Presentation file generation method and device based on target big language generative model

Abstract

The embodiment of the invention discloses a presentation manuscript generation method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining a manuscript generation material input by a user, carrying out the processing of the manuscript content based on the manuscript generation material through employing a target large language generative model, and obtaining the reference manuscript content, and performing content filling on the presentation file template based on the reference file content to generate a target presentation file.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a presentation generation method, device, storage medium and electronic device based on a target large language generative model. Background Art

[0002] With the rapid development of computer technology, the production level of presentations is also improving rapidly. Presentations can clearly display data and information and facilitate communication. In daily life, when making presentations, users mainly manually add the collected materials to the specified presentation one by one, and manually complete the paging and typesetting. Summary of the invention

[0003] The present application provides a method, device, storage medium and electronic device for generating a presentation, and the technical solution is as follows:

[0004] In a first aspect, an embodiment of the present application provides a method for generating a presentation based on a target large language generative model, the method comprising:

[0005] Obtain the document generation material input by the user;

[0006] Based on the manuscript generation material, a target large language generative model is used to process the manuscript content to obtain reference manuscript content, and based on the reference manuscript content, the presentation template is filled with content to generate a target presentation.

[0007] Optionally, the step of generating the material based on the manuscript using a target large language generative model to process the manuscript content to obtain reference manuscript content, and filling the presentation template with content based on the reference manuscript content includes:

[0008] Determining that the manuscript generation material is a document material;

[0009] Based on the document material, a target large language generative model is used to extract the project name to obtain a demonstration project name;

[0010] The document content slice corresponding to the presentation project name is determined based on the document material, and the presentation template is filled with the slice content based on the document content slice.

[0011] Optionally, extracting the project name based on the document material using a target large language generative model to obtain the demonstration project name includes:

[0012] Generate a project name extraction prompt word for the document material, input the project name extraction prompt word and the document material into the target large language generation model to extract the project name, and obtain a demonstration project name.

[0013] Optionally, determining the document content slice corresponding to the demonstration project name from the document material includes:

[0014] Vectorizing the document material to obtain document vector image data; vectorizing the demonstration project name to obtain a demonstration project name vector;

[0015] The demonstration project name vector performs a vector index search process on the demonstration project name vector in the document vector image data to obtain the name position information corresponding to the demonstration project name;

[0016] The document material is processed by document content segmentation based on the name location information to obtain document content slices corresponding to the demonstration project name.

[0017] Optionally, the step of generating the material based on the manuscript using a target large language generative model to process the manuscript content to obtain reference manuscript content, and filling the presentation template with content based on the reference manuscript content includes:

[0018] Determining that the manuscript generation material is a reference presentation manuscript set, wherein the reference presentation manuscript set includes at least one reference presentation manuscript;

[0019] Based on the reference presentation manuscript set, a target large language generative model is used to perform comprehensive processing on the presentation title content to obtain a title presentation content corresponding to at least one presentation title;

[0020] The presentation template is filled with title content based on the title presentation content.

[0021] Optionally, the step of performing comprehensive processing of presentation title content using a target large language generative model based on the reference presentation manuscript set to obtain title presentation content corresponding to at least one presentation title includes:

[0022] Performing document page vectorization processing on the reference presentation document set to obtain a reference document page vector corresponding to at least one reference presentation document page;

[0023] Performing vector clustering processing on all reference manuscript page vectors to obtain at least one reference cluster center, and obtaining a reference demonstration title corresponding to the reference cluster center;

[0024] Generate a presentation title content extraction prompt word based on the reference presentation title and the reference presentation manuscript set;

[0025] The demonstration title content extraction prompt words, the reference demonstration title and the reference presentation manuscript set are input into the target large language generation model for comprehensive processing of the demonstration title content to obtain the title demonstration content corresponding to at least one demonstration title.

[0026] Optionally, the method further includes:

[0027] Acquire a presentation generation prompt word input by a user, and generate a presentation outline using a target language generation model based on the presentation generation prompt word to obtain a presentation outline;

[0028] The step of obtaining the manuscript input by the user to generate the material includes:

[0029] Receive the document generation material input by the user for the presentation outline.

[0030] Optionally, the method further includes:

[0031] If the manuscript generation material is a document material, a presentation project name is obtained, the presentation project name is generated based on the document material using a target language generation model, and the presentation outline is adjusted based on the presentation project name; and / or,

[0032] If the document generation material is a reference presentation set, at least one presentation title is obtained, the presentation title is generated based on the reference presentation set, and the presentation outline is adjusted based on the presentation title.

[0033] In a second aspect, an embodiment of the present application provides a presentation document generation device based on a target large language generative model, the device comprising:

[0034] The material acquisition module is used to acquire the manuscript generated by the user input;

[0035] The document processing module is used to process the document content based on the document generation material using a target large language generative model to obtain reference document content, and to fill the presentation template with content based on the reference document content to generate a target presentation.

[0036] Optionally, the document processing module includes:

[0037] A material determination unit, used to determine that the manuscript generation material is a document material;

[0038] A name extraction unit, used to extract the project name based on the document material using a target large language generative model to obtain a demonstration project name;

[0039] The content filling unit is used to determine the document content slice corresponding to the presentation project name based on the document material, and fill the presentation template with slice content based on the document content slice.

[0040] Optionally, the name extraction unit is used to:

[0041] Generate a project name extraction prompt word for the document material, input the project name extraction prompt word and the document material into the target large language generation model to extract the project name, and obtain a demonstration project name.

[0042] Optionally, determining the document content slice corresponding to the demonstration project name from the document material includes:

[0043] Vectorizing the document material to obtain document vector image data; vectorizing the demonstration project name to obtain a demonstration project name vector;

[0044] The demonstration project name vector performs a vector index search process on the demonstration project name vector in the document vector image data to obtain the name position information corresponding to the demonstration project name;

[0045] The document material is processed by document content segmentation based on the name location information to obtain document content slices corresponding to the demonstration project name.

[0046] Optionally, the document processing module is used to:

[0047] Determining that the manuscript generation material is a reference presentation manuscript set, wherein the reference presentation manuscript set includes at least one reference presentation manuscript;

[0048] Based on the reference presentation manuscript set, a target large language generative model is used to perform comprehensive processing on the presentation title content to obtain a title presentation content corresponding to at least one presentation title;

[0049] The presentation template is filled with title content based on the title presentation content.

