is determined according to the first genre information, the second genre information, the second instruction and the adjusted target content.

[0096] For example, the third genre information may be determined based on the second instruction, and the determination process is similar to determining the first genre information based on the first instruction. For another example, the fourth genre information may be determined based on the adjusted target content, and the determination process is similar to determining the second genre information based on the target content before adjustment.

[0097] The genre information determined based on the second instruction and the adjusted target content may be given a higher weight, and then the updated target genre information may be determined in combination with the first genre information and the second genre information. For example, a genre with a weighted probability higher than a threshold, or a genre with the highest probability, or a genre determined based on the intersection of the first to fourth genre information, etc. may be selected. Then, a layout template corresponding to the updated target genre information may be determined.

[0098] In step S54, the typeset, adjusted target content is output based on the typeset template corresponding to the adjusted target content.

[0099] Therefore, when determining the layout template of the adjusted target content, the historical dialogue between the user and the agent can be referred to, so that after the target content is adjusted, the layout template can also be updated if necessary. Therefore, the user can focus on the adjusted content and does not need to make extra operations to adjust the layout, which improves the user's operation efficiency.

[0100] The interactive interface between the user and the agent may include both a dialogue area and a preview area. For example, when the first instruction is a message sent by the user, the first instruction sent by the user to the agent may be displayed in the dialogue area of ​​the interface, wherein the interface also includes a preview area for displaying the target content after typeset.

[0101] Figure 6 FIG. 1 is a schematic diagram showing an interactive interface between a user and an agent according to some embodiments of the present disclosure. Figure 6 As shown, the interface 6 includes a dialogue area 61 and a preview area 62 .

[0102] In the dialogue area 61, a message 611 sent by the user "Help me write a paper on the application of artificial intelligence in the medical field" and a response message 612 of the intelligent agent "Generate a paper for you on the application of artificial intelligence in the medical field..." are displayed. The message 612 may include the body of the generated paper, or may include an introduction to the generated paper, or a download control. According to the message 611 and the generated content, it can be determined that the generated target content is a paper genre.

[0103] The dialogue area 61 may also include an input component 613, which may include one or more controls for the user to provide various types of input. If the user is not satisfied with the generated content, the user may also instruct the agent to adjust the target content through the input component 613.

[0104] The dialogue area 61 may also include an operation bar, including a trigger control 614 for the text polishing function of the agent, a trigger control 615 for the expansion function, etc. The user can quickly call the function of the agent by triggering these controls.

[0105] In the preview area 62 , the generated target content is displayed using the essay genre.

[0106] In this way, during the interaction with the agent, the user can preview the automatically typeset target content generated by the agent in real time, and the user can browse and adjust it efficiently.

[0107] The preview area 62 may provide an editing function. For example, the user may directly edit the generated content in the preview area 62, so that the user can make detailed modifications. Moreover, the user's operation may be perceived by the agent. That is, the dialogue area 61 and the preview area 62 may be bidirectionally linked.

[0108] In some embodiments, in response to detecting a first instruction for generating content, an operation bar is displayed in a dialog area, the operation bar includes an operation function, and the operation function is used to edit the target content; in response to a user triggering the operation function, the target content is edited based on the operation function; and in a preview area, the edited target content is displayed.

[0109] refer to Figure 6 In the interface 6 shown, before the first instruction is detected, the dialogue area 61 may not include the operation bar 614, or the operation bar displayed in the dialogue area 61 includes a default function (such as a music generation function, a picture generation function, etc.). After the first instruction is detected, the function in the operation bar is displayed or updated, so that the displayed operation bar includes functions related to the editing target content. In addition, the user's triggering of the operation function can affect the display result of the preview area, that is, the preview area can display the edited target content in real time. Thereby, the user's editing and preview efficiency is improved.

[0110] In addition to triggering the skills or functions of the agent, the user can also instruct the agent to edit the target content by sending a message to the agent. In some embodiments, the editing instructions sent by the user to the agent are displayed in the dialogue area, and the editing instructions include at least one of the content editing instructions and the typesetting editing instructions; in the preview area, the target content edited based on the editing instructions is displayed. This method can facilitate users to modify the target content more flexibly, and supports real-time preview of the modification results, which also improves the editing and preview efficiency of users.

