Information processing method and device, electronic equipment and storage medium

Abstract

The invention provides an information processing method and device, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence. According to the specific implementation scheme, entity relation extraction is conducted on an original creation text, and a first knowledge graph is generated; performing entity community division according to node attributes of nodes in the first knowledge graph, and updating the first knowledge graph based on the divided entity communities and hierarchical structures of the entity communities to obtain a second knowledge graph; and acquiring plot information of the to-be-created chapter, and generating a text of the to-be-created chapter according to the plot information and the second knowledge graph.

Description

Technical Field

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

[0002] For article writers, the big model can be used to learn and understand the creative ideas and techniques of excellent works by disassembling popular works, providing inspiration for their own creations. At the same time, this method can also help authors sort out their own creative outlines and article ideas, and assist authors in creating subsequent chapter content.

[0003] In the prior art, when the current chapter needs to be created, the large model will use the content of the previous chapter and information such as character introductions, without understanding the overall story outline of the novel. The logical disorder of the article may occur, and the final result is likely to deviate from the main line of the story. It cannot be used directly as the main text content and the author needs to make a lot of modifications. Summary of the invention

[0004] The present disclosure provides an information processing method, device, electronic device and storage medium.

[0005] According to one aspect of the present disclosure, there is provided an information processing method, comprising: extracting entity relationships from an original creative text to generate a first knowledge graph; dividing entity communities according to node attributes of nodes in the first knowledge graph, and updating the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph; acquiring plot information of a chapter to be created, and generating the main text of the chapter to be created based on the plot information and the second knowledge graph.

[0006] According to another aspect of the present disclosure, there is provided an information processing device, including: a first generation module, used to extract entity relationships from an original creative text to generate a first knowledge graph; an update module, used to divide entity communities according to node attributes of nodes in the first knowledge graph, and update the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph; a second generation module, used to obtain plot information of a chapter to be created, and generate the main text of the chapter to be created based on the plot information and the second knowledge graph.

[0007] According to another aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the information processing method described in the above-mentioned one aspect embodiment.

[0008] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, on which a computer program / instructions are stored, and the computer instructions are used to enable the computer to execute the information processing method described in the above-mentioned embodiment.

[0009] According to another aspect of the present disclosure, a computer program product is provided, including a computer program / instruction, wherein when the computer program / instruction is executed by a processor, the information processing method described in the above-mentioned embodiment is implemented.

[0010] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0012] Figure 1 A flowchart of an information processing method provided by an embodiment of the present disclosure;

[0013] Figure 2 A flowchart of another information processing method provided by an embodiment of the present disclosure;

[0014] Figure 3 A schematic diagram of the process of adding a new knowledge graph provided in an embodiment of the present disclosure;

[0015] Figure 4 A flowchart of another information processing method provided by an embodiment of the present disclosure;

[0016] Figure 5 A schematic diagram of a process for generating a second knowledge graph provided in an embodiment of the present disclosure;

[0017] Figure 6 A flowchart of another information processing method provided by an embodiment of the present disclosure;

[0018] Figure 7 A schematic diagram of a process for generating the body of a chapter to be created provided by an embodiment of the present disclosure;

[0019] Figure 8 A schematic diagram of a process of splitting chapters in an information processing method provided in an embodiment of the present disclosure;

[0020] Fig. 9 A schematic diagram of a chapter splitting process provided in an embodiment of the present disclosure;

[0021] Fig.10A schematic diagram of the structure of an information processing device provided by an embodiment of the present disclosure;

[0022] Fig.11 The present invention is a block diagram of an electronic device for implementing the information processing method according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0024] The information processing method, apparatus, and electronic device according to the embodiments of the present disclosure are described below with reference to the accompanying drawings.

[0025] Artificial Intelligence (AI) is a discipline that studies how computers can simulate certain thought processes and intelligent behaviors of human beings (such as learning, reasoning, thinking, planning, etc.). It includes both hardware-level technologies and software-level technologies. AI hardware technologies generally include computer vision technology, speech recognition technology, natural language processing technology, as well as learning / deep learning, big data processing technology, knowledge graph technology, and other aspects.

[0026] Figure 1 A flowchart of an information processing method provided by an embodiment of the present disclosure.

[0027] like Figure 1 As shown, the information processing method may include:

[0028] S101, extract entity relationships from the original creative text to generate a first knowledge graph.

[0029] It should be noted that the execution subject of the information processing method in the embodiment of the present disclosure may be a hardware device with data processing capabilities and / or the necessary software required to drive the hardware device to work. Optionally, the execution subject may include a server, a user terminal and other intelligent devices. Optionally, the user terminal includes but is not limited to a mobile phone, a computer, an intelligent voice interaction device, etc. Optionally, the server includes but is not limited to a network server, an application server, and may also be a server of a distributed system, or a server combined with a blockchain, etc. The embodiment of the present disclosure is not specifically limited.

[0030] In some implementations, the original creative text may be split to obtain multiple text blocks, entities and entity relationships may be extracted from the text blocks, and a first knowledge graph may be generated based on the entities and entity relationships. Optionally, the original creative text may be split according to its chapters.

[0031] In some implementations, the chapter titles in the original creative text are identified and the original creative text is split into text blocks equal in number to the chapter titles, that is, a text block corresponding to each chapter can be split.