[0050] Optionally, the document processing module is used to:

[0051] Performing document page vectorization processing on the reference presentation document set to obtain a reference document page vector corresponding to at least one reference presentation document page;

[0052] Performing vector clustering processing on all reference manuscript page vectors to obtain at least one reference cluster center, and obtaining a reference demonstration title corresponding to the reference cluster center;

[0053] Generate a presentation title content extraction prompt word based on the reference presentation title and the reference presentation manuscript set;

[0054] The demonstration title content extraction prompt words, the reference demonstration title and the reference presentation manuscript set are input into the target large language generation model for comprehensive processing of the demonstration title content to obtain the title demonstration content corresponding to at least one demonstration title.

[0055] Optionally, the device is also used for:

[0056] Acquire a presentation generation prompt word input by a user, and generate a presentation outline using a target language generation model based on the presentation generation prompt word to obtain a presentation outline;

[0057] The material acquisition module is used to:

[0058] Receive the document generation material input by the user for the presentation outline.

[0059] Optionally, the document processing module is used to:

[0060] If the manuscript generation material is a document material, a presentation project name is obtained, the presentation project name is generated based on the document material using a target language generation model, and the presentation outline is adjusted based on the presentation project name; and / or,

[0061] If the document generation material is a reference presentation set, at least one presentation title is obtained, the presentation title is generated based on the reference presentation set, and the presentation outline is adjusted based on the presentation title.

[0062] In a third aspect, an embodiment of the present application provides a computer storage medium, wherein the computer storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the above-mentioned method steps.

[0063] In a fourth aspect, an embodiment of the present application provides an electronic device, which may include: a processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the above-mentioned method steps.

[0064] The beneficial effects brought about by the technical solutions provided by some embodiments of the present application include at least:

[0065] In one or more embodiments of the present application, the electronic device obtains the document generation material input by the user, and then uses the target large language generative model to process the document content based on the document generation material to obtain the reference document content. Then, the presentation template can be filled with the content based on the reference document content to generate the target presentation. The whole process does not require a lot of time for manual production of the presentation. The efficiency of presentation processing is improved based on the target large language generative model, which saves presentation processing time and optimizes the presentation production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 paying creative work.

[0067] Figure 1 It is a flowchart of a presentation generation method provided in an embodiment of the present application;

[0068] Figure 2 It is a flowchart of a content filling process provided by an embodiment of the present application;

[0069] Figure 3 is a flowchart of another content filling process provided by an embodiment of the present application;

[0070] Figure 4 is a flowchart of another content filling process provided by an embodiment of the present application;

[0071] Figure 5 It is a structural schematic diagram of a presentation document generating device provided in an embodiment of the present application;

[0072] Figure 6 It is a structural schematic diagram of a document processing module provided in an embodiment of the present application;

[0073] Figure 7 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application;

[0074] Figure 8 It is a schematic diagram of the structure of the operating system and user space provided in the embodiment of the present application;

[0075] Fig. 9 yes Figure 8 The architecture diagram of the Android operating system;

[0076] Fig.10 yes Figure 8 Architecture diagram of the IOS operating system. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0078] In the description of the present application, it should be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In the description of the present application, it should be noted that, unless otherwise clearly specified and limited, "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood in specific circumstances. In addition, in the description of the present application, unless otherwise specified, "multiple" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone. The character " / " generally indicates that the associated objects before and after are an "or" relationship.

[0079] In the related art, manually making a presentation is a time-consuming and tedious task that requires a lot of time and effort. Especially when the presentation needs to be frequently updated and modified, this task becomes even more difficult and time-consuming.

[0080] The present application is described in detail below with reference to specific embodiments.

[0081] In one embodiment, Figure 1 As shown, a presentation generation method based on a target large language generative model is proposed, which can be implemented by a computer program and can be run on a presentation generation device based on the von Neumann system. The computer program can be integrated in an application or run as an independent tool application. The presentation generation device can be an electronic device, including but not limited to: a personal computer, a tablet computer, a handheld device, a vehicle-mounted device, a wearable device, a computing device or other processing device connected to a wireless modem. In different networks, terminal devices can be called different names, such as: user equipment, access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication device, user agent or user device, cellular phone, cordless phone, 5G network or electronic device in future evolution network, etc.

[0082] Specifically, the presentation generation method includes:

[0083] S102: Obtaining the manuscript generation material input by the user;

[0084] The presentation involved in this manual refers to a playable dynamic file made of text, pictures, audio, etc., with some special effects and dynamic display effects. The presentation is composed of many slides, which are independent and interrelated contents in the presentation. Usually, users can insert rich contents such as pictures, animations, tables, notes and handouts into the slides, making the complex contents of the presentation easy to understand and leaving a deeper impression.

[0085] Document generation materials can be understood as relevant filling content used to generate presentations, which can be pictures, audios, videos, tables, document files, etc. collected and organized by users. Document generation materials can be files collected by users in their daily work and life. The files can be saved in the local storage of the device after being collected by the user, or they can be saved in the cloud storage space of the device.

[0086] It should be noted that the format of the document generation material can be any format, such as picture format, pdf format, word format, rar format, zip format or folder format, etc.

[0087] Optionally, the electronic device may provide a human-computer interaction interface, which may be a target interface for automatically generating a presentation, wherein the target interface includes at least a presentation operation area and a presentation generation interaction area, wherein the presentation operation area is used to display display elements related to the presentation function, and the presentation generation interaction area may be used for a user to input presentation generation materials and instruct the target large language generation model to process the presentation content;

[0088] Exemplarily, a user may upload a document generation material required for automatically generating a presentation in the presentation generation interactive area, and then the electronic device executes the step of "processing the document content using the target large language generative model based on the document generation material to obtain reference document content, and filling the presentation template with content based on the reference document content to generate a target presentation";

[0089] Exemplarily, a user may upload a document generation material required for automatically generating a presentation in the presentation operation area, and then input a document content processing instruction in the presentation generation interactive area, and the electronic device executes the step of "processing the document content based on the document generation material using a target large language generative model to obtain reference document content, and filling the presentation template with content based on the reference document content to generate a target presentation"; or, the electronic device detects that the document generation material does not require the user to input a document content processing instruction, and the electronic device automatically executes the step of processing the document content based on the document generation material using a target large language generative model to obtain reference document content, and filling the presentation template with content based on the reference document content to generate a target presentation";

[0090] S104: Based on the manuscript generation material, a target large language generative model is used to process the manuscript content to obtain reference manuscript content, and based on the reference manuscript content, a presentation template is filled with content to generate a target presentation.

[0091] Optionally, a universal basic big language model can be directly obtained as a target big language generative model, and the model can be deployed for a document content processing scenario. Then, based on the document generation material, prompt words instructing the target big language generative model to process the document content are determined. Then, the prompt words and the document generation material are input into the target big language generative model to process the document content to obtain reference document content. Based on the reference document content, the presentation template is filled with content to generate a target presentation.