[0111] For example, a natural language understanding (NLU) module and a command execution module can be constructed. The NLU module uses semantic parsing technology to convert user instructions into structured operation instructions. For example, the NLU module can use grammatical analysis trees, semantic role labeling and other technologies to understand the semantic structure of user instructions, extract operation objects (such as paragraphs, titles, pictures, etc.) and operation actions (such as setting fonts, adjusting line spacing, moving positions, etc.). The command execution module is used to adjust the layout of the target content according to the instructions. For example, based on the parsing results, the corresponding layout function is called to modify the style of the target content. In order to improve the accuracy and efficiency of instruction parsing, an instruction template library can be established to predefine and optimize the matching of common instructions.

[0112] When the genre of the target content's publishing platform can be determined, the simulated interface of the platform can be displayed during preview, and the typeset target content can be displayed in the simulated interface. For example, in response to the first genre information or the second genre information including the target content's publishing platform, the simulated interface corresponding to the platform is displayed in the preview area; based on the typeset template, the typeset target content is displayed in the simulated interface.

[0113] Figure 7 FIG. 1 is a schematic diagram showing an interactive interface between a user and an agent according to some embodiments of the present disclosure. Figure 7 As shown, the interface 7 includes a dialogue area 71 and a preview area 72 .

[0114] In the dialogue area 71, a message 711 sent by the user "Help me write a Beijing travel guide and post it on a certain APP" and a response message 712 of the intelligent agent "The following is a Beijing travel guide for you to be posted on the APP..." are displayed. According to messages 711 and 712, it can be determined that the generated target content is the genre of a certain APP.

[0115] In the preview area 72, the interface 720 of the APP is displayed, and the target content after typeset is displayed in the style of the article in the APP. Therefore, the user can obtain the publishing effect on the APP more intuitively. The user does not need to paste the generated content to the APP for testing, and then return to the interactive interface with the intelligent agent to edit and adjust. After obtaining the ideal effect in the current interface, go to the APP to publish. Therefore, the user's operating efficiency is improved.

[0116] The typesetting template and the target content can be two independent files, or the typesetting format determined based on the typesetting template can be embedded into the target content. For example, in the latter case, the typesetting template and the target content can be processed using a machine learning model to obtain the code output by the machine learning model, and the code is used to typeset the target content based on the typesetting template. That is, the input code includes the target content and the format indicator embedded in the target content. For example, the output can be HTML code, which includes both the target content and the format tags for controlling the display style. Of course, HTML is just an example, and the target content to which the template is applied can also be expressed in other languages. Thereby, the typesetting target content has stronger portability.

[0117] Alternatively, a dedicated code generation model can be trained. For example, a sequence-to-sequence (Seq2Seq) architecture can be used, such as a Transformer-based encoder-decoder model. The model input includes user instructions, document genre information, and encoded representations of text content, and the output is front-end programming language code (such as HTML and CSS code snippets). During the training process, a large amount of typesetting example data is used, including user instructions, document content, and corresponding front-end code, so that the model learns the mapping relationship between instructions and code.

[0118] The above describes the methods of various embodiments of the present disclosure. The following further describes the apparatuses for executing the above methods.

[0119] Figure 8 FIG. 2 shows a schematic diagram of the structure of an information processing device according to some embodiments of the present disclosure. Figure 8 As shown, the information processing device 8 of this embodiment includes: a generation module 81, configured to generate target content based on a first instruction of a user; a determination module 82, configured to determine first genre information according to information of one or more dimensions carried by the first instruction of the user, determine second genre information according to one or more features extracted from the target content, and determine a typesetting template corresponding to the target content according to the first genre information and the second genre information; and an output module 83, configured to output the typeset target content based on the typesetting template.

[0120] In some embodiments, the determination module 82 is further configured to: perform semantic understanding of the user's first instruction to determine one or more dimensions of information carried by the first instruction; for each of the one or more dimensions, determine the alternative genre information corresponding to the dimension information; and determine the first genre information based on the alternative genre information corresponding to the information of each dimension.

[0121] In some embodiments, the determination module 82 is further configured to: in response to the user not indicating a genre through the first instruction, determine the first genre information based on the union of the alternative genre information corresponding to the information of each dimension; or, in response to the user indicating a genre through the first instruction, filter out information in the alternative genre information that contradicts the genre indicated by the user, and determine the first genre information based on the genre indicated by the user and the remaining information in the alternative genre information.