[0032] Furthermore, key information such as chapter summaries, character introductions, and writing techniques are extracted from the text blocks, and entities are extracted from the key information. The entities may include: character relationships, plot points, and story background, and the relationships between character relationships, plot points, and story background are taken as entity relationships.

[0033] S102, dividing entity communities according to node attributes of nodes in the first knowledge graph, and updating the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph.

[0034] In some implementations, the nodes in the first knowledge graph are entities. By obtaining the node attributes of the nodes in the first knowledge graph and dividing the entities into communities based on the node attributes, multiple entity communities can be obtained, and then the hierarchical structure of the entity communities can be determined based on the relationships between the entity communities. The second knowledge graph can be obtained by generating a summary of each entity community and updating the summary and the hierarchical structure to the first knowledge graph. Among them, the node attributes include but are not limited to: characters, chapter summaries, etc.

[0035] Optionally, entities with the same node attributes may be divided into an entity community. For example, if entity A, entity B, and entity C are all person nodes, entity A, entity B, and entity C are divided into the same entity community.

[0036] Optionally, in order to reduce the complexity of data, the entity communities may be divided according to a certain number of communities. Optionally, if the number of entity communities reaches a set community threshold, the community division is terminated to obtain multiple entity communities.

[0037] S103, obtaining the plot information of the chapter to be created, and generating the main text of the chapter to be created based on the plot information and the second knowledge graph.

[0038] Optionally, the plot information of the chapter to be created can be obtained from the user's input information. Optionally, the plot information of the chapter to be created can also be determined based on the original creation text. The outline of the original creation text is extracted, and the plot information of the chapter to be created is generated by the large model based on the outline. For example, a pre-trained plot generation model can be used to generate the plot information.

[0039] Optionally, the plot information includes information such as characters and writing techniques, and target key information related to the plot information can be retrieved from the second knowledge graph based on the plot information to generate the main text of the chapter to be created based on the target key information.

[0040] Exemplarily, assuming that the plot information includes character A, a search is performed in the second knowledge graph based on character A to retrieve character B, character C, and plot A related to character A in the second knowledge graph, and character A, character B, character C, and plot A are used as target key information to generate the main text based on the target key information.

[0041] Optionally, a pre-trained text generation model can be used to generate the text based on the target key information.

[0042] According to the information processing method provided by the embodiment of the present disclosure, by extracting entities and entity relationships from the original creative text, a first knowledge graph can be generated based on the entities and entity relationships. Furthermore, in order to improve the efficiency of information retrieval and query, the entities in the first knowledge graph can be divided into entity communities, and according to the divided entity communities and the hierarchical structure of the entity communities, the first knowledge graph is updated to obtain the second knowledge graph. Furthermore, the main text of the chapter to be created can be generated based on the plot information of the chapter to be created and the second knowledge graph to obtain the main text of the chapter to be created. In the present disclosure, by generating a knowledge graph from the original creative text and displaying it to the user, it can be convenient for the user to sort out creative ideas, frameworks, narrative rhythms, character settings and other information to provide inspiration for their own creation. The generation of the main text of the article based on the knowledge graph can greatly enhance the text logic of the main text and improve the readability of the text.

[0043] Figure 2 A flowchart of an information processing method provided by an embodiment of the present disclosure.

[0044] like Figure 2 As shown, the information processing method may include:

[0045] S201, extract entity relationships from the original creative text to generate a first knowledge graph.

[0046] The relevant contents of step S201 can be found in the above embodiment and will not be described again here.

[0047] S202, based on the hierarchical Leiden algorithm, recursively perform community clustering on the nodes in the first knowledge graph according to the node attributes.

[0048] Optionally, nodes with the same node attributes may be recursively clustered based on the hierarchical Leiden algorithm to obtain multiple entity communities. For example, nodes with the node attribute of person may be recursively clustered to obtain entity communities containing person entities.

[0049] S203: In response to the number of entity communities reaching a set community threshold, the community division is terminated and a plurality of entity communities are acquired.

[0050] In some implementations, in order to avoid over-dividing communities, thereby saving computing resources and time and improving computing efficiency, a threshold for the number of communities can be set to determine whether to end community division based on the set community threshold. If the number of entity communities reaches the set community threshold, the community division is ended, and multiple entity communities with the set community threshold can be obtained.

[0051] S204: Based on the divided entity communities and the hierarchical structure of the entity communities, the first knowledge graph is updated to obtain a second knowledge graph.

[0052] In some implementations, the first knowledge graph can be updated based on the summary and hierarchy of the entity community, thereby ensuring that the entities and relationships in the knowledge graph are more accurate and improving the accuracy and completeness of the knowledge graph. Optionally, the hierarchy of the entity community can be determined based on the entity relationships between communities, and a summary of each entity community can be generated based on the large model. Further, based on the summary and hierarchy of the entity community, the first knowledge graph is updated to obtain the second knowledge graph.

[0053] In some implementations, during the creation process, there will be new chapters that need to be updated in the knowledge graph. Optionally, the local knowledge graph corresponding to the new chapters can be added to the second knowledge graph to improve the timeliness and completeness of the knowledge graph.

[0054] Optionally, by obtaining the newly added text and constructing a local knowledge graph of the newly added text, the local knowledge graph can be constructed by extracting entities and entity relationships in the newly added text.