[0092] Optionally, in this specification, a target large language generative model is pre-trained, and the target large language generative model is obtained by adapting the basic large language model to the document content processing scenario, so that the basic large language model can be quickly applied to a new document content processing field without retraining a new model, and only fine-tuning the basic large language model is required;

[0093] The basic large language model may be, for example, a ChatGPT model, a Wenxinyiyan model, an LLM model, etc. The type of the basic large language model is not limited in this specification.

[0094] In a feasible implementation, a training method of a target large language generative model is illustrated as follows:

[0095] Model creation: First, the basic large language model is obtained, and then the initial target large language generative model is constructed. The initial target large language generative model can be composed of a scene adaptation module and a basic large language module based on the basic large language model.

[0096] Sample data acquisition: acquiring a large amount of sample data, the sample data is to construct a document content processing sample data set, the document content processing sample data set contains a large number of document content processing samples; the document content processing samples can be data of the document generation material type, and the document content processing samples can be presentation document generation prompt words (i.e. prompt description words that instruct the model to perform document content processing);

[0097] Sample data annotation: Based on the needs of document content processing scenarios, expert services are introduced to manually annotate document content processing samples with corresponding document content processing result labels.

[0098] Model training process: Input the document content processing samples into the initial target large language generative model for at least one round of model training to obtain the predicted document content processing results (equivalent to the reference document content in the actual application stage), and use the model loss function to determine the model loss value based on the predicted document content processing results and the document content processing result labels. Based on the model loss value, adjust the model parameters of the scene adaptation module to control the model parameters of the basic large language model unchanged until the model training end conditions are met to obtain the feature matching model.

[0099] Optionally, the model training end condition of the model may include, for example, the value of the loss function is less than or equal to a preset loss function threshold, the number of iterations reaches a preset number threshold, etc. The specific model training end condition can be determined based on actual conditions and is not specifically limited here.

[0100] It should be noted that the scene adaptation module involved in one or more embodiments of the present specification is created based on a machine learning model, and the machine learning model includes but is not limited to a convolutional neural network (CNN) model, a deep neural network (DNN) model, a recurrent neural network (RNN) model, an embedding model, a gradient boosting decision tree (GBDT) model, a logistic regression (LR) model, and other machine learning models. One or more fittings.

[0101] In a feasible implementation, the document generation material may be a prompt word in text form, that is, a prompt description word that instructs the target large language generation model to process the document content. For example, if a user expects to make a year-end performance review presentation, the presentation generation prompt word input by the user may be "assuming that you are a presentation production expert, please help me make a mid-year performance review presentation";

[0102] After the electronic device obtains the presentation generation prompt words input by the user, it uses the target language generative model to generate a presentation outline based on the presentation generation prompt words to obtain the presentation outline; then the user can further insert the presentation content based on the presentation outline, that is, the user can further input the presentation content required to be filled in the outline based on the presentation outline. After the electronic device obtains the presentation content, it generates the page content of the presentation page under the presentation outline based on the presentation content, and fills the page content into each presentation page of the presentation template. After completing the content filling, the target presentation is generated.

[0103] Among them, the presentation template can be a presentation template selected by the user or a pre-default presentation template. The presentation template can play a role in content layout by constraining the content filling format after the output content of the target large language generation model.

[0104] In one or more embodiments of the present application, the electronic device obtains the document generation material input by the user, and then uses the target large language generative model to process the document content based on the document generation material to obtain the reference document content. Then, the presentation template can be filled with the content based on the reference document content to generate the target presentation. The whole process does not require a lot of time for manual production of the presentation. The efficiency of presentation processing is improved based on the target large language generative model, which saves presentation processing time and optimizes the presentation production process.

[0105] See also Figure 2 , Figure 2 This is a flow chart of a content filling process proposed in this application. Specifically:

[0106] S202: Determine that the manuscript generation material is a document material;

[0107] Document material can be understood as the source data input by the user without being extracted and processed for presentation elements. In general, during the presentation production process, the content of the presentation needs to be further processed, organized, refined, typeset, and other production steps based on the document material in order to generate the final target presentation. The document material contains a large amount of text data of text type.

[0108] S204: extracting the project name based on the document material using a target large language generative model to obtain a demonstration project name;

[0109] In this manual, the user can only upload the document material, and then the electronic device will automatically generate a project name extraction prompt word for the document material. The project name extraction prompt word is a prompt used to indicate which kind of document processing the target large language generation model should perform on the document material. Based on the project name extraction prompt word, the target large language generation model can automatically extract the project name of the document material to obtain a demonstration project name.

[0110] Optionally, the user uploads the document material and triggers the project name extraction instruction, and then the electronic device automatically generates a project name extraction prompt word for the document material, and the project name extraction prompt word may be, for example, "assuming that you are a presentation production expert, please combine these document materials to summarize the project name in the document";

[0111] Exemplarily, after the electronic device obtains the document generation material input by the user, it determines that the document generation material is a document material, and then automatically generates a project name extraction prompt word for the document material, inputs the project name extraction prompt word and the document material into the target large language generation model to extract the project name, and obtains the demonstration project name, and the number of demonstration project names can be multiple.

[0112] S206: Determine the document content slice corresponding to the presentation project name based on the document material, and fill the presentation template with the slice content based on the document content slice.

[0113] Specifically, after the electronic device uses the target large language generative model to extract project names based on the document material to obtain several demonstration project names, the electronic device then locates the document description position in the document material based on these demonstration project names, and then obtains the document content related to the demonstration project name at the document description position, and then divides the "document content related to the demonstration project name" to obtain document content slices corresponding to each demonstration project name, and then fills the document content slices into the presentation template. After all document content filling is completed (the user may input a confirmation content filling completion instruction, and the electronic device determines that all document content filling is completed in response to the content filling completion instruction), the final target presentation is obtained. If there is a next content filling process, S202 is executed.

[0114] In a feasible implementation, the document content slice corresponding to the demonstration project name may be determined based on the document material in a vectorized manner and a vector graph indexing method. For details, refer to the following method:

[0115] A2: performing vector processing on the document material to obtain document vector image data, and performing vector processing on the demonstration project name to obtain a demonstration project name vector;

[0116] Schematically, the document material is usually composed of more content. In order to quickly locate, the present specification can use the graph index method to construct a graph index for the document material. The constructed graph index is the document vector graph data. Then the graph index method performs index search processing in the document vector graph based on the demonstration project name, so as to determine the name location information corresponding to the demonstration project name;

[0117] For example, the graph indexing method may be a Hierarchical Navigable Small World (HNSW) method, in which the document material is vectorized using the HNSW method to obtain document vector graph data, which is a vector space data of an HNSW graph structure. In the graph construction stage, randomness is introduced by randomly inserting vector nodes to construct a vector small world graph as a graph index, thereby facilitating subsequent rapid retrieval.

