Intelligent novel-script conversion method and system

Abstract

The invention relates to the technical field of natural language processing, in particular to a novel-script intelligent conversion method and system, and the method comprises the steps: obtaining an input novel text, and splitting the novel text to obtain divided volumes; processing each chapter to obtain a chapter mark and a chapter outline, and selecting roles appearing in each chapter to obtain an outgoing figure list; acquiring role labeling information based on the outline of each chapter and section, and acquiring roll-dividing outline information based on the outline of each chapter and section; combining the role labeling information and the roll division outline information corresponding to each roll division to obtain a combined set; distributing plot volume information and role distribution information corresponding to each set based on the set number request; merging based on the plot volume information and the role allocation information to obtain a set outline, and studying, judging and adjusting the content of the set outline; generating the script content corresponding to each set based on the set outline, and combining the script content to output a complete script of the novel text. The method and the device are used for improving the conversion effect from a large volume of novel to a script.

Description

Technical Field

[0001] The present application relates to the technical field of natural language processing, and more particularly to a novel-script intelligent conversion method and system. Background Art

[0002] Currently, with the continuous and healthy development of novels, more and more novels are chosen to be remade and adapted into film and television works, dramas, and short videos. The way of converting novels into scripts has also changed from manual conversion in the past to automated conversion auxiliary tools.

[0003] In the process of converting novels to scripts, important tasks include understanding and analyzing the plot of the novel, marking the characters, generating the script outline, etc. Although some current conversion tools can achieve a certain degree of auxiliary conversion functions, such as script generation based on generating large models, which usually gives the content of the novel chapters as input and lets the large model convert and rewrite the script, it has the following problems: First, current tools usually only support text processing at the episode level from partial chapters or single chapters of a novel to scripts, and cannot handle the content of very long texts or entire novels; at the same time, the format they generate is often not standardized and the structure is unclear, which brings additional editing tasks to screenwriters; the number of episodes, the number of characters and the amount of plot in each episode of the scripts generated by existing tools are all uncontrollable, and there is often a problem of unbalanced content volume in each episode. Summary of the invention

[0004] In order to improve the conversion effect of large-volume novels into scripts, the present application provides a novel-script intelligent conversion method and system.

[0005] In the first aspect, the present application provides a novel-script intelligent conversion method, which adopts the following technical solution: A novel-script intelligent conversion method comprises the following steps: Obtaining an input novel text, splitting the novel text into sub-volumes, wherein the sub-volumes contain at least one chapter; Marking each of the chapters to obtain a chapter mark, reasoning each of the chapters to obtain a chapter outline corresponding to each of the chapters, and selecting characters appearing in each of the chapters to obtain a character list; Based on the chapter outlines, the character information of each character in the character list in the sub-volume is marked, and based on the chapter outlines, characteristic content analysis is performed to obtain the sub-volume outline information corresponding to each sub-volume; Merging the character annotation information corresponding to each of the sub-volumes and the sub-volume outline information to obtain a merged set, and performing multi-source content processing on the merged set to respectively generate complete character cards and full text outlines for each of the novel texts; Obtain an episode request, and assign plot volume information and role assignment information corresponding to each episode based on the episode request combined with the complete role card and the full text outline; Based on the plot volume information and the role allocation information, the corresponding multiple chapters are merged to obtain a collection outline, and the content of the collection outline is analyzed and adjusted according to a preset text rule; The script content corresponding to each episode is generated based on the collection outline and merged to output a complete script of the novel text.

[0006] In some embodiments, based on each chapter outline, the character information of each character in the character list in the sub-volume is annotated, and based on each chapter outline, feature content analysis is performed to obtain the sub-volume outline information corresponding to each sub-volume, including the following steps: Obtaining the outline of each chapter and the list of characters appearing in the chapter; Select a single character in the character list, and annotate the relevant plot corresponding to the single character in each chapter outline based on the prompt word project to obtain character annotation information; Based on the prompt word project, each chapter outline is analyzed to screen the plots corresponding to the individual characters in each chapter to retain important plots and eliminate irrelevant plots, and the development logic and logical relationships in the chapter outlines are connected in series to obtain the volume outline information.

[0007] In some embodiments, the character annotation information corresponding to each of the sub-volumes and the sub-volume outline information are merged to obtain a merged set, and multi-source content processing is performed on the merged set to generate complete character cards and full text outlines for each of the novel texts, respectively, including the following steps: Acquire the role annotation information corresponding to each of the sub-volumes, and merge several role annotation information corresponding to the same role to obtain a role merge set; Deleting the repeated role labeling information in the role merge set and sorting the roles according to a preset sorting rule to obtain the complete role card; Obtaining the sub-volume outline information corresponding to each sub-volume and merging them to obtain a full-text outline merged set; The duplicate contents in the merged set of full-text outlines are removed, the missing contents are supplemented or replaced, and the content consistency and logic of the merged set of full-text outlines are detected to obtain the full-text outline.

[0008] In some embodiments, obtaining an episode request, and assigning plot volume information and role assignment information corresponding to each episode based on the episode request combined with the complete role card and the full text outline, includes the following steps: Calculating a total volume score based on the full text outline, the total volume score consisting of a first volume score corresponding to the dialogue content, a second volume score corresponding to the behavior content, and a third volume score corresponding to the emotional content; The total volume score is evenly divided into a collective volume score corresponding to each episode, and the full text outline is divided into a plurality of individual volume ranges based on the collective volume score, and the plot volume information corresponding to each episode is obtained based on the individual volume ranges; Based on the individual group mass range, select the set roles that contribute the group mass points in the complete role card, and assign corresponding set plots to each set role according to the group mass points contributed by each set role to obtain the role assignment information; Among them, each episode character is divided into weakly related characters, main characters and special characters based on the appearance frequency, number of chapters and chapter position corresponding to the episode characters, and there are at least two main characters and / or special characters in each episode.