[0055] Furthermore, the local knowledge graph and the second knowledge graph are matched with roles to obtain the role matching result of the local knowledge graph, and the role matching result includes whether the role in the local knowledge graph is the same role in the second knowledge graph. In other words, role matching can be performed according to the character identifier. If the character identifier is the same, it can be determined whether the role in the local knowledge graph is the same role in the second knowledge graph. Then, based on the role matching result, the role in the local knowledge graph can be associated with the second knowledge graph, and the chapter information of the text can be added to the second knowledge graph synchronously.

[0056] Optionally, the relevant information of the associated roles in the second knowledge graph may be updated based on the role matching results, and / or the novel outline corresponding to the second knowledge graph may be updated.

[0057] For example, Figure 3 The flowchart of adding a new knowledge graph is shown in the figure. For the newly added text, a local knowledge graph of the newly added text is constructed, and the role relationship, character information and other contents are searched in the second knowledge graph to match the local knowledge graph with the second knowledge graph to obtain the role matching result. The role matching result includes whether the role in the local knowledge graph is the same role in the second knowledge graph, as well as the plot and writing techniques associated with the same role. Figure 3 In the example, role A and role M are the same role. Furthermore, based on the role matching result, the role in the local knowledge graph can be associated with the second knowledge graph, and the chapter information of the text can be added to the second knowledge graph synchronously.

[0058] S205, obtaining the plot information of the chapter to be created, and generating the main text of the chapter to be created based on the plot information and the second knowledge graph.

[0059] The relevant contents of step S205 can be found in the above embodiment and will not be described again here.

[0060] According to the information processing method provided by the embodiment of the present disclosure, by extracting entities and entity relationships from the original creative text, a first knowledge graph can be generated based on the entities and entity relationships. Further, the hierarchical Leiden algorithm is used to divide the entities in the first knowledge graph into entity communities, and according to the divided entity communities and the hierarchical structure of the entity communities, the first knowledge graph is updated to obtain a second knowledge graph. Furthermore, the body of the chapter to be created can be generated based on the plot information of the chapter to be created and the second knowledge graph to obtain the body of the chapter to be created. In the present disclosure, by generating a knowledge graph from the original creative text and displaying it to the user, it can be convenient for the user to sort out information such as creative ideas, frameworks, narrative rhythms, character settings, etc., to provide inspiration for their own creation. The generation of the body of the article based on the knowledge graph can greatly enhance the text logic of the body and improve the readability of the text.

[0061] Figure 4 A flowchart of an information processing method provided by an embodiment of the present disclosure.

[0062] like Figure 4 As shown, the information processing method may include:

[0063] S401, extract key information from the chapter text of the original creative text.

[0064] In some implementations, key information can be extracted from the text blocks corresponding to the chapters of the original creative text to improve information acquisition efficiency. Optionally, by splitting the chapters of the original creative text, a text block corresponding to each chapter is obtained, and key information is extracted from the chapter body of the text block based on the large model, and the key information at least includes chapter summary, character information, key plot, writing techniques and chapter outline.

[0065] S402, extracting entity relationships from key information to obtain entity relationship triples.

[0066] Optionally, entity relationships can be determined by extracting entities from the key information. Optionally, entity relationships between characters, plots, and backgrounds can be determined as entity relationship triples by extracting entities such as characters, plots, and backgrounds.

[0067] S403: Generate a first knowledge graph based on the entity relationship triples.

[0068] In some implementations, an initial third knowledge graph may be generated based on entity relationship triples, and entity disambiguation may be performed on the third knowledge graph to obtain a first knowledge graph, so as to eliminate redundant information in the knowledge graph, making the graph more concise and clear, thereby improving the quality and accuracy of the knowledge graph. Optionally, the first knowledge graph may be obtained by identifying duplicate entities from the third knowledge graph and merging and disambiguating the duplicate entities.

[0069] S404, dividing entity communities according to node attributes of nodes in the first knowledge graph, and updating the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph.

[0070] S405, obtaining the plot information of the chapter to be created, and generating the main text of the chapter to be created based on the plot information and the second knowledge graph.

[0071] The relevant contents of steps S404-S405 can be found in the above embodiment and will not be repeated here.

[0072] According to the information processing method provided by the embodiment of the present disclosure, by extracting entities from the text block of the original creative text, the entity relationship triples can be determined, and the third knowledge graph can be generated based on the entity relationship triples, and then entity disambiguation can be performed to obtain a concise and clear first knowledge graph. Further, in order to improve the efficiency of information retrieval and query, the entities in the first knowledge graph can be divided into entity communities, and the first knowledge graph can be updated according to the divided entity communities and the hierarchical structure of the entity communities to obtain the second knowledge graph. Further, the body of the chapter to be created can be generated based on the plot information of the chapter to be created and the second knowledge graph to obtain the body of the chapter to be created. In the present disclosure, by generating a knowledge graph from the original creative text and displaying it to the user, it can be convenient for the user to sort out the creative ideas, framework, narrative rhythm, character settings and other information, and provide inspiration for their own creation. The generation of the body of the article based on the knowledge graph can greatly enhance the text logic of the body and improve the readability of the text.