[0118] In the graph retrieval stage, feature engineering can be used to vectorize the demonstration project names to obtain demonstration project name vectors;

[0119] A4: performing a vector index search process on the demonstration project name vector in the document vector graph data based on the demonstration project name vector to obtain name position information corresponding to the demonstration project name;

[0120] A6: Perform document content segmentation processing on the document material based on the name location information to obtain document content slices corresponding to the demonstration project name.

[0121] Specifically, after obtaining the demonstration project name vector, the demonstration project name vector is input into the document vector graph data, and a graph indexing method (such as the HNSW method) is used to perform index search processing based on the demonstration project name in the document vector graph, so that the name position information corresponding to the demonstration project name can be determined. The name position information is the k nearest neighbor nodes of the demonstration project name vector determined based on the graph indexing method (such as the HNSW method), and the k nearest neighbor nodes (node ​​vectors corresponding to certain document contents associated with the demonstration project name) are the name position information corresponding to the demonstration project name, and then based on these name position information, the document content position related to the demonstration project name can be reversely indexed; then, based on the document content position related to the demonstration project name, the document material is subjected to document content segmentation processing, and the document content slice corresponding to each demonstration project name can be obtained.

[0122] S208: After completing the content filling, generate the target presentation;

[0123] It can be understood that the document content slices corresponding to each demonstration project name are obtained, and then the document content slices are filled into the presentation template. After all the document content is filled (the user may input a confirmation content filling completion instruction, and the electronic device determines that all the document content is filled in response to the content filling completion instruction), the final target presentation is obtained. If there is a next content filling process, S202 is executed.

[0124] In one or more embodiments of the present specification, the process of automatic presentation generation is shown, which can greatly reduce the time spent by users in making presentations and improve overall efficiency. The entire process does not require a large amount of time spent on manual presentation production. When the presentation generation material is a document material, the efficiency of presentation processing is improved based on the target large language generative model, which saves presentation processing time and optimizes the presentation production process.

[0125] See also Figure 3 , Figure 3 This is a flow chart of a content filling process proposed in this application. Specifically:

[0126] S302: Determine that the document generation material is a reference presentation document set, wherein the reference presentation document set includes at least one reference presentation document;

[0127] The reference presentation collection includes multiple reference presentations, such as multiple PPTs.

[0128] It can be understood that the multiple reference presentations in the reference presentation collection are materials for the user to automatically generate a new target presentation. For example, the user expects to automatically integrate the multiple reference presentations in the reference presentation collection to generate a new target presentation for summarizing the characteristics of the multiple reference presentations and referring to the multiple reference presentations.

[0129] For example, a plurality of reference presentations in the reference presentation collection may be historical presentations;

[0130] For example, the multiple reference presentations in the reference presentation set may be weekly presentations, and the target presentation may be a monthly presentation;

[0131] S304: Based on the reference presentation manuscript set, a target large language generative model is used to perform comprehensive processing on presentation title content to obtain title presentation content corresponding to at least one presentation title;

[0132] In a feasible implementation, the user uploads a reference presentation collection, and the electronic device vectorizes each reference presentation page of each reference presentation, obtains the central title of each class by vector clustering, and then uses the target large language generation model to perform comprehensive processing on the presentation title content to obtain the title presentation content corresponding to at least one presentation title. Finally, the title presentation content is inserted into the presentation template to generate a new target presentation.

[0133] B2: performing document page vectorization processing on the reference presentation document set to obtain a reference document page vector corresponding to at least one reference presentation document page;

[0134] Schematically, feature vector engineering is used to vectorize each reference presentation page of each reference presentation in the reference presentation set to obtain a reference page vector corresponding to each reference presentation page, and all reference page vectors are mapped to the same high-dimensional vector space.

[0135] B4: performing vector clustering processing on all reference manuscript page vectors to obtain at least one reference cluster center, and obtaining a reference demonstration title corresponding to the reference cluster center;

[0136] Indicatively, vector clustering processing is then performed on all reference manuscript page vectors to obtain a number of reference cluster centers, and reference demonstration titles corresponding to the reference cluster centers are obtained;

[0137] For example, the reference document page vectors corresponding to all reference document pages are clustered using the k-means method; the reference document page title corresponding to each cluster center is obtained;

[0138] Exemplarily, the process of clustering the reference document page vectors is explained as follows:

[0139] The reference document page vectors are clustered. In the specific implementation, the number of clusters x extracted from the reference document page titles will be preset (it can be a default value). The purpose of the clustering process is to cluster the data set composed of all reference document page vectors to obtain the number of sets indicated by the number of clusters x.

[0140] During clustering:

[0141] 1. Randomly select x reference document page vectors from the data set (the set of all reference document page vectors) as the centroid;

[0142] 2. For each reference document page vector in the data set, calculate the distance (such as Euclidean distance or Manhattan distance) between the reference document page vector and each centroid, and divide the reference document page vector into the set to which the centroid indicated by the shortest distance belongs;

[0143] 3. Then recalculate the centroid for each set based on the centroid calculation formula;

[0144] 4. Calculate the target distance between the new centroid and the original centroid, and determine whether the clustering process is terminated based on the distance. If terminated, sort the reference document page vectors in the category by cluster size, and take the top X words in each category as the reference document page title; if not terminated, execute the above steps 2-4.

[0145] Optionally, whether the clustering process is terminated is determined based on the distance, and a distance threshold may be set. When the target distance is less than the distance threshold, the process is terminated. Otherwise, the above steps 2-4 are continued.

[0146] Optionally, the distance (between each pair of feature vectors) may be calculated using at least one of a Euclidean distance formula, a Manhattan distance formula, a cosine distance formula, a correlation coefficient distance formula, and the like.

[0147] B6: Generate a presentation title content extraction prompt word based on the reference presentation title and the reference presentation manuscript set;

[0148] Illustratively, generating a presentation title content extraction prompt word based on the reference presentation title and the reference presentation manuscript set;

[0149] B8: Input the demonstration title content extraction prompt words, the reference demonstration title and the reference presentation manuscript set into the target large language generation model to perform comprehensive processing on the demonstration title content, and obtain the title demonstration content corresponding to at least one demonstration title.

[0150] Schematically, the demo title content extraction prompt words, the reference demo title and the reference demo manuscript set are input into the target large language generative model for comprehensive processing of the demo title content, and the title demo content corresponding to at least one demo title is obtained, that is, all the demo contents of each category are summarized and typeset using the target large language generative model, and the theme demo contents corresponding to several demo titles can be obtained;

[0151] S306: Filling the presentation template with title content based on the title presentation content;

[0152] Illustratively, after extracting a number of title presentation contents from a reference presentation set using a target large language generative model, the presentation template is used to insert the number of title presentation contents into the presentation template to fill in the title presentation contents related to the reference presentation title.