[0009] In some embodiments, merging corresponding multiple chapters based on the plot volume information and the role allocation information to obtain an outline includes the following steps: Acquire chapter information that matches the plot volume information, and merge a corresponding number of continuous chapters or skipped chapters based on the chapter information to obtain a script chapter group; The plot corresponding to the role matching the role allocation information in each chapter outline corresponding to the script chapter group is obtained to generate the collection outline.

[0010] In some embodiments, judging and adjusting the content of the collection outline according to preset text rules includes the following steps: Obtaining element information and node information in each of the outlines, wherein the element information includes action elements, character elements, scene elements, environmental feature elements, and dialogue elements, and the node information includes a starting point, a supporting point, a turning point, and a joining point; When each of the outlines satisfies the preset text rules, each of the outlines at least contains character elements, scene elements, environmental feature elements, and contains the action elements or dialogue elements; Adjust the order and position of the contents in the outlines so that each outline satisfies the preset text rules. Specifically, the beginning and ending positions in each outline include at least one of the starting point, the supporting point, the turning point and the joining point, and the node information content at the beginning and the end of adjacent outlines is consistent with coherence.

[0011] In some of the embodiments, obtaining the node information in each of the clusters includes the following steps: When new character elements, scenario elements, environmental characteristic elements, role relationships, and forces appear in the collection outline, the starting point is added to the corresponding position in the collection outline; Starting from the starting point, after the starting point, the support point is added to the corresponding position in the set outline every time the measurement length passes until the character elements, scene elements, environmental feature elements, role relationships, and forces in the set outline change, wherein the measurement length includes the word length and the collective measurement length; When an explicit turning point and / or an implicit turning point occurs in the collection outline, the turning point is added at the corresponding position in the collection outline, wherein the explicit turning point is characterized by a contradiction between the character elements, the character relationships, and the forces corresponding to the same character, and the implicit turning point is characterized by a contradiction between the emotional information corresponding to the action elements and the dialogue elements determined based on the prompt word engineering and / or a contradiction between the plot development and the logical development in the chapter outline; When the outline shows changes in character elements, scene elements, environmental feature elements, role relationships, and forces after the starting point, the convergence point is added at the corresponding position before the change occurs.

[0012] In some embodiments, the node information contents at the beginning and the end of adjacent sets of the outlines are consistent with each other, comprising the following steps: Obtaining the compactness corresponding to each set based on the collective quantity score corresponding to each set outline, and selecting a repetition mode or a continuous mode based on the compactness; Obtain the set outlines corresponding to two adjacent sets. When the adjacent set outlines meet the coherence, The repetition mode is characterized in that in the adjacent collection outlines, the beginning of the collection outline of the next collection repeats the content of the end of the collection outline of the previous collection; The continuity is characterized by that in adjacent episodes, if the scene elements and the environmental feature elements are the same, then based on the prompt word engineering, the plot coherence and logical coherence of the character elements, action elements, and dialogue elements at the beginning of the episode of the next episode and the character elements, action elements, and dialogue elements at the end of the episode of the previous episode are guaranteed; if the scene elements and the environmental feature elements are different, then it is guaranteed that the end of the episode of the previous episode corresponds to the conjunction point and the beginning of the episode of the next episode corresponds to the starting point.

[0013] In some embodiments, generating script content corresponding to each episode based on the episode outline includes the following steps: Adding corresponding content to a preset script template based on the element information corresponding to the collection outline, wherein the script template includes a background section, a character section, and a plot section; The scene elements and the environmental feature elements are added to the background section, the character elements are added to the character section, and the action elements and dialogue elements are added to the plot section.

[0014] In the second aspect, the present application provides a novel-script intelligent conversion system, which adopts the following technical solutions: A novel-script intelligent conversion system, comprising: A text volume segmentation module, used for obtaining an input novel text, and splitting the novel text to obtain volumes, wherein the volumes contain at least one chapter; A chapter marking module, used for marking each of the chapters to obtain chapter marks, and performing reasoning on each of the chapters to obtain a chapter outline corresponding to each of the chapters; A character selection module, used for selecting characters appearing in each of the chapters to obtain a list of characters appearing; A role labeling module, used for labeling the role information of each character in the character list in the sub-volume based on each chapter outline; An outline analysis module, used for performing characteristic content analysis based on each of the chapter outlines to obtain the volume outline information corresponding to each of the volumes; A role merging module, used for merging the role annotation information corresponding to each of the sub-volumes to obtain a merged set, and performing multi-source content processing on the merged set to generate a complete role card for each of the novel texts; An outline merging module, used for merging the sub-volume outline information corresponding to each sub-volume to obtain a merged set, and performing multi-source content processing on the merged set to generate a full-text outline for each novel text; A script conversion module is used to obtain an episode request, and to allocate plot volume information and role allocation information corresponding to each episode based on the episode request combined with the complete role card and the full text outline; it is also used to merge corresponding multiple chapters based on the plot volume information and the role allocation information to obtain an episode outline, and to judge and adjust the content of the episode outline according to preset text rules; based on the episode outline, the script content corresponding to each episode is generated and merged to output a complete script of the novel text.

[0015] The technical solution provided by the embodiments of the present application has the following technical effects: In particular, the novel is split into volumes and asynchronous parallel processing of character annotation and outline generation is implemented. Finally, the results of analysis and annotation of multiple volumes are merged, which can effectively process super-long novels and solve the limitations of current technology in processing super-long texts. The script generation module can refer to the content of the previous episode when generating each episode script to ensure the coherence and consistency of the plot between episodes. At the same time, it verifies the format, logic, and structure when converting each episode script to reduce the abnormal script conversion situation. Determine and control the script volume and characters of each episode to avoid unbalanced content. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a block diagram of a building sunlight simulation analysis method provided in this embodiment.