[0122] In some embodiments, the determination module 82 is further configured to: extract one or more features from the target content; determine a representation of the target content based on the one or more features; and determine the second genre information of the target content using the representation of the target content.

[0123] In some embodiments, the determination module 82 is further configured to: process the representation of the target content using a classification model, and determine the second genre information of the target content based on the classification result; or, match the representation of the target content with one or more preset representations, and determine the genre information corresponding to the matched preset representations as the second genre information of the target content.

[0124] In some embodiments, the determination module 82 is further configured to: determine the target genre according to the intersection of the first genre information and the second genre information; and determine the typesetting template corresponding to the target genre as the typesetting template corresponding to the target content.

[0125] In some embodiments, the target genre includes at least one of a literary genre and a publishing platform genre.

[0126] In some embodiments, the determination module 82 is further configured to: determine a target element from the target content; and determine a display style of the target element, wherein the target element is displayed in the display style in the laid out target content.

[0127] In some embodiments, the determination module 82 is further configured to: in response to the target element being a term in a specified field, determine the display style of the target element as highlighting; or, in response to the target element being a specified data type, determine the display style pre-configured for the data type as the display style of the target element.

[0128] In some embodiments, the generation module 81 is further configured to: receive a second instruction from the user, the second instruction is used to adjust the target content, and based on the second instruction, generate the adjusted target content; the determination module 82 is further configured to determine the typesetting template corresponding to the adjusted target content based on the first genre information, the second genre information, the second instruction and the adjusted target content; based on the typesetting template corresponding to the adjusted target content, output the typeset, adjusted target content.

[0129] In some embodiments, the output module 83 is further configured to: utilize a machine learning model to process the typesetting template and the target content, and obtain a code output by the machine learning model, where the code is used to typeset the target content based on the typesetting template.

[0130] In some embodiments, the output module 83 is further configured to: display the first instruction sent by the user to the agent in the dialogue area of ​​the interface, wherein the interface also includes a preview area for displaying the target content after typeset.

[0131] In some embodiments, the output module 83 is further configured to: in response to detecting a first instruction for generating content, display an operation bar in the dialogue area, the operation bar including an operation function, and the operation function is used to edit the target content; in response to a user triggering the operation function, edit the target content based on the operation function; and in the preview area, display the edited target content.

[0132] In some embodiments, the output module 83 is further configured to: display in the dialogue area the editing instructions sent by the user to the agent, the editing instructions including at least one of content editing instructions and typesetting editing instructions; and display in the preview area the target content edited based on the editing instructions.

[0133] In some embodiments, the output module 83 is further configured to: in response to the first genre information or the second genre information including a publishing platform of the target content, display a simulation interface corresponding to the platform in the preview area; based on the layout template, display the layout target content in the simulation interface.

[0134] Fig. 9 A block diagram of an electronic device according to some embodiments of the present disclosure is shown.

[0135] The memory 91 is used to store one or more computer-readable instructions. The memory 91 may include any combination of various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory, including but not limited to random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), read-only memory (ROM), and flash memory. The memory 91 may store, for example, an operating system, an application, a boot loader (BootLoader), a database, and other programs, and may also store various applications and various data.

[0136] The processor 92 is used to run computer-readable instructions to implement the method described in any of the above embodiments. The specific implementation of each step of the method can be referred to the above embodiments, and the repeated parts are not repeated here.

[0137] The processor 92 may be configured to execute the steps in the methods of the embodiments of the present disclosure. The processor 92 may be embodied as various processing devices, such as a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The central processing unit (CPU) may be an X86 or ARM architecture, etc.

[0138] The processor 92 and the memory 91 may communicate with each other directly or indirectly. For example, the processor 92 and the memory 91 may communicate with each other via a network. The network may include a wireless network, a wired network, and / or any combination of a wireless network and a wired network. The processor 92 and the memory 91 may also communicate with each other via a system bus, which is not limited in the present disclosure.

[0139] It should be noted that Fig. 9 The components of the electronic device 9 shown are only exemplary and non-restrictive. The electronic device 9 may also have other components according to actual application requirements. The processor 92 may control other components in the electronic device 9 to perform desired functions.

[0140] The electronic device 9 may be implemented by software, firmware and / or hardware, and may be integrated into a device installed with relevant application programs.

[0141] Fig.10 A block diagram of an electronic device according to some other embodiments of the present disclosure is shown.