[0153] Furthermore, the presentation template may be determined based on the reference presentation templates corresponding to all reference presentations in the reference presentation set, and the reference presentation template that is used most times in the reference presentation set may be taken as the final presentation template.

[0154] S308: After completing the content filling, generate the target presentation;

[0155] The target presentation can be a PPT presentation (PPT, PowerPoint) file;

[0156] It can be understood that the document content slices corresponding to each demonstration project name are obtained, and then the document content slices are filled into the presentation template. After all the document content is filled (it can be that the user inputs a confirmation content filling completion instruction, and the electronic device responds to the content filling completion instruction to determine that all the document content is filled), the final target presentation is obtained. If there is a next content filling process, the step of obtaining the document generation material input by the user is executed.

[0157] It should be noted that, in the process of generating the target presentation, the user can input different document generation materials multiple times. The multiple document generation materials can be of different types, for example, they can be presentation generation prompt word types, document material types, and reference presentation collection types. Different types of document generation materials can execute different content filling processes. For example, after determining that the currently input document generation material is a reference presentation collection, S304 can be executed. For example, after determining that the currently input document generation material is a document material, S204 can be executed, and so on, until all document contents are filled in. For example, the user inputs a confirmation content filling completion instruction, and the electronic device determines that all document contents are filled in response to the content filling completion instruction.

[0158] In one or more embodiments of the present specification, the electronic device generates material by obtaining a document input by a user, and when the document generation material is a reference presentation set, the document content is processed by using a target large language generative model based on the reference presentation set to obtain a title presentation content. The presentation template can then be filled with content based on the title presentation content to generate a target presentation. The entire process does not require a large amount of time for manual production of presentations, and the efficiency of presentation processing is improved based on the target large language generative model, thereby saving presentation processing time and optimizing the presentation production process.

[0159] See also Figure 4 , Figure 4 This is a flowchart of a content filling processing method proposed in this application. Specifically:

[0160] S402: Acquire presentation generation prompt words input by the user, and generate a presentation outline using a target language generative model based on the presentation generation prompt words to obtain a presentation outline;

[0161] The presentation generation prompt words can be understood as prompt description words that instruct the target large language generation model to process the content of the document. For example, if the user expects to make a year-end performance review presentation, the presentation generation prompt words input by the user can be "assuming that you are a presentation production expert, please help me make a mid-year performance review presentation";

[0162] After the electronic device obtains the presentation generation prompt words input by the user, it uses the target language generation model to generate a presentation outline based on the presentation generation prompt words to obtain a presentation outline;

[0163] Exemplarily, the electronic device may provide a human-computer interaction interface, which may be a target interface for providing automatic generation of presentations, wherein the target interface includes at least a presentation operation area and a presentation generation interaction area, wherein the presentation operation area is used to display display elements related to presentation functions, and the presentation generation interaction area may be used for a user to input presentation generation materials and instruct a target large language generation model to process presentation content;

[0164] Furthermore, the user can input presentation generation prompt words in the presentation generation interactive area, for example, "Suppose you are a presentation production expert, please help me make a mid-year performance review presentation." The electronic device then obtains the presentation generation prompt words input by the user, and then generates a document outline based on the presentation generation prompt words using the target large language generative model, thereby obtaining a presentation outline.

[0165] The content of the presentation outline generally includes a title and one or more first-level outline nodes, and the first-level outline nodes respectively include several levels of child nodes in a tree structure, such as second-level nodes, third-level nodes, and fourth-level nodes. In this specification, the presentation outline is automatically generated by the target language generative model according to the presentation generation prompt words; since the presentation outline includes the outline content under the multi-level results, the user can subsequently input the document materials for content filling based on the presentation outline.

[0166] Furthermore, the electronic device can automatically update the presentation outline to the presentation operation area, and the user can intuitively view and conveniently operate in the presentation operation area dedicated to presentation processing.

[0167] S404: receiving the document generation material input by the user for the presentation outline.

[0168] Exemplarily, the electronic device obtains a presentation outline and can display the presentation outline to the user, and the user can upload a document to generate material based on the outline content of the presentation outline.

[0169] Furthermore, the user can upload the presentation generation material in the presentation generation interactive area and can input the corresponding material processing information, such as the material processing information is used to instruct the target large language generation model how to process based on the presentation generation material;

[0170] S406: Based on the manuscript generation material, a target large language generative model is used to process the manuscript content to obtain reference manuscript content, and the presentation template is filled with content based on the reference manuscript content;

[0171] The electronic device may determine the type of the manuscript generation material, and optionally, may determine that the manuscript generation material is a type of document material or a type of reference presentation manuscript collection;

[0172] Optionally, if the manuscript generation material is a document material, then steps S202 and the like are started to be executed. In addition, after the electronic device obtains the name of the demonstration project, the demonstration project name is generated based on the document material using the target large language generation model after the steps S202 and the like are started to be executed, and the electronic device can adjust the presentation outline based on the demonstration project name in parallel, such as updating the outline node content of the presentation outline based on the demonstration project name;

[0173] Optionally, if the document generation material is a reference presentation set, then steps S302 and the like are started to be executed. After the external electronic device obtains at least one presentation title, the presentation title is generated based on the reference presentation set after the steps S302 and the like are started to be executed. The electronic device can adjust the presentation outline based on the presentation title in parallel. For example, the outline node content of the presentation outline is updated based on the presentation title;

[0174] S408: After completing the content filling, generate the target presentation

[0175] In one or more embodiments of the present specification, the process of automatic presentation generation is shown, which can greatly reduce the time spent by users in making presentations and improve overall efficiency. The entire process does not require a large amount of time spent on manual presentation production. When the presentation generation material is a document material, the efficiency of presentation processing is improved based on the target large language generative model, which saves presentation processing time and optimizes the presentation production process.

[0176] The following will be combined Figure 5 , the presentation document generation device provided in the embodiment of the present application is introduced in detail. It should be noted that, Figure 5 The presentation generating device shown is used to execute the present application Figure 1 to Figure 4 For the convenience of explanation, only the part related to the embodiment of the present application is shown. For the specific technical details not disclosed, please refer to the present application. Figure 1 to Figure 4 The embodiment shown.

[0177] See also Figure 5, which shows a schematic diagram of the structure of the presentation generating device of the embodiment of the present application. The presentation generating device 1 can be implemented as all or part of the user terminal through software, hardware or a combination of both. According to some embodiments, the presentation generating device 1 includes a material acquisition module 11 and a document processing module 12, which are specifically used to:

[0178] The material acquisition module 11 is used to acquire the manuscript input by the user to generate the material;

[0179] The document processing module 12 is used to process the document content based on the document generation material using a target large language generative model to obtain reference document content, and to fill the presentation template with content based on the reference document content to generate a target presentation.