[0017] Figure 2 It is a frame diagram of obtaining surrounding building information of a building to be analyzed based on geographic location information provided by an embodiment of the present application. DETAILED DESCRIPTION

[0018] To more clearly understand the purpose, technical solutions and advantages of the present application, the present application is described and illustrated below in conjunction with the accompanying drawings and embodiments. However, it should be understood by those of ordinary skill in the art that the present application can be implemented without these details. In some cases, in order to avoid unnecessary descriptions that make various aspects of the present application obscure, well-known methods, processes, systems, components and / or circuits that have been described at a higher level will not be described in detail. For those of ordinary skill in the art, it is obvious that various changes can be made to the embodiments disclosed in the present application, and without departing from the principles and scope of the present application, the general principles defined in the present application can be applied to other embodiments and application scenarios. Therefore, the present application is not limited to the embodiments shown, but conforms to the broadest scope consistent with the scope claimed for protection of the present application.

[0019] It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention. In addition, the technical features involved in each embodiment of the present invention described below can be combined with each other as long as there is no conflict between them.

[0020] In the description of this application, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed", etc. are understood to exclude the number itself, and "above", "below", "within", etc. are understood to include the number itself. If there is a description of "first" or "second", it is only used to distinguish the technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0021] In the description of the present application, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples.

[0022] The embodiment of the present application discloses a novel-script intelligent conversion method.

[0023] like Figure 1 As shown, a novel-script intelligent conversion method comprises the following steps: S100, obtaining an input novel text, splitting the novel text into sub-volumes, each sub-volume containing at least one chapter.

[0024] First, in the initial stage, the received super-long novel text needs to be split. Splitting means splitting the full text into several sub-contents in a way that a single chapter is the smallest unit. According to different situations, different numbers of chapters can be selected as sub-volumes for splitting. In the embodiment of the present application, the sub-volume includes chapters 1-100, and the number of sub-volumes can be the same or different.

[0025] S200, marking each chapter to obtain a chapter mark, inferring each chapter to obtain a chapter outline corresponding to each chapter, and selecting a character appearing in each chapter to obtain a character list.

[0026] Then we move on to the chapter analysis phase, marking the specific important content in each chapter to obtain the chapter number and corresponding chapter outline of each chapter.

[0027] At the same time, the character selection phase is carried out, and the characters appearing in each chapter are marked based on the prompt word project to obtain a list of characters appearing.

[0028] S300, based on the outline of each chapter, character marking information of each character in the character list in the sub-volume is analyzed, and based on the outline of each chapter, feature content analysis is performed to obtain sub-volume outline information corresponding to each sub-volume.

[0029] Based on the prompt word project, the characteristic parts in each volume are annotated to obtain the character annotation. The character annotation is characterized by annotating the plot, performance and other related plot contents corresponding to the appearing characters.

[0030] At the same time, the annotation work is carried out in parallel, that is, the annotation work on the character content in each volume is carried out simultaneously.

[0031] Based on the prompt word project, content analysis and reasoning are performed on the characteristic parts of each volume to obtain the corresponding information of the volume outline. The volume outline information is represented by the summary of the full text content reflected by the content of each volume.

[0032] S400, merging the character labeling information and the volume outline information corresponding to each volume to obtain a merged set, and performing multi-source content processing on the merged set to generate complete character cards and full text outlines for each novel text respectively.

[0033] After asynchronous parallel annotation and analysis of the volumes, the character annotation information and volume outline information of each volume are merged, so that the character annotation information and volume outline information corresponding to several split contents are combined as a whole to obtain the annotation and analysis results of the full text of the novel, including the complete character information cards and full text outline corresponding to the full text of the novel.

[0034] The above-mentioned prompt word engineering is characterized as a method that can obtain a certain result without adjusting the parameters or weights of the large model. It can guide the system model to generate more accurate and targeted output text that is more in line with the prompt words through the input prompt words that are carefully designed and adjusted in the early stage. Accurate and targeted prompt words are provided by adding explanations, examples, thought chain reasoning, adding prompt templates, adding dialogue history, etc. The system performs reasoning analysis and annotation on the content related to the prompt words in the content of the volume or chapter based on the prompt words.

[0035] S500, obtaining an episode request, and assigning plot volume information and role allocation information corresponding to each episode based on the episode request combined with the complete role card and the full text outline.

[0036] The episode request is a user-selected or system-default request, which is characterized by a request to convert the novel text into a script containing a certain number of episodes, and the plot volume and appearance characters of each episode are allocated according to the episode request and the corresponding complete role cards and full text outline.

[0037] S600, based on the plot volume information and the role allocation information, the corresponding multiple chapters are merged to obtain a collection outline, and the content of each collection outline is analyzed and adjusted according to the preset text rules.

[0038] After obtaining the preliminary plot volume and characters of each episode, the novel chapters corresponding to the plot and characters required for the episode are merged to obtain the plot outline for the episode, and the content of each episode outline is judged and adjusted through the set text rules. The text rules are represented by the evaluation rules for the format, logic, plot coherence and other contents of the script.

[0039] S700, generating script content corresponding to each episode based on the collection outline and merging them to output a complete script of the novel text.

[0040] The script content of each episode is generated based on the adjusted outline of each episode, and the script content of all episodes is merged to output the complete script content of the novel.

[0041] Through the above steps, the novel is split into volumes, and asynchronous parallel processing of character labeling and outline generation is realized. Finally, the results of analysis and labeling of multiple volumes are merged. This can effectively handle very long novels and solve the limitations of current technology in processing very long texts. The script generation module can refer to the content of the previous episode when generating each episode of the script to ensure the continuity and consistency of the plot between episodes. At the same time, the format, logic and structure are verified when each episode is converted to reduce abnormal script conversion. The script volume and characters of each episode are judged and controlled to avoid the problem of unbalanced content in the episodes.