[0073] The flowchart of generating the second knowledge graph is shown in Figure 5. By disassembling and splitting the original creative text, the text blocks corresponding to each chapter can be obtained, and the chapter summary, character information, key plot, writing techniques and chapter outline can be extracted from the text blocks as key information, and then entities can be extracted from the key information to form the third knowledge graph, and the repeated entities in the third knowledge graph are merged and disambiguated to obtain the first knowledge graph. Figure 5 In the example, if there are two roles A in the first knowledge graph, the duplicate roles A can be merged to obtain the first knowledge graph. Furthermore, the entities in the first knowledge graph are divided into communities to obtain multiple entity communities and a hierarchical structure of the entity communities. A second knowledge graph is obtained by generating a summary of the entity community and updating the summary and hierarchical structure of the entity community to the first knowledge graph. After obtaining the second knowledge graph, the second knowledge graph can be displayed on the user interface for the user to view, and the second knowledge graph can also be stored for subsequent use.

[0074] Figure 6 A flowchart of an information processing method provided by an embodiment of the present disclosure.

[0075] like Figure 6 As shown, the information processing method may include:

[0076] S601, extract entity relationships from the original creative text to generate a first knowledge graph.

[0077] S602, dividing entity communities according to node attributes of nodes in the first knowledge graph, and updating the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph.

[0078] The relevant contents of steps S601-S602 can be found in the above embodiment and will not be repeated here.

[0079] S603, obtaining the plot information of the chapter to be created.

[0080] In some implementations, the plot information of the chapter to be created can be obtained based on the user's input information, and the plot information of the chapter to be created can also be generated based on the original creation text. That is, if the user inputs the plot points, the plot information of the chapter to be created can be generated based on the plot points; if the user does not input the plot points, the plot information of the chapter to be created can be generated based on the plot of the original creation text. Generating plot information based on the user's input makes the creation more creative and personalized, and generating plot information based on the original creation text ensures the coherence of the plot, thereby enhancing the readability of the text.

[0081] Optionally, by obtaining user input information, identifying plot points from the user input information, and determining the plot information of the chapter to be created based on the plot points. Optionally, by extracting the above outline from the original creation text and calling a pre-trained plot generation model, the plot generation model generates the plot information of the chapter to be created according to the above outline.

[0082] S604, determining the end of the previous text according to the chapter to be created, and extracting relevant search information according to the end of the previous text and plot information.

[0083] Optionally, information extraction can be performed on the ending of the above text and the plot information, and the extracted identical information can be used as relevant search information. For example, if character A is extracted from both the ending of the above text and the plot information, character A can be used as relevant search information.

[0084] S605: Perform graph retrieval on the second knowledge graph based on relevant search information to obtain target key information for creation.

[0085] In some implementations, entities and entity relationships related to the relevant search information can be retrieved from the second knowledge graph as target key information to improve the accuracy and relevance of the search results and enhance the ability to understand the information.

[0086] Optionally, the chapter number and role ID of the chapter to be created can be determined based on relevant search information, and key information can be matched in different entity communities in the second knowledge graph based on the chapter number and role ID to obtain target key information. In other words, the chapter number and role ID can be matched in the chapter community and role community in the second knowledge graph to obtain target key information that matches the chapter number and role ID.

[0087] S606, generating the main text of the chapter to be created based on the target key information and plot information.

[0088] In some implementations, in order to improve creation efficiency and shorten creation time, a large model can be used to generate the text. By calling a pre-trained text generation model, the text generation model generates the text of the chapter to be created based on the target key information and plot information.

[0089] like Figure 7 The flowchart of generating the main text of the chapter to be created is shown. By judging whether there are plot points in the user's input information, if there are plot points, the plot information of the chapter to be created is determined according to the plot points; if there are no plot points, the above outline is obtained from the original creation text, and the plot generation model generates the plot information of the chapter to be created according to the above outline.

[0090] Furthermore, relevant search information is extracted from the plot information, and the relevant search information is used to perform graph search in the second knowledge graph to obtain target key information, and then the text generation model can be used to generate the text of the chapter to be created based on the target key information. The target key information includes but is not limited to: character information, outline of the original creation text, writing techniques, etc.

[0091] According to the information processing method provided by the embodiment of the present disclosure, by extracting entities and entity relationships from the original creative text, a first knowledge graph can be generated based on the entities and entity relationships. Furthermore, in order to improve the efficiency of information retrieval and query, the entities in the first knowledge graph can be divided into entity communities, and according to the divided entity communities and the hierarchical structure of the entity communities, the first knowledge graph is updated to obtain a second knowledge graph. Furthermore, a graph search can be performed on the second knowledge graph based on the plot information of the chapter to be created to determine the target key information, and the main text of the chapter to be created can be generated based on the target key information and the plot information. In the present disclosure, by generating a knowledge graph from the original creative text and displaying it to the user, it can be convenient for the user to sort out the creative ideas, framework, narrative rhythm, character settings and other information to provide inspiration for their own creation. The generation of the main text of the article based on the knowledge graph can greatly enhance the text logic of the main text and improve the readability of the text.

[0092] Based on the above embodiments, the present disclosure can explain the process of chapter splitting. Figure 8 As shown, the chapter splitting process may include:

[0093] S801, split the original creative text into chapters according to the preset first matching rule to obtain the number of split chapters.