[0180] Optional, such as Figure 6 As shown, the document processing module 12 includes:

[0181] A material determination unit 121, configured to determine that the manuscript generation material is a document material;

[0182] A name extraction unit 122 is used to extract the project name based on the document material using a target large language generative model to obtain a demonstration project name;

[0183] The content filling unit 123 is used to determine the document content slice corresponding to the presentation project name based on the document material, and fill the presentation template with the slice content based on the document content slice.

[0184] Optionally, the name extraction unit 122 is used to:

[0185] Generate a project name extraction prompt word for the document material, input the project name extraction prompt word and the document material into the target large language generation model to extract the project name, and obtain a demonstration project name.

[0186] Optionally, the content filling unit 123 is used to:

[0187] Vectorizing the document material to obtain document vector image data; vectorizing the demonstration project name to obtain a demonstration project name vector;

[0188] The demonstration project name vector performs a vector index search process on the demonstration project name vector in the document vector image data to obtain the name position information corresponding to the demonstration project name;

[0189] The document material is processed by document content segmentation based on the name location information to obtain document content slices corresponding to the demonstration project name.

[0190] Optionally, the document processing module 12 is used to:

[0191] Determining that the manuscript generation material is a reference presentation manuscript set, wherein the reference presentation manuscript set includes at least one reference presentation manuscript;

[0192] Based on the reference presentation manuscript set, a target large language generative model is used to perform comprehensive processing on the presentation title content to obtain a title presentation content corresponding to at least one presentation title;

[0193] The presentation template is filled with title content based on the title presentation content.

[0194] Optionally, the document processing module 12 is used to:

[0195] Performing document page vectorization processing on the reference presentation document set to obtain a reference document page vector corresponding to at least one reference presentation document page;

[0196] Performing vector clustering processing on all reference manuscript page vectors to obtain at least one reference cluster center, and obtaining a reference demonstration title corresponding to the reference cluster center;

[0197] Generate a presentation title content extraction prompt word based on the reference presentation title and the reference presentation manuscript set;

[0198] The demonstration title content extraction prompt words, the reference demonstration title and the reference presentation manuscript set are input into the target large language generation model for comprehensive processing of the demonstration title content to obtain the title demonstration content corresponding to at least one demonstration title.

[0199] Optionally, the device 1 is also used for:

[0200] Acquire a presentation generation prompt word input by a user, and generate a presentation outline using a target language generation model based on the presentation generation prompt word to obtain a presentation outline;

[0201] The material acquisition module is used to:

[0202] Receive the document generation material input by the user for the presentation outline.

[0203] Optionally, the document processing module 12 is used to:

[0204] If the manuscript generation material is a document material, a presentation project name is obtained, the presentation project name is generated based on the document material using a target language generation model, and the presentation outline is adjusted based on the presentation project name; and / or,

[0205] If the document generation material is a reference presentation set, at least one presentation title is obtained, the presentation title is generated based on the reference presentation set, and the presentation outline is adjusted based on the presentation title.

[0206] It should be noted that the presentation generation device provided in the above embodiment only uses the division of the above functional modules as an example when executing the presentation generation method. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the presentation generation device provided in the above embodiment and the presentation generation method embodiment belong to the same concept, and the implementation process thereof is detailed in the method embodiment, which will not be repeated here.

[0207] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0208] In one or more embodiments of the present application, the electronic device obtains the document generation material input by the user, and then uses the target large language generative model to process the document content based on the document generation material to obtain the reference document content. Then, the presentation template can be filled with the content based on the reference document content to generate the target presentation. The whole process does not require a lot of time for manual production of the presentation. The efficiency of presentation processing is improved based on the target large language generative model, which saves presentation processing time and optimizes the presentation production process.

[0209] The present application also provides a computer storage medium, which can store multiple instructions, and the instructions are suitable for being loaded and executed by a processor as described above. Figure 1 to Figure 4 The method for generating a presentation in the embodiment shown in the figure can be found in the specific execution process. Figure 1 to Figure 4 The specific description of the illustrated embodiment will not be repeated here.

[0210] The present application also provides a computer program product, which stores at least one instruction, and the at least one instruction is loaded and executed by the processor as described above. Figure 1 to Figure 4 The method for generating a presentation in the embodiment shown in the figure can be found in the specific execution process. Figure 1 to Figure 4 The specific description of the illustrated embodiment will not be repeated here.

[0211] Please refer to Figure 7, which shows a block diagram of the structure of an electronic device provided by an exemplary embodiment of the present application. The electronic device in the present application may include one or more of the following components: a processor 110, a memory 120, an input device 130, an output device 140, and a bus 150. The processor 110, the memory 120, the input device 130, and the output device 140 may be connected via the bus 150.

[0212] The processor 110 may include one or more processing cores. The processor 110 uses various interfaces and lines to connect various parts of the entire electronic device, and executes various functions and processes data of the electronic device 100 by running or executing instructions, programs, code sets or instruction sets stored in the memory 120, and calling data stored in the memory 120. Optionally, the processor 110 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 110 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 110, but may be implemented separately through a communication chip.

[0213] The memory 120 may include a random access memory (RAM) or a read-only memory (ROM). Optionally, the memory 120 includes a non-transitory computer-readable storage medium. The memory 120 may be used to store instructions, programs, codes, code sets or instruction sets. The memory 120 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as a touch function, a sound playback function, an image playback function, etc.), instructions for implementing the following various method embodiments, etc. The operating system may be an Android system, including a system deeply developed based on the Android system, an IOS system developed by Apple, including a system deeply developed based on the IOS system or other systems. The data storage area may also store data created by the electronic device during use, such as a phone book, audio and video data, chat record data, etc.

[0214] See also Figure 8 As shown, the memory 120 can be divided into an operating system space and a user space. The operating system runs in the operating system space, and native and third-party applications run in the user space. In order to ensure that different third-party applications can achieve good operating results, the operating system allocates corresponding system resources to different third-party applications. However, different application scenarios in the same third-party application also have different requirements for system resources. For example, in the local resource loading scenario, the third-party application has higher requirements for disk reading speed; in the animation rendering scenario, the third-party application has higher requirements for GPU performance. The operating system and third-party applications are independent of each other, and the operating system often cannot perceive the current application scenario of the third-party application in a timely manner, resulting in the operating system being unable to perform targeted system resource adaptation according to the specific application scenario of the third-party application.

[0215] In order for the operating system to distinguish the specific application scenarios of third-party applications, it is necessary to open up data communication between third-party applications and the operating system so that the operating system can obtain the current scenario information of third-party applications at any time, and then perform targeted system resource adaptation based on the current scenario.