[0042] In other embodiments, based on the outline of each chapter, the character information of each character in the character list in the sub-volume is annotated, and based on the outline of each chapter, characteristic content analysis is performed to obtain the sub-volume outline information corresponding to each sub-volume, including the following steps: S310, obtaining the outline of each chapter and the list of characters appearing in the chapter.

[0043] Based on the prompt word project, important paragraphs are selected in each chapter, such as the beginning, the end, the middle climax paragraph, the paragraphs with multiple characters, etc., to obtain the main plot overview. The content of several main plot overviews is cleaned and the logical order is adjusted, and then the paragraphs are integrated. The integrated result is used as the chapter outline of the chapter.

[0044] The character information of the characters corresponding to each chapter is integrated to obtain a character list, which contains the names, appearance frequencies, appearance locations, etc. of all the characters.

[0045] S320, selecting a single character in the character list, and annotating the relevant plots corresponding to the single character in each chapter outline based on the prompt word project to obtain character annotation information.

[0046] Receive the chapter outlines of each chapter and the overall character list, select individual characters in the character list in order of appearance, and annotate the content related to the character information in the chapter outlines corresponding to the sub-volume based on the prompt word project. The annotated plot includes the character's performance, personality, growth trajectory, behavior, etc. in each chapter, and integrate the character annotation information into the character card corresponding to the sub-volume content.

[0047] Specifically, carefully designed prompts are used to guide the model to understand what the character did and what results the character achieved in each volume or chapter, what impact and contribution their behavior and results had on the overall plot, and what objects they obtained. The above-mentioned analysis and reasoning are then used to label the character's corresponding performance, personality, growth trajectory, and other content.

[0048] S330, analyzing each chapter outline based on the prompt word project to screen the plots corresponding to individual characters in each chapter to retain important plots and eliminate irrelevant plots, and connecting the development logic and logical relationship in each chapter outline to obtain the volume outline information.

[0049] Analyze the outlines of each chapter corresponding to the volume, retain the relevant plots of the characters, delete the plots used for description, explanation, modification, etc., and analyze the chapter outlines corresponding to each chapter according to the development logic of the behaviors and actions corresponding to the plots, as well as the analysis of the writing logic such as flashbacks, supplements, and flashbacks to obtain the volume outline information.

[0050] Specifically, the chapter outline and each character are analyzed through the prompt word project to determine the corresponding relationship between each character and the plot; the flashback plot, fantasy plot and parts unrelated to the main story are sorted out from the order of plot development; the corresponding perspective of the main characters is understood, and the plot development and character development are combined with the perspective.

[0051] In other embodiments, the character labeling information and the volume outline information corresponding to each volume are merged to obtain a merged set, and the merged set is processed with multi-source content to generate complete character cards and full text outlines for each novel text, including the following steps: S410, obtaining the role annotation information corresponding to each volume, and merging several role annotation information corresponding to the same role to obtain a role merge set.

[0052] The character annotation information of the same character in each sub-volume is merged to generate a character merge set corresponding to the character, which contains all the annotated plot content related to the character.

[0053] S420, deleting repeated role labeling information in the role merging set and sorting the roles according to a preset sorting rule to obtain a complete role card.

[0054] The character annotations that appear repeatedly and represent the same plot content are deleted in the character merging set to obtain the character information cards corresponding to each character. The character information cards contain the development, personality changes, contributions, etc. of the character in the full text of the novel.

[0055] S430, obtaining the sub-volume outline information corresponding to each sub-volume and merging them to obtain a full-text outline merged set.

[0056] The full text outline analysis information of the independent analysis of each volume is merged to obtain a merged set of full text outlines corresponding to the full text of the novel.

[0057] S440, de-duplicate the duplicate contents in the merged set of full-text outlines, interpolate or replace the missing contents, and detect the content consistency and logic of the merged set of full-text outlines to obtain the full-text outline.

[0058] For the merged set of full-text outlines, duplicate content with the same meaning or the same plot description is removed, and the missing parts between the plot development and the logical connection are interpolated or replaced by combining prompt words and context analysis algorithms; finally, it is determined whether there is a plot conflict or logic conflict in the content of the merged set of full-text outlines after deduplication and interpolation or replacement of missing content. If so, the plot with a high probability of being correct is obtained according to the context of each chapter outline and the abnormal plot or logic is replaced to ensure content consistency and logical correctness. At the same time, when it is determined that the conflicting content is accurate according to the context of each chapter outline, the conflicting content is retained and the logical order is adjusted according to the chapter position where the full-text outline analysis information appears.

[0059] In some other embodiments, obtaining an episode request, and assigning plot volume information and role assignment information corresponding to each episode based on the episode request combined with complete role cards and the full text outline, includes the following steps: S510, calculating a total volume score based on the full text outline, the total volume score being composed of a first volume score corresponding to the dialogue content, a second volume score corresponding to the behavior content, and a third volume score corresponding to the emotion content.

[0060] The total volume score is used to evaluate the overall content volume of a novel. The evaluation criteria specifically combine the amount of dialogue, the amount of behavior, and the complex relationship of emotions, and each evaluation factor is independently scored.

[0061] In general scenarios, the full text outline is the essence of the entire novel. The more dialogues there are, the more dialogues there are in the novel text, and the greater the volume score of the dialogues; when the novel corresponds to martial arts, cultivation, war and other types, it must contain a lot of fighting, battles and other actions. So when performing the script, the more actions there are, the larger the plot volume needs to be occupied, so the corresponding volume score will be larger; when the novel corresponds to romance, family, youth and other types, the emotions between characters are often more complex and changeable. So when generating the corresponding interpretation script, the emotional conversion and emotional acquisition between characters need to occupy more plot volume, so the more complex the emotional relationship, the larger the corresponding volume score will be.