[0094] In some implementations, rules such as content continuity, plot development, and chapter length can be used as preset first matching rules, and the original creative text can be split into chapters based on the first matching rules to split the original creative text into chapter numbers, chapter titles, and chapter texts. The number of split chapters can be determined based on the chapter numbers.

[0095] S802, obtaining the total length of the original creative text, and determining a judgment threshold according to the total length of the text.

[0096] S803, in response to the number of chapters being less than the judgment threshold, the large model identifies chapter rules of the original creative text and constructs a second matching rule.

[0097] In some implementations, the correctness of the splitting can be ensured, and whether the first matching rule is successfully split can be determined based on the decision threshold. Optionally, the decision threshold can be determined based on the total length of the original creative text. For example, 0.01% of the total length of the text is used as the decision threshold.

[0098] Furthermore, the number of chapters is compared with the judgment threshold. If the number of chapters is less than the judgment threshold, the original creative text can be split again; if the number of chapters is greater than or equal to the judgment threshold, the chapter text is determined based on the split result to obtain the text block of each chapter.

[0099] In some implementations, in order to ensure the accuracy of the splitting and improve the splitting efficiency, the large model can be used to determine the second matching rule of the splitting, and the second matching rule can be used to split the original creative text into chapters. Optionally, the original creative text can be split from the middle position to obtain two text segments, and the text content of the target length can be extracted from the two text segments, and then the large model can identify the regularity of the chapter title of the text content of the target length, and generate the second matching rule according to the regularity of the chapter title.

[0100] For example, a text content of 10,000 words in length can be extracted from two text fragments, and the large model can be used to recognize the text content to obtain the chapter titles of the text content, and the regularity of the chapter titles can be determined to generate a second matching rule.

[0101] S804, based on the second matching rule, the original creative text is divided into chapters to obtain the chapter text corresponding to each chapter.

[0102] In some implementations, the chapter titles in the original creative text can be identified according to the second matching rule, and the original creative text can be split according to the identified chapter titles to obtain the chapter text corresponding to each chapter. According to the content of the text, the text block corresponding to the chapter can be determined.

[0103] In some implementations, after the chapter titles are split, the chapter titles split from the original creative text can be normalized and missing chapter titles can be searched to optimize the split chapter titles and obtain the final chapter titles of the original creative text.

[0104] According to the information processing method provided by the embodiment of the present disclosure, the original creative text is split into chapters by using the first matching rule, and when the splitting fails, the large model recognizes the chapter title pattern of the original creative text to generate a second matching rule, and the second matching rule is used to split the chapters to obtain the chapter text corresponding to each chapter, which can ensure the accuracy of chapter splitting and improve the efficiency of chapter title recognition.

[0105] like Fig. 9 The flowchart of chapter splitting is shown. For the original creative text, the first matching rule is used to match the original creative text with rules to split the original creative text into chapters, and the information such as the split chapter serial number, chapter title, and chapter text is obtained, and the number of chapters is determined according to the chapter serial number. By judging whether the number of chapters is less than the judgment threshold, when the number of chapters is less than the judgment threshold, the large model is called, and the large model recognizes the chapter titles of the original creative text, and extracts regularities from the chapter titles to construct the second matching rule.

[0106] Further, the second matching rule is used to match the original creative text to determine the chapter text corresponding to each chapter. During the splitting process, the chapter titles can also be checked for normalization and missing chapter titles can be searched to optimize the split chapter titles and obtain the final splitting result.

[0107] Corresponding to the information processing methods provided in the above-mentioned embodiments, an embodiment of the present disclosure further provides an information processing device. Since the information processing device provided in the embodiment of the present disclosure corresponds to the information processing methods provided in the above-mentioned embodiments, the implementation methods of the above-mentioned information processing methods are also applicable to the information processing device provided in the embodiment of the present disclosure and will not be described in detail in the following embodiments.

[0108] Fig.10 A schematic diagram of the structure of an information processing device provided in an embodiment of the present disclosure.

[0109] like Fig.10 As shown, the information processing device 1000 of the embodiment of the present disclosure includes a first generating module 1001 , an updating module 1002 and a second generating module 1003 .

[0110] The first generation module 1001 is used to extract entity relationships from the original creative text and generate a first knowledge graph;

[0111] An updating module 1002 is used to divide entity communities according to node attributes of nodes in the first knowledge graph, and update the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph;

[0112] The second generation module 1003 is used to obtain the plot information of the chapter to be created, and generate the text of the chapter to be created based on the plot information and the second knowledge graph.

[0113] In one embodiment of the present disclosure, the updating module 1002 is further used to: recursively perform community clustering on the nodes in the first knowledge graph according to the node attributes based on the hierarchical Leiden algorithm; terminate the community division in response to the number of the entity communities reaching a set community threshold, and obtain multiple entity communities.

[0114] In one embodiment of the present disclosure, the update module 1002 is further used to: determine the hierarchical structure of the entity community based on the entity relationship between communities; generate a summary of each entity community based on the big model; and update the first knowledge graph based on the summary of the entity community and the hierarchical structure to obtain a second knowledge graph.

[0115] In one embodiment of the present disclosure, the first generation module 1001 is also used to: extract key information from the chapter text of the original creative text; perform entity relationship extraction on the key information to obtain entity relationship triples; and generate the first knowledge graph based on the entity relationship triples.