[0216] Taking the Android operating system as an example, the programs and data stored in the memory 120 are as follows: Fig. 9As shown, the memory 120 may store a Linux kernel layer 320, a system runtime library layer 340, an application framework layer 360 and an application layer 380, wherein the Linux kernel layer 320, the system runtime library layer 340 and the application framework layer 360 belong to the operating system space, and the application layer 380 belongs to the user space. The Linux kernel layer 320 provides underlying drivers for various hardware of electronic devices, such as display drivers, audio drivers, camera drivers, Bluetooth drivers, Wi-Fi drivers, power management, etc. The system runtime library layer 340 provides the main feature support for the Android system through some C / C++ libraries. For example, the SQLite library provides database support, the OpenGL / ES library provides 3D drawing support, and the Webkit library provides browser kernel support, etc. The Android runtime library (Android runtime) is also provided in the system runtime library layer 340, which mainly provides some core libraries that allow developers to use the Java language to write Android applications. The application framework layer 360 provides various APIs that may be used when building applications. Developers can also use these APIs to build their own applications, such as activity management, window management, view management, notification management, content provider, package management, call management, resource management, and location management. At least one application runs in the application layer 380. These applications can be native applications that come with the operating system, such as contact applications, text messaging applications, clock applications, camera applications, etc.; they can also be third-party applications developed by third-party developers, such as game applications, instant messaging applications, photo beautification applications, etc.

[0217] Taking the operating system as an IOS system as an example, the programs and data stored in the memory 120 are as follows: Fig.10As shown, the IOS system includes: a core operating system layer 420 (Core OS layer), a core service layer 440 (Core Services layer), a media layer 460 (Media layer), and a touchable layer 480 (Cocoa Touch Layer). The core operating system layer 420 includes the operating system kernel, drivers, and underlying program frameworks, which provide functions closer to the hardware for use by the program framework located in the core service layer 440. The core service layer 440 provides system services and / or program frameworks required by the application, such as the foundation framework, account framework, advertising framework, data storage framework, network connection framework, geographic location framework, motion framework, etc. The media layer 460 provides audio-visual interfaces for the application, such as graphics and image related interfaces, audio technology related interfaces, video technology related interfaces, and wireless playback (AirPlay) interfaces for audio and video transmission technologies. The touchable layer 480 provides various commonly used interface-related frameworks for application development, and the touchable layer 480 is responsible for the user's touch interaction operations on the electronic device. For example, local notification service, remote push service, advertising framework, game tool framework, message user interface (UI) framework, user interface UIKit framework, map framework, etc.

[0218] exist Fig.10 Among the frameworks shown, the frameworks related to most applications include but are not limited to: the basic framework in the core service layer 440 and the UIKit framework in the touchable layer 480. The basic framework provides many basic object classes and data types, provides the most basic system services for all applications, and has nothing to do with UI. The classes provided by the UIKit framework are basic UI class libraries for creating touch-based user interfaces. iOS applications can provide UIs based on the UIKit framework, so it provides the basic architecture of applications for building user interfaces, drawing, processing and user interaction events, responding to gestures, etc.

[0219] Among them, the method and principle of implementing data communication between third-party applications and the operating system in the IOS system can be referred to the Android system, and this application will not go into details here.

[0220] Among them, the input device 130 is used to receive input instructions or data, and the input device 130 includes but is not limited to a keyboard, a mouse, a camera, a microphone or a touch device. The output device 140 is used to output instructions or data, and the output device 140 includes but is not limited to a display device and a speaker. In one example, the input device 130 and the output device 140 can be combined, and the input device 130 and the output device 140 are touch screen displays, which are used to receive touch operations on or near the user using any suitable object such as a finger or a touch pen, and to display the user interface of each application. The touch screen display is usually set on the front panel of the electronic device. The touch screen display can be designed as a full screen, a curved screen or a special-shaped screen. The touch screen display can also be designed as a combination of a full screen and a curved screen, or a combination of a special-shaped screen and a curved screen, which is not limited in the embodiments of the present application.

[0221] In addition, those skilled in the art will appreciate that the structure of the electronic device shown in the above drawings does not constitute a limitation on the electronic device, and the electronic device may include more or fewer components than shown, or combine certain components, or arrange the components differently. For example, the electronic device also includes a radio frequency circuit, an input unit, a sensor, an audio circuit, a wireless fidelity (WiFi) module, a power supply, a Bluetooth module and other components, which will not be described in detail here.

[0222] In the embodiment of the present application, the execution subject of each step may be the electronic device described above. Optionally, the execution subject of each step is the operating system of the electronic device. The operating system may be an Android system, an IOS system, or other operating systems, which is not limited in the embodiment of the present application.

[0223] The electronic device of the embodiment of the present application may also be equipped with a display device, which may be any device capable of realizing a display function, such as a cathode ray tube display (CR), a light-emitting diode display (LED), an electronic ink screen, a liquid crystal display (LCD), a plasma display panel (PDP), etc. The user may use the display device on the electronic device 101 to view displayed text, images, videos and other information. The electronic device may be a smart phone, a tablet computer, a gaming device, an AR (Augmented Reality) device, a car, a data storage device, an audio playback device, a video playback device, a notebook, a desktop computing device, a wearable device such as an electronic watch, an electronic glasses, an electronic helmet, an electronic bracelet, an electronic necklace, electronic clothing and other devices.

[0224] exist Figure 7 In the electronic device shown, the processor 110 may be used to call an application stored in the memory 120 and specifically perform the following operations:

[0225] Obtain the document generation material input by the user;

[0226] Based on the manuscript generation material, a target large language generative model is used to process the manuscript content to obtain reference manuscript content, and based on the reference manuscript content, the presentation template is filled with content to generate a target presentation.

[0227] In one embodiment, the processor 110 performs the following operations when executing the step of generating the material based on the document using the target large language generative model to process the document content to obtain reference document content and filling the presentation template with content based on the reference document content:

[0228] Determining that the manuscript generation material is a document material;

[0229] Based on the document material, a target large language generative model is used to extract the project name to obtain a demonstration project name;

[0230] The document content slice corresponding to the presentation project name is determined based on the document material, and the presentation template is filled with the slice content based on the document content slice.

[0231] In one embodiment, the processor 110 performs the following operations when performing the project name extraction based on the document material using the target large language generative model to obtain the demonstration project name:

[0232] Generate a project name extraction prompt word for the document material, input the project name extraction prompt word and the document material into the target large language generation model to extract the project name, and obtain a demonstration project name.