[0062] By calculating and adding up the volume scores in multiple dimensions, the total volume score corresponding to the entire full text outline can be obtained. The larger the total volume score of the full text outline, the more script space is required for the plot to unfold.

[0063] S520, the total volume score is evenly divided into collective volume scores corresponding to each episode, and the full text outline is divided into a number of individual volume ranges based on the collective volume scores, and the plot volume information corresponding to each episode is obtained based on the individual volume ranges.

[0064] The total volume score is evenly divided into the number that meets the episode request, and each episode is guaranteed to have the same total volume score, so that each episode can maintain the same compactness.

[0065] Each episode is averaged in the full text outline through the collective volume to obtain a certain single-volume mass range, that is, the full text outline is split into content ranges corresponding to several episodes, and the plot volume information corresponding to each episode is obtained according to the single-volume mass range.

[0066] S530, selecting the episode roles that contribute to the collective points in the complete role cards based on the individual episode mass range, and assigning corresponding episode plots to each episode role according to the collective points contributed by each episode role to obtain role assignment information.

[0067] As each character's dialogue, actions, emotions, etc. account for different proportions in each episode, the volume score they contribute in each episode is different. Therefore, it is necessary to obtain the corresponding required characters and the plot that each character needs to perform in the episode from the complete character cards of the full text of the novel based on the volume range corresponding to each episode.

[0068] In this way, based on the collective volume requirements of each episode, we can determine the characters that need to appear in the episode and the plot that each character needs to go through, such as the lines that need to be said, the actions that need to be performed, and the emotional output that needs to be made.

[0069] S540, dividing the characters in each episode into weakly associated characters, main characters and special characters based on the appearance frequency, number of chapters and chapter positions corresponding to the characters in the episode, and there are at least two main characters and / or special characters in each episode.

[0070] At the same time, each character is divided into weakly related characters, main characters and special characters according to the different volumes they occupy in the plot of the novel or script and their necessity for advancing the plot.

[0071] First, weakly related characters are screened by the frequency of character appearances. Weakly related characters are characterized as characters that are irrelevant or not very relevant to the entire novel, such as minor characters and irrelevant characters. If a character's appearance frequency is low, it is considered that the character is irrelevant or unimportant to the plot. In the embodiment of the present application, weakly related characters generally do not occupy any collective quantity points.

[0072] The remaining characters in the list are important characters that have a great connection with the novel.

[0073] Determine the number of chapters and the chapter positions where each character appears. Generally, the main characters include decent characters and villains. In a complete novel, decent characters usually appear throughout the entire text of the novel or in the early part of the novel, and appear in more chapters. Villain characters usually appear throughout the entire text of the novel or in the later part of the novel and continue to the end of the novel, and appear in more chapters. In this way, the main characters can be extracted by appearing in more chapters at the chapter positions that meet the preset requirements.

[0074] Special characters are generally characterized as those who play a special role in promoting the plot or have a special influence on the characters, such as the hermit masters in the immortal cultivation novels, the informants or witnesses in the mystery novels, the undercover agents in the police and gangster novels, etc. They usually do not appear at the beginning or early position of the novel, but play a role in reversing the plot or pushing it to a climax in the middle or late position of the novel. Then the number of chapters in which such characters appear is generally not large, and the positions in which they appear are concentrated in the middle or late positions of the full text of the novel. Therefore, special characters can be extracted through the corresponding number of chapters and chapter positions.

[0075] In order to make each episode's script complete, there must be at least two main characters or two main characters or one main character and one special character in each episode.

[0076] Specifically, special characters generally play an important role in the climax and turning points of the plot. Therefore, when calculating the collective scores corresponding to the characters, different calculation weights can be set for special characters and main characters respectively. When the volume parameters such as dialogue, behavior, and emotion of special characters are consistent or similar to those of the main characters, a higher collective score can be calculated for the special characters.

[0077] In some other embodiments, merging corresponding multiple chapters based on the plot volume information and the role allocation information to obtain an outline includes the following steps: S610, obtaining chapter information that matches the scenario volume information, and merging a corresponding number of continuous chapters or skipped chapters based on the chapter information to obtain a script chapter group.

[0078] After obtaining the plot volume information corresponding to each episode, multiple required chapters are merged into a script chapter group corresponding to each episode according to the plot volume required for each episode and the chapter outline of each chapter in the novel text.

[0079] The number of chapters in the script chapter group corresponding to each episode is generally different, and several chapters in a script chapter group can be continuous or discontinuous. Since novels may use flashbacks, flashbacks, supplementary narrations and other writing techniques when writing, and the logic of script interpretation may be different from the logic of novel writing, when converting the flashback logic of the novel into the sequential logic required for an episode of the script, one or some subsequent chapters can be placed in script chapter groups at other locations.

[0080] At the same time, each chapter is divided and distributed into multiple different script chapter groups according to the specific chapter outline.

[0081] S620, obtaining the plots corresponding to the roles that match the role allocation information in the outlines of each chapter corresponding to the script chapter group to generate a collection outline.

[0082] The roles corresponding to the role assignment information and the plot content related to the roles, such as actions, dialogues, emotions, role relationships, etc., are extracted from the chapter outlines corresponding to each chapter in the script chapter group, and combined into the episode outline corresponding to the episode.

[0083] In other embodiments, judging and adjusting the content of the collection outline according to preset text rules includes the following steps: S630, obtaining element information and node information in each set of outlines.

[0084] Element information includes action elements, character elements, scene elements, environmental feature elements, and dialogue elements.

[0085] Based on the prompt word project, the characteristic elements and content nodes in each set of outlines are obtained. The characteristic elements include the characters appearing, the content of the dialogue, the actions, the scenes, the environmental information, etc.