[0116] In one embodiment of the present disclosure, the first generation module 1001 is further used to: generate an initial third knowledge graph based on the entity relationship triples; identify duplicate entities from the third knowledge graph, and merge and disambiguate the duplicate entities to obtain the first knowledge graph.

[0117] In one embodiment of the present disclosure, the first generation module 1001 is further used to: split the chapters of the original creative text to obtain text blocks corresponding to each chapter; extract the key information from the chapter text of the text block based on the large model, and the key information includes at least chapter summaries, character information, key plots, writing techniques and chapter outlines.

[0118] In one embodiment of the present disclosure, the second generation module 1003 is further used to: obtain user input information, identify plot points from the user input information, and determine the plot information of the chapter to be created based on the plot points.

[0119] In one embodiment of the present disclosure, the second generation module 1003 is further used to: extract the context outline from the original creative text; call a pre-trained plot generation model, and the plot generation model generates the plot information of the chapter to be created according to the context outline.

[0120] In one embodiment of the present disclosure, the second generation module 1003 is further used to: determine the end of the previous text according to the chapter to be created, and extract relevant retrieval information according to the end of the previous text and the plot information; perform graph retrieval on the second knowledge graph based on the relevant retrieval information to obtain target key information for creation; generate the main text of the chapter to be created based on the target key information and the plot information.

[0121] In one embodiment of the present disclosure, the second generation module 1003 is further used to: determine the chapter number and role identification of the chapter to be created based on the relevant search information; and match key information in different entity communities of the second knowledge graph based on the chapter number and the role identification to obtain the target key information.

[0122] In one embodiment of the present disclosure, the second generation module 1003 is further used to: call a pre-trained text generation model, and the text generation model generates the text of the chapter to be created according to the target key information and the plot information.

[0123] In one embodiment of the present disclosure, the first generation module 1001 is further used to: split the original creative text into chapters according to a preset first matching rule to obtain the number of split chapters; obtain the total text length of the original creative text, and determine a judgment threshold based on the total text length; in response to the number of chapters being less than the judgment threshold, use the large model to identify chapter rules for the original creative text and construct a second matching rule; based on the second matching rule, split the original creative text into chapters to obtain the chapter text corresponding to each chapter.

[0124] In one embodiment of the present disclosure, the first generation module 1001 is further used to: split the original creative text from the middle position to obtain two text segments, and extract the main text content of the target length from the two text segments; use the large model to identify the regularity of chapter titles on the text content of the target length, and generate the second matching rule based on the regularity of the chapter titles.

[0125] In one embodiment of the present disclosure, the device further includes: normalizing the chapter titles split from the original creative text and searching for missing chapter titles to optimize the split chapter titles to obtain the final chapter titles of the original creative text.

[0126] In one embodiment of the present disclosure, the device also includes: obtaining the newly added text and constructing a local knowledge graph of the newly added text; performing role matching on the local knowledge graph and the second knowledge graph, and obtaining the role matching result of the local knowledge graph, the role matching result including whether the role in the local knowledge graph is the same role in the second knowledge graph; according to the role matching result, associating the role of the local knowledge graph to the second knowledge graph, and synchronizing the chapter information of the newly added text to the second knowledge graph.

[0127] In one embodiment of the present disclosure, the device further includes: updating relevant information of the associated roles in the second knowledge graph according to the role matching result; and / or updating the novel outline corresponding to the second knowledge graph.

[0128] According to the information processing device provided by the embodiment of the present disclosure, by extracting entities and entity relationships from the original creative text, a first knowledge graph can be generated based on the entities and entity relationships. Furthermore, in order to improve the efficiency of information retrieval and query, the entities in the first knowledge graph can be divided into entity communities, and according to the divided entity communities and the hierarchical structure of the entity communities, the first knowledge graph is updated to obtain the second knowledge graph. Furthermore, the main text of the chapter to be created can be generated based on the plot information of the chapter to be created and the second knowledge graph to obtain the main text of the chapter to be created. In the present disclosure, by generating a knowledge graph from the original creative text and displaying it to the user, it can be convenient for the user to sort out creative ideas, frameworks, narrative rhythms, character settings and other information to provide inspiration for their own creation. The generation of the main text of the article based on the knowledge graph can greatly enhance the text logic of the main text and improve the readability of the text.

[0129] In the technical solution disclosed herein, the acquisition, storage and application of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0130] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0131] Fig.11A schematic block diagram of an example electronic device 1100 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0132] like Fig.11 As shown, the device 1100 includes a computing unit 1101, which can perform various appropriate actions and processes according to a computer program / instruction stored in a read-only memory (ROM) 1102 or a computer program / instruction loaded from a storage unit 1106 to a random access memory (RAM) 1103. In the RAM 1103, various programs and data required for the operation of the device 1100 can also be stored. The computing unit 1101, the ROM 1102, and the RAM 1103 are connected to each other via a bus 1104. An input / output (I / O) interface 1105 is also connected to the bus 1104.