[0233] In one embodiment, the processor 110 performs the following operations when determining the document content slice corresponding to the presentation project name based on the document material:

[0234] Vectorizing the document material to obtain document vector image data; vectorizing the demonstration project name to obtain a demonstration project name vector;

[0235] Based on the demonstration project name vector, a vector index search process is performed on the demonstration project name vector in the document vector image data to obtain name position information corresponding to the demonstration project name;

[0236] The document material is processed by document content segmentation based on the name location information to obtain document content slices corresponding to the demonstration project name.

[0237] In one embodiment, the processor 110 performs the step of processing the manuscript content using the target large language generative model based on the manuscript generation material to obtain reference manuscript content, and filling the presentation template with content based on the reference manuscript content, including:

[0238] Determining that the manuscript generation material is a reference presentation manuscript set, wherein the reference presentation manuscript set includes at least one reference presentation manuscript;

[0239] Based on the reference presentation manuscript set, a target large language generative model is used to perform comprehensive processing on the presentation title content to obtain a title presentation content corresponding to at least one presentation title;

[0240] The presentation template is filled with title content based on the title presentation content.

[0241] In one embodiment, the processor 110 performs the following operations when performing the comprehensive processing of the presentation title content based on the reference presentation set using the target large language generative model to obtain the title presentation content corresponding to at least one presentation title:

[0242] Performing document page vectorization processing on the reference presentation document set to obtain a reference document page vector corresponding to at least one reference presentation document page;

[0243] Performing vector clustering processing on all reference manuscript page vectors to obtain at least one reference cluster center, and obtaining a reference demonstration title corresponding to the reference cluster center;

[0244] Generate a presentation title content extraction prompt word based on the reference presentation title and the reference presentation manuscript set;

[0245] The demonstration title content extraction prompt words, the reference demonstration title and the reference presentation manuscript set are input into the target large language generation model for comprehensive processing of the demonstration title content to obtain the title demonstration content corresponding to at least one demonstration title.

[0246] In one embodiment, the processor 110 performs the following operations when executing the presentation generation method:

[0247] Acquire a presentation generation prompt word input by a user, and generate a presentation outline using a target language generation model based on the presentation generation prompt word to obtain a presentation outline;

[0248] The step of obtaining the manuscript input by the user to generate the material includes:

[0249] Receive the document generation material input by the user for the presentation outline.

[0250] In one embodiment, the processor 110 performs the following operations when executing the presentation generation method:

[0251] If the manuscript generation material is a document material, a presentation project name is obtained, the presentation project name is generated based on the document material using a target language generation model, and the presentation outline is adjusted based on the presentation project name; and / or,

[0252] If the document generation material is a reference presentation set, at least one presentation title is obtained, the presentation title is generated based on the reference presentation set, and the presentation outline is adjusted based on the presentation title.

[0253] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0254] The above disclosure is only the preferred embodiment of the present application, which certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A presentation generation method based on a target large language generative model, characterized in that: The method comprises: Obtain the document generation material input by the user; Based on the manuscript generation material, a target large language generative model is used to process the manuscript content to obtain reference manuscript content, and based on the reference manuscript content, the presentation template is filled with content to generate a target presentation.

2. The method according to claim 1, characterized in that The method of generating the material based on the manuscript using the target large language generative model to process the manuscript content to obtain reference manuscript content, and filling the presentation template with content based on the reference manuscript content includes: Determining that the manuscript generation material is a document material; Based on the document material, a target large language generative model is used to extract the project name to obtain a demonstration project name; The document content slice corresponding to the presentation project name is determined based on the document material, and the presentation template is filled with the slice content based on the document content slice.

3. The method according to claim 2, characterized in that The project name is extracted based on the document material using a target large language generative model to obtain a demonstration project name, including: Generate a project name extraction prompt word for the document material, input the project name extraction prompt word and the document material into the target large language generation model to extract the project name, and obtain a demonstration project name.

4. The method according to claim 2, characterized in that: The determining, based on the document material, the document content slice corresponding to the demonstration project name includes: Vectorizing the document material to obtain document vector image data; vectorizing the demonstration project name to obtain a demonstration project name vector; Based on the demonstration project name vector, a vector index search process is performed on the demonstration project name vector in the document vector image data to obtain name position information corresponding to the demonstration project name; The document material is processed by document content segmentation based on the name location information to obtain document content slices corresponding to the demonstration project name.

5. The method according to claim 1, characterized in that The method of generating the material based on the manuscript using the target large language generative model to process the manuscript content to obtain reference manuscript content, and filling the presentation template with content based on the reference manuscript content includes: Determining that the manuscript generation material is a reference presentation manuscript set, wherein the reference presentation manuscript set includes at least one reference presentation manuscript; Based on the reference presentation manuscript set, a target large language generative model is used to perform comprehensive processing on the presentation title content to obtain a title presentation content corresponding to at least one presentation title; The presentation template is filled with title content based on the title presentation content.

6. The method according to claim 5, characterized in that The method of performing comprehensive processing of presentation title content using a target large language generative model based on the reference presentation manuscript set to obtain title presentation content corresponding to at least one presentation title includes: Performing document page vectorization processing on the reference presentation document set to obtain a reference document page vector corresponding to at least one reference presentation document page; Performing vector clustering processing on all reference manuscript page vectors to obtain at least one reference cluster center, and obtaining a reference demonstration title corresponding to the reference cluster center; Generate a presentation title content extraction prompt word based on the reference presentation title and the reference presentation manuscript set; The demonstration title content extraction prompt words, the reference demonstration title and the reference presentation manuscript set are input into the target large language generation model for comprehensive processing of the demonstration title content to obtain the title demonstration content corresponding to at least one demonstration title.

7. The method according to claim 1, characterized in that The method further comprises: Acquire a presentation generation prompt word input by a user, and generate a presentation outline using a target language generation model based on the presentation generation prompt word to obtain a presentation outline; The step of obtaining the manuscript input by the user to generate the material includes: Receive the document generation material input by the user for the presentation outline.

8. The method according to claim 7, characterized in that The method further comprises: If the manuscript generation material is a document material, a presentation project name is obtained, the presentation project name is generated based on the document material using a target language generation model, and the presentation outline is adjusted based on the presentation project name; and / or, If the document generation material is a reference presentation set, at least one presentation title is obtained, the presentation title is generated based on the reference presentation set, and the presentation outline is adjusted based on the presentation title.

9. A presentation manuscript generation device based on a target large language generative model, characterized in that: The device comprises: The material acquisition module is used to acquire the manuscript generated by the user input; The document processing module is used to process the document content based on the document generation material using a target large language generative model to obtain reference document content, and to fill the presentation template with content based on the reference document content to generate a target presentation.

10. An electronic device, characterized in that: include: A processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the method steps as claimed in any one of claims 1 to 8.

Images