[0086] The above elements are together the script elements required for a script, and serve as the constituent elements of the background construction, actor lines, actor actions, narration and other contents in the script presentation.

[0087] Node information includes starting point, supporting point, turning point and junction point.

[0088] Articles, novels, scripts and other artistic forms of presentation all contain a complete chain of introduction, development, turn and conclusion. The introduction represents the beginning of a new plot or new content, the development represents the continuation and development of the content, the turn represents the turning point, reversal and climax of the plot or content, and the conclusion represents the summary of the plot or content.

[0089] According to the scenes, characters' behavior language, emotional output, etc. analyzed in the novel text, combined with the prompt word project, corresponding nodes can be added to the corresponding position content in the novel text.

[0090] Because novel texts contain more words and have more complete stories, there are often multiple nodes corresponding to the beginning, development, transition, and conclusion in a novel.

[0091] S640, when each set of outlines meets the preset text rules, each set of outlines at least contains character elements, scene elements, environmental feature elements, and contains action elements or dialogue elements.

[0092] In order to obtain a reasonable and satisfactory script for each episode, the corresponding outline of each episode needs to meet the text rules of the plot elements. That is, each episode must contain characters, scenes and environment content, which are necessary elements in the interpretation of the script. As for action or dialogue, at least one of them may exist according to the actual interpretation situation or the actual script content. For example, if an episode depicts the inner dynamics of a character or a fighting scene, only action elements may exist.

[0093] S650, adjusting the order and position of the contents in the collection outlines so that each collection outline meets the preset text rules. Specifically, the beginning and ending positions in each collection outline at least include one of the starting point, the supporting point, the turning point and the converging point, and the beginning and ending node information contents in adjacent collection outlines are consistent.

[0094] Each episode's script must also meet the rules of reasonable and complete text, that is, the beginning and end of each episode's outline must correspond to at least one node information, because the nodes of introduction, development, turning point and conclusion all correspond to important characteristic positions of the script interpretation or plot development. If the key positions at the beginning and end of an episode do not correspond to important nodes of the plot, then the script will be incomplete and rather bland.

[0095] At the same time, the content of two adjacent episodes must be consistent and coherent, that is, the content of the end of the previous episode and the beginning of the next episode must be consistent. Therefore, when analyzing and adjusting the content of each episode, the content of the previous episode will be referred to. In this way, in the process of generating scripts episode by episode in this application, the occurrence of incoherence problems such as loss of plot due to broken chapters and deviation of plot logic can be avoided.

[0096] The beginning position of each outline may be the first 100 to 300 words of the outline, and the ending position may be the last 100 to 300 words of the outline.

[0097] In some other embodiments, obtaining node information in each set of outlines includes the following steps: S631, when new character elements, situational elements, environmental characteristic elements, role relationships, and forces appear in the collection outline, add a starting point at the corresponding position in the collection outline.

[0098] The starting point usually represents the beginning of a new plot. The scenes, environments, characters, and other contents between different plots often change. For example, if the first act takes place in the classroom and the second act changes to home, it often represents the emergence of a new plot. Therefore, a starting point needs to be added to this type of position.

[0099] At the same time, when new characters or new forces appear, new plots may also appear, so starting points can also be added at the corresponding locations.

[0100] S632, starting from the starting point, after each measurement length, a supporting point is added to the corresponding position in the outline until the character elements, scene elements, environmental feature elements, role relationships, and forces in the outline change.

[0101] The measurement length includes word length and collective length.

[0102] The support point represents that the content at the corresponding position is a further description and advancement of the content at the starting point. Therefore, the overall environment, scene, characters, forces, etc. of the support point should remain the same or similar to those of the starting point. Therefore, starting from the starting point, when the content is advanced, after a preset length, a support point information is added when the overall environment, scene, etc. have not changed, until the above content changes to enter a new plot or a subsequent turning point plot.

[0103] The measurement length can be added after each number of words according to the actual situation, or after each collective measurement of a score.

[0104] S633, when an explicit feature change and / or an implicit turning point change occurs in the set outline, a turning point is added at the corresponding position in the set outline.

[0105] The explicit characteristic transition is manifested as conflicts between the character elements, role relationships, and forces corresponding to the same role.

[0106] When there are obvious changes in characters, relationships between characters, or forces in the episode outline, it is considered that the plot has turned, so a turning point is added at the corresponding position. For example, A and B go from being lovers to being enemies, or A changes from force a to force b.

[0107] The implicit feature transition is characterized by the contradiction between the emotional information corresponding to the action elements and dialogue elements and / or the contradiction between the plot development and logical development in the chapter outline based on the prompt word engineering.

[0108] When the prompt word project determines that there are contradictory action emotions, dialogue emotions, or contradictory plot developments and logical developments in the episode outline, such as when there are contradictions in the dialogue content and emotions before and after the characteristic transition words "but" and "turn", or when the action changes from a hug to a fight, or when character a changes from helping b to contradictory attacking b, the character has a story turning point or reversal, and a turning point is added at the corresponding position.

[0109] S634, when changes in character elements, situation elements, environmental characteristics, role relationships, and forces appear after the starting point of the collection outline, a junction point is added to the corresponding position before the change occurs.

[0110] When a plot outline shows changes in the above-mentioned evaluation elements after the starting point, it is considered the end of one scene and the beginning of the next scene. At this time, a conjunction point is added to the position before the plot elements change.

[0111] In some other embodiments, the beginning and ending node information in adjacent clusters are consistent with coherence, including the following steps: S651, obtaining the compactness corresponding to each set based on the collective score corresponding to each set outline, and selecting a repetition mode or a continuous mode based on the compactness.

[0112] In order to maintain the coherence and consistency of the development of each episode, it is necessary to determine the coherence between the beginning and the end of the adjacent episode outlines. First, the compactness of each episode is obtained through the collective score of the episode outline.