[0133] A number of components in the device 1100 are connected to the I / O interface 1105, including: an input unit 1106 such as a keyboard, a mouse, etc.; an output unit 1107 such as various types of displays, speakers, etc.; a storage unit 1108 such as a disk, an optical disk, etc.; and a communication unit 1109 such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 1109 allows the device 1100 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0134] The computing unit 1101 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 1101 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 1101 performs the various methods and processes described above, such as information processing methods. For example, in some embodiments, the information processing method may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 1106. In some embodiments, part or all of the computer program / instructions may be loaded and / or installed on the device 1100 via the ROM 1102 and / or the communication unit 1109. When the computer program / instructions are loaded into the RAM 1103 and executed by the computing unit 1101, one or more steps of the information processing method described above may be performed. Alternatively, in other embodiments, the computing unit 1101 may be configured to execute the information processing method in any other appropriate manner (eg, by means of firmware).

[0135] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs / instructions that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0136] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0137] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0138] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0139] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), the Internet, and a blockchain network.

[0140] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs / instructions running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0141] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in the disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in the disclosure can be achieved, and this document does not limit them here.

[0142] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. An information processing method, wherein: The method comprises: Extract entity relationships from the original creative text to generate the first knowledge graph; Dividing entity communities according to node attributes of nodes in the first knowledge graph, and updating the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph; The plot information of the chapter to be created is obtained, and the main text of the chapter to be created is generated according to the plot information and the second knowledge graph.

2. The method according to claim 1, wherein: The dividing of entity communities according to node attributes of nodes in the first knowledge graph includes: Based on the hierarchical Leiden algorithm, recursively performing community clustering on the nodes in the first knowledge graph according to the node attributes; In response to the number of the entity communities reaching a set community threshold, the community division is terminated and a plurality of entity communities are acquired.

3. The method according to claim 1 or 2, wherein: The updating of the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph includes: Based on the entity relationships among the communities, determining a hierarchical structure of the entity communities; generating a summary of each entity community based on the macro model; Based on the summary of the entity community and the hierarchical structure, the first knowledge graph is updated to obtain a second knowledge graph.

4. The method according to claim 1, wherein: The extracting entity relationships from the original creative text to generate a first knowledge graph includes: extracting key information from the body of a chapter of said original creative text; Extracting entity relationships from the key information to obtain entity relationship triples; Generate the first knowledge graph based on the entity relationship triples.

5. The method according to claim 4, wherein: The generating the first knowledge graph according to the entity relationship triples includes: Generate an initial third knowledge graph according to the entity relationship triples; Duplicate entities are identified from the third knowledge graph, and the duplicate entities are merged and disambiguated to obtain the first knowledge graph.

6. The method according to claim 4, wherein: The extracting of key information from the chapter text of the original creative text includes: Splitting the chapters of the original creative text to obtain text blocks corresponding to each chapter; The key information is extracted from the chapter text of the text block based on the large model, and the key information at least includes chapter summary, character information, key plot, writing techniques and chapter outline.

7. The method according to claim 1, wherein: The step of obtaining the plot information of the chapter to be created includes: User input information is obtained, and plot points are identified from the user input information, and based on the plot points, plot information of the chapter to be created is determined.

8. The method according to claim 1, wherein: The step of obtaining the plot information of the chapter to be created includes: Extract the above outline from said original creative text; A pre-trained plot generation model is called, and the plot generation model generates the plot information of the chapter to be created according to the above outline.

9. The method according to any one of claims 1-2 or 4-8, wherein: The step of generating the text of the chapter to be created according to the plot information and the second knowledge graph includes: According to the chapter to be created, determining the end of the previous text, and extracting relevant search information according to the end of the previous text and the plot information; Performing graph retrieval on the second knowledge graph based on the relevant search information to obtain target key information for creation; Based on the target key information and the plot information, the main text of the chapter to be created is generated.

10. The method according to claim 9, wherein: The performing graph retrieval on the second knowledge graph based on the relevant search information to obtain target key information for creation includes: Determine the chapter number and role identification of the chapter to be created according to the relevant search information; According to the chapter number and the role identifier, key information is matched in different entity communities of the second knowledge graph to obtain the target key information.

11. The method according to claim 9, wherein: The step of generating the text of the chapter to be created based on the target key information and the plot information includes: A pre-trained text generation model is called, and the text generation model generates the text of the chapter to be created according to the target key information and the plot information.

12. The method according to claim 6, wherein: The chapters of the original creative text are split to obtain text blocks corresponding to each chapter, including: According to a preset first matching rule, the original creative text is divided into chapters to obtain the number of divided chapters; Obtaining the total length of the original creative text, and determining a judgment threshold according to the total length of the text; In response to the number of chapters being less than the judgment threshold, the large model identifies chapter regularities of the original creative text and constructs a second matching rule; Based on the second matching rule, the original creative text is divided into chapters to obtain the chapter text corresponding to each chapter.

13. The method according to claim 12, wherein: The large model identifies chapter rules of the original creative text and constructs a second matching rule, including: Splitting the original creative text from the middle position to obtain two text segments, and extracting text content of a target length from the two text segments; The large model identifies the regularity of chapter titles for the text content of the target length, and generates the second matching rule according to the regularity of chapter titles.

14. The method according to claim 12 or 13, wherein: The method further comprises: The chapter titles split out of the original creative text are normalized and missing chapter titles are searched to optimize the split chapter titles and obtain the final chapter titles of the original creative text.