[0142] Fig.10 The electronic device 10 shown may be a computer system with a dedicated hardware structure, which can execute corresponding functions when a relevant application program is installed.

[0143] Electronic devices include, but are not limited to, mobile terminals such as smart phones, laptops, personal digital assistants (PDA), tablet personal computers (Tablet PC), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), wearable devices, etc., and fixed terminals such as digital televisions, desktop computers, etc.

[0144] like Fig.10 As shown, a central processing unit (CPU) 101 performs various processes according to a program stored in a read-only memory (ROM) 102 or a program loaded from a storage section 108 to a random access memory (RAM) 103. In the RAM 103, data required when the CPU 101 performs various processes, etc., is stored as needed. The central processing unit is merely exemplary, and it may also be other types of processors, such as the various processors described above. The ROM 102, the RAM 103, and the storage section 108 may be various forms of computer-readable storage media. It should be noted that although Fig.10 ROM 102, RAM 103 and storage section 108 are shown separately in FIG. 1 , but one or more of them may be combined or located in the same or different memory or storage modules.

[0145] The CPU 101, the ROM 102, and the RAM 103 are connected to each other via a bus 104. To the bus 104, an input / output interface 105 is also connected.

[0146] The following components are connected to the input / output interface 105: an input section 106, such as a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output section 107, including a display, such as a cathode ray tube (CRT), a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage section 108, including a hard disk, a magnetic tape, etc.; and a communication section 109, including a network interface card such as a LAN card, a modem, etc. The communication section 109 allows communication processing to be performed via a network such as the Internet. It is easy to understand that although Fig.10 Parts of the electronic device 10 are shown to communicate via the bus 104, but they may also communicate via a network or other means, wherein the network may include a wireless network, a wired network, and / or any combination of wireless networks and wired networks.

[0147] A drive 1010 is also connected to the input / output interface 105 as needed. A removable medium 1011 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is mounted on the drive 1010 as needed so that a computer program read therefrom is installed into the storage section 108 as needed.

[0148] When the above-described series of processing is realized by software, a program constituting the software can be installed from a network such as the Internet or a storage medium such as the removable medium 1011 .

[0149] 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, some embodiments of the present disclosure include a computer program product, which, when the computer program product is run on a computer, enables the computer to implement the method described in any of the aforementioned embodiments. The computer program product includes a computer instruction carried on a computer-readable medium, containing a program code for executing the method shown in the flowchart. In such an embodiment, the computer instruction can be downloaded and installed from the network through the communication part 109, or installed from the storage part 108, or installed from the ROM 102. When the computer program is executed by the CPU 101, the method of the embodiment of the present disclosure is executed.

[0150] It should be noted that, in the context of the present disclosure, a computer-readable medium may be a tangible medium that may contain or store a program for use by an instruction execution system, apparatus, or device or for use in conjunction with an instruction execution system, apparatus, or device.

[0151] The computer readable medium may be a computer readable storage medium, or a computer readable signal medium, or any combination of the two.

[0152] Computer-readable storage media include, but are not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or components, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to, electrical connections with one or more wires, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in conjunction with an instruction execution system, device, or device. Computer instructions are stored on a computer-readable storage medium, and when the instructions are executed by a processor, the method described in any of the foregoing embodiments is implemented.

[0153] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than a computer readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0154] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0155] In some embodiments, a computer program is further provided, comprising: instructions, which, when executed by a processor, cause the processor to execute the method described in any of the above embodiments. For example, the instructions may be embodied as computer program codes.

[0156] In embodiments of the present disclosure, computer program code for performing the operations of the present disclosure may be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on a user's computer, partially on a user's computer, as a separate software package, partially on a user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0157] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present disclosure. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0158] The functions described above may be performed at least in part by one or more hardware logic components. For example, exemplary hardware logic components that may be used include, without limitation, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0159] Although some specific embodiments of the present disclosure have been described in detail by way of example, it should be understood by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present disclosure. It should be understood by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.< / h1>

Claims

1. An information processing method, comprising: generating target content based on the user's first instruction; Determine first genre information according to the information of one or more dimensions carried by the first instruction of the user; Determining second genre information according to one or more features extracted from the target content; Determining a typesetting template corresponding to the target content according to the first genre information and the second genre information; Based on the layout template, the target content after layout is output.