[0113] The larger the collective score of an episode outline, the more compact the plot of the episode is, because it may contain more dialogues, actions, etc.

[0114] Get the set outlines corresponding to two adjacent sets, when the adjacent set outlines meet the coherence: S652, the repetition mode is characterized in that in adjacent collection outlines, the beginning of the collection outline of the next collection repeats the ending content of the previous collection outline.

[0115] If the compactness of the episode outline is low, which means that there is not much content in each episode, then in order to maintain the coherence and consistency between adjacent episodes, the first 100 words to 300 words at the end of the previous episode can be repeated and added to the beginning of the next episode outline, so that the next episode outline can continue to be generated based on the previous episode. This can simply and effectively avoid the break or loss of the plot between the two episodes.

[0116] S653, the continuity mode is characterized by that in adjacent episodes, if the scene elements and environmental feature elements are the same, then based on the prompt word engineering, the plot coherence and logical coherence of the character elements, action elements, and dialogue elements at the beginning of the episode of the next episode and the character elements, action elements, and dialogue elements at the end of the episode of the previous episode are guaranteed; if the scene elements and environmental feature elements are different, then it is guaranteed that the end of the episode of the previous episode corresponds to the joining point and the opening of the episode of the next episode corresponds to the starting point.

[0117] If the episode outline is compact, it means that there is more content in each episode. In order to avoid including irrelevant and repeated plot content in each episode, the repetition method is not suitable.

[0118] There are two specific forms of continuous mode: When the scene elements and environmental elements in two adjacent episodes have not changed, it means that the ending and beginning parts of the two episodes are in the same plot scene. In this case, it is necessary to use prompt word engineering to judge and select the dialogues and behaviors that are coherent as the end of the previous episode and the beginning of the next episode, respectively, to ensure that the dialogue or action at the beginning of the next episode can match the dialogue or action at the end of the previous episode, so as to ensure the continuity of the content.

[0119] When the scene elements and environmental elements in two adjacent episodes change, it means that the ending and beginning of the two episodes are not in the same plot scene, but in two different plots. Therefore, it is necessary to ensure that the ending of the previous episode corresponds to the convergence point to end the plot at the end of the previous episode, and to ensure that the beginning of the next episode can start a new plot to correspond to a starting point. Therefore, when the ending of the previous episode is the convergence point and the beginning of the next episode is the starting point, even if the plot, scene, dialogue, etc. between the two episodes are not related, it is considered that there is coherence and consistency between the two episodes.

[0120] In some other embodiments, generating the script content corresponding to each episode based on the episode outline includes the following steps: S710, adding corresponding content to a preset script template based on the element information corresponding to the collection outline, the script template including a background template, a character template and a plot template.

[0121] Each episode of the script corresponds to a script template, which contains several necessary sections of the script. The background section is used to display the historical background, environmental background, scene construction content, etc. of the development of the plot. The character template is used to display the characters that appear in the episode and the corresponding role relationships, forces, ages and other information. The plot template is used to display the actions, dialogues, emotional interpretations and other content between characters.

[0122] S720, adding scene elements and environmental feature elements to the background section, adding character elements to the character section, and adding action elements and dialogue elements to the plot section.

[0123] Add the corresponding element information to the corresponding section to obtain the corresponding script content for each episode.

[0124] The present application also discloses a novel-script intelligent conversion system, comprising: The text volume segmentation module is used to obtain an input novel text and split the novel text into volumes, each of which contains at least one chapter.

[0125] The chapter marking module is used to mark each chapter to obtain chapter marks, and to infer each chapter to obtain the chapter outline corresponding to each chapter.

[0126] The role selection module is used to select the characters that appear in each chapter to obtain a list of characters.

[0127] The role labeling module is used to label the role information of each character in the character list in each volume based on the outline of each chapter.

[0128] The outline analysis module is used to perform feature content analysis based on the outlines of each chapter to obtain the outline information corresponding to each volume.

[0129] The role merging module is used to merge the role annotation information corresponding to each volume to obtain a merged set, and perform multi-source content processing on the merged set to generate a complete role card for each novel text.

[0130] The outline merging module is used to merge the sub-volume outline information corresponding to each sub-volume to obtain a merged set, and perform multi-source content processing on the merged set to generate a full-text outline for each novel text.

[0131] The script conversion module is used to obtain episode requests, and allocate plot volume information and role allocation information corresponding to each episode based on the episode requests combined with complete role cards and the full text outline; it is also used to merge corresponding multiple chapters based on the plot volume information and role allocation information to obtain an episode outline, and to analyze and adjust the content of the episode outline according to preset text rules; based on the episode outline, the script content corresponding to each episode is generated and merged to output a complete script of the novel text.

[0132] The implementation principle is: In particular, the novel is split into volumes and asynchronous parallel processing of character labeling and outline generation is implemented. Finally, the results of analysis and labeling of multiple volumes are merged. This can effectively handle very long novels and solve the limitations of current technology in processing very long texts. The script generation module can refer to the content of the previous episode when generating each episode of the script to ensure the continuity and consistency of the plot between episodes. At the same time, the format, logic, and structure are verified when each episode of the script is converted to reduce abnormal script conversion. The script volume and characters of each episode are judged and controlled to avoid the problem of unbalanced content in the episodes.

[0133] It should be understood that although the steps in the flowchart of the accompanying drawings are shown in sequence according to the instructions of the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise clearly stated in this document, the execution of these steps is not strictly limited in order and can be performed in other orders.