15. The method according to any one of claims 1-2 or 4-8, wherein: The method further comprises: Obtaining the newly added text and constructing a local knowledge graph of the newly added text; Performing role matching on the local knowledge graph and the second knowledge graph, and obtaining a role matching result of the local knowledge graph, wherein the role matching result includes whether the role in the local knowledge graph is the same role in the second knowledge graph; According to the role matching result, the role of the local knowledge graph is associated with the second knowledge graph, and the chapter information of the newly added text is synchronized to the second knowledge graph.

16. The method according to claim 15, wherein: The method further comprises: According to the role matching result, updating relevant information of the associated roles in the second knowledge graph; and / or, Update the novel outline corresponding to the second knowledge graph.

17. An information processing device, wherein: The device comprises: A first generation module is used to extract entity relationships from the original creative text and generate a first knowledge graph; An updating module, configured to divide entity communities according to node attributes of nodes in the first knowledge graph, and update the first knowledge graph based on the divided entity communities and the hierarchical structure of the entity communities to obtain a second knowledge graph; The second generation module is used to obtain the plot information of the chapter to be created, and generate the text of the chapter to be created based on the plot information and the second knowledge graph.

18. The device according to claim 17, wherein: The update module is further used for: Based on the hierarchical Leiden algorithm, recursively performing community clustering on the nodes in the first knowledge graph according to the node attributes; In response to the number of the entity communities reaching a set community threshold, the community division is terminated and a plurality of entity communities are acquired.

19. The device according to claim 17 or 18, wherein: The update module is further used for: Based on the entity relationships among the communities, determining a hierarchical structure of the entity communities; generating a summary of each entity community based on the macro model; Based on the summary of the entity community and the hierarchical structure, the first knowledge graph is updated to obtain a second knowledge graph.

20. The device according to claim 17, wherein: The first generating module is further used for: extracting key information from the body of a chapter of said original creative text; Extracting entity relationships from the key information to obtain entity relationship triples; Generate the first knowledge graph based on the entity relationship triples.

21. The device according to claim 20, wherein: The first generating module is further used for: Generate an initial third knowledge graph according to the entity relationship triples; Duplicate entities are identified from the third knowledge graph, and the duplicate entities are merged and disambiguated to obtain the first knowledge graph.

22. The device according to claim 20, wherein: The first generating module is further used for: Splitting the chapters of the original creative text to obtain text blocks corresponding to each chapter; The key information is extracted from the chapter text of the text block based on the large model, and the key information at least includes chapter summary, character information, key plot, writing techniques and chapter outline.

23. The device according to claim 17, wherein: The second generating module is further used for: User input information is obtained, and plot points are identified from the user input information, and based on the plot points, plot information of the chapter to be created is determined.

24. The device according to claim 17, wherein: The second generating module is further used for: Extract the above outline from said original creative text; A pre-trained plot generation model is called, and the plot generation model generates the plot information of the chapter to be created according to the above outline.

25. The device according to any one of claims 17-18 or 20-24, wherein The second generating module is further used for: According to the chapter to be created, determining the end of the previous text, and extracting relevant search information according to the end of the previous text and the plot information; Performing graph retrieval on the second knowledge graph based on the relevant search information to obtain target key information for creation; Based on the target key information and the plot information, the main text of the chapter to be created is generated.

26. The device according to claim 25, wherein The second generating module is further used for: Determine the chapter number and role identification of the chapter to be created according to the relevant search information; According to the chapter number and the role identifier, key information is matched in different entity communities of the second knowledge graph to obtain the target key information.

27. The device according to claim 25, wherein: The second generating module is further used for: A pre-trained text generation model is called, and the text generation model generates the text of the chapter to be created according to the target key information and the plot information.

28. The device according to claim 22, wherein: The first generating module is further used for: According to a preset first matching rule, the original creative text is divided into chapters to obtain the number of divided chapters; Obtaining the total length of the original creative text, and determining a judgment threshold according to the total length of the text; In response to the number of chapters being less than the judgment threshold, the large model identifies chapter regularities of the original creative text and constructs a second matching rule; Based on the second matching rule, the original creative text is divided into chapters to obtain the chapter text corresponding to each chapter.

29. The device according to claim 28, wherein The first generating module is further used for: Splitting the original creative text from the middle position to obtain two text segments, and extracting text content of a target length from the two text segments; The large model identifies the regularity of chapter titles for the text content of the target length, and generates the second matching rule according to the regularity of chapter titles.

30. The device according to claim 28 or 29, wherein: The device also includes: The chapter titles split out of the original creative text are normalized and missing chapter titles are searched to optimize the split chapter titles and obtain the final chapter titles of the original creative text.

31. The device according to any one of claims 17-18 or 20-24, wherein The device also includes: Obtaining the newly added text and constructing a local knowledge graph of the newly added text; Performing role matching on the local knowledge graph and the second knowledge graph, and obtaining a role matching result of the local knowledge graph, wherein the role matching result includes whether the role in the local knowledge graph is the same role in the second knowledge graph; According to the role matching result, the role of the local knowledge graph is associated with the second knowledge graph, and the chapter information of the newly added text is synchronized to the second knowledge graph.

32. The device according to claim 31, wherein The device also includes: According to the role matching result, updating relevant information of the associated roles in the second knowledge graph; and / or, Update the novel outline corresponding to the second knowledge graph.

33. An electronic device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 16.

34. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-16.

35. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the method according to any one of claims 1 to 16 is implemented.

Images