2. The information processing method according to claim 1, wherein: The determining the first genre information according to the information of one or more dimensions carried by the first instruction of the user includes: Performing semantic understanding on the first instruction of the user to determine one or more dimensions of information carried by the first instruction; For each dimension of the one or more dimensions, determining candidate genre information corresponding to the information of the dimension; The first genre information is determined according to the candidate genre information corresponding to the information of each dimension.

3. The information processing method according to claim 2, wherein: The determining the first genre information according to the candidate genre information corresponding to the information of each dimension includes: In response to the user not indicating a genre through the first instruction, determining the first genre information according to a union of candidate genre information corresponding to the information of each dimension; or In response to the user indicating a genre through the first instruction, information in the candidate genre information that contradicts the genre indicated by the user is filtered out, and the first genre information is determined based on the genre indicated by the user and the remaining information in the candidate genre information.

4. The information processing method according to claim 1, wherein: Determining the second genre information according to one or more features extracted from the target content includes: Extracting the one or more features from the target content; Determining a representation of the target content according to the one or more features; Second genre information of the target content is determined using the representation of the target content.

5. The information processing method according to claim 4, wherein: The determining the second genre information of the target content by using the representation of the target content includes: Processing the representation of the target content using a classification model, and determining the second genre information of the target content according to the classification result; or The representation of the target content is matched with one or more preset representations, and the genre information corresponding to the matched preset representations is determined as the second genre information of the target content.

6. The information processing method according to claim 1, wherein: The determining, according to the first genre information and the second genre information, a typesetting template corresponding to the target content comprises: determining a target genre according to an intersection of the first genre information and the second genre information; The typesetting template corresponding to the target genre is determined as the typesetting template corresponding to the target content.

7. The information processing method according to claim 6, wherein: The target genre includes at least one of a literary genre and a publishing platform genre.

8. The information processing method according to any one of claims 1 to 7, further comprising: determining a target element from the target content; A display style of the target element is determined, wherein the target element is displayed in the display style in the target content after typeset.

9. The information processing method according to claim 8, wherein: Determining the display style of the target element includes: In response to the target element being a term in a specified field, determining that the display style of the target element is highlighted; or In response to the target element being a specified data type, a display style preconfigured for the data type is determined as the display style of the target element.

10. The information processing method according to any one of claims 1 to 7, further comprising: receiving a second instruction from the user, where the second instruction is used to adjust the target content; Based on the second instruction, generating adjusted target content; Determining a typesetting template corresponding to the adjusted target content according to the first genre information, the second genre information, the second instruction, and the adjusted target content; The adjusted target content is output after typeset based on a typeset template corresponding to the adjusted target content.

11. The information processing method according to any one of claims 1 to 7, wherein: The outputting of the target content after typeset based on the typeset template includes: The layout template and the target content are processed using a machine learning model to obtain a code output by the machine learning model, wherein the code is used to layout the target content based on the layout template.

12. The information processing method according to any one of claims 1 to 7, further comprising: In the dialogue area of ​​the interface, the first instruction sent by the user to the agent is displayed, wherein the interface also includes a preview area for displaying the target content after typeset.

13. The information processing method according to claim 12, further comprising: In response to detecting that the first instruction is used to generate content, displaying an operation bar in the dialog area, the operation bar including an operation function, and the operation function is used to edit the target content; In response to the user triggering the operation function, editing the target content based on the operation function; In the preview area, the edited target content is displayed.

14. The information processing method according to claim 12, further comprising: Displaying, in the dialogue area, an editing instruction sent by the user to the agent, wherein the editing instruction includes at least one of a content editing instruction and a layout editing instruction; In the preview area, the target content edited based on the editing instruction is displayed.

15. The information processing method according to claim 12, wherein: The outputting of the target content after typeset based on the typeset template includes: In response to the first genre information or the second genre information including a publishing platform of the target content, displaying a simulation interface corresponding to the platform in the preview area; Based on the layout template, the target content after layout is displayed in the simulation interface.

16. An electronic device, comprising: Memory; as well as A processor coupled to the memory, wherein the processor is configured to execute the information processing method according to any one of claims 1 to 15 based on instructions stored in the memory.

17. A computer-readable storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the information processing method according to any one of claims 1 to 15 is implemented. 18 . A computer program product, which, when executed on a computer, enables the computer to implement the information processing method according to claim 1 .

Images