[0134] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A novel-script intelligent conversion method, characterized in that: The following steps are involved: Obtaining an input novel text, splitting the novel text into sub-volumes, wherein the sub-volumes contain at least one chapter; Marking each of the chapters to obtain a chapter mark, reasoning each of the chapters to obtain a chapter outline corresponding to each of the chapters, and selecting characters appearing in each of the chapters to obtain a character list; Based on each of the chapter outlines, character labeling information of each character in the character list in the sub-volume is obtained, and based on each of the chapter outlines, feature content analysis is performed to obtain sub-volume outline information corresponding to each of the sub-volumes; Merging the character annotation information corresponding to each of the sub-volumes and the sub-volume outline information to obtain a merged set, and performing multi-source content processing on the merged set to respectively generate complete character cards and full text outlines for each of the novel texts; Obtain an episode request, and assign plot volume information and role assignment information corresponding to each episode based on the episode request combined with the complete role card and the full text outline; Based on the plot volume information and the role allocation information, the corresponding multiple chapters are merged to obtain a collection outline, and the content of the collection outline is analyzed and adjusted according to a preset text rule; The script content corresponding to each episode is generated based on the collection outline and merged to output a complete script of the novel text.

2. The intelligent novel-script conversion method according to claim 1, characterized in that: Based on each chapter outline, the character information of each character in the character list in the sub-volume is annotated, and based on each chapter outline, characteristic content analysis is performed to obtain sub-volume outline information corresponding to each sub-volume, including the following steps: Obtaining the outline of each chapter and the list of characters appearing in the chapter; Select a single character in the character list, and annotate the relevant plot corresponding to the single character in each chapter outline based on the prompt word project to obtain character annotation information; Based on the prompt word project, each chapter outline is analyzed to screen the plots corresponding to the individual characters in each chapter to retain important plots and eliminate irrelevant plots, and the development logic and logical relationships in the chapter outlines are connected in series to obtain the volume outline information.

3. The intelligent novel-script conversion method according to claim 2 is characterized in that: The character labeling information corresponding to each of the sub-volumes and the sub-volume outline information are merged to obtain a merged set, and multi-source content processing is performed on the merged set to generate complete character cards and full text outlines for each of the novel texts, including the following steps: Acquire the role annotation information corresponding to each of the sub-volumes, and merge several role annotation information corresponding to the same role to obtain a role merge set; Deleting the repeated role labeling information in the role merge set and sorting the roles according to a preset sorting rule to obtain the complete role card; Obtaining the sub-volume outline information corresponding to each of the sub-volumes and merging them to obtain a full-text outline merged set; The duplicate contents in the merged set of full-text outlines are removed, the missing contents are supplemented or replaced, and the content consistency and logic of the merged set of full-text outlines are detected to obtain the full-text outline.

4. The intelligent novel-script conversion method according to claim 1, characterized in that: Obtaining an episode request, and assigning plot volume information and role assignment information corresponding to each episode based on the episode request combined with the complete role card and the full text outline, including the following steps: Calculating a total volume score based on the full text outline, the total volume score consisting of a first volume score corresponding to the dialogue content, a second volume score corresponding to the behavior content, and a third volume score corresponding to the emotional content; The total volume score is evenly divided into a collective volume score corresponding to each episode, and the full text outline is divided into a plurality of individual volume ranges based on the collective volume score, and the plot volume information corresponding to each episode is obtained based on the individual volume ranges; Based on the individual group mass range, select the set roles that contribute the group mass points in the complete role card, and assign corresponding set plots to each set role according to the group mass points contributed by each set role to obtain the role assignment information; Among them, each episode character is divided into weakly related characters, main characters and special characters based on the appearance frequency, number of chapters and chapter position corresponding to the episode characters, and there are at least two main characters and / or special characters in each episode.

5. The intelligent novel-script conversion method according to claim 4 is characterized in that: Based on the plot volume information and the role allocation information, corresponding multiple chapters are merged to obtain an outline, including the following steps: Acquire chapter information that matches the plot volume information, and merge a corresponding number of continuous chapters or skipped chapters based on the chapter information to obtain a script chapter group; The plot corresponding to the role matching the role allocation information in each chapter outline corresponding to the script chapter group is obtained to generate the collection outline.

6. The intelligent novel-script conversion method according to claim 1, characterized in that: Analyzing and adjusting the content of the outline according to preset text rules includes the following steps: Obtaining element information and node information in each of the outlines, wherein the element information includes action elements, character elements, scene elements, environmental feature elements, and dialogue elements, and the node information includes a starting point, a supporting point, a turning point, and a joining point; When each of the outlines satisfies the preset text rules, each of the outlines at least contains character elements, scene elements, environmental feature elements, and contains the action elements or dialogue elements; Adjust the order and position of the contents in the outlines so that each outline satisfies the preset text rules. Specifically, the beginning and ending positions in each outline include at least one of the starting point, the supporting point, the turning point and the joining point, and the node information content at the beginning and the end of adjacent outlines is consistent with coherence.

7. The intelligent novel-script conversion method according to claim 6 is characterized in that: Obtaining the node information in each of the above-mentioned sets of outlines includes the following steps: When new character elements, scenario elements, environmental characteristic elements, role relationships, and forces appear in the collection outline, the starting point is added to the corresponding position in the collection outline; Starting from the starting point, after the starting point, the support point is added to the corresponding position in the set outline every time the measurement length passes until the character elements, scene elements, environmental feature elements, role relationships, and forces in the set outline change, wherein the measurement length includes the word length and the collective measurement length; When an explicit turning point and / or an implicit turning point occurs in the collection outline, the turning point is added at the corresponding position in the collection outline, wherein the explicit turning point is characterized by a contradiction between the character elements, the character relationships, and the forces corresponding to the same character, and the implicit turning point is characterized by a contradiction between the emotional information corresponding to the action elements and the dialogue elements determined based on the prompt word engineering and / or a contradiction between the plot development and the logical development in the chapter outline; When the outline shows changes in character elements, scene elements, environmental feature elements, role relationships, and forces after the starting point, the convergence point is added at the corresponding position before the change occurs.

8. The intelligent novel-script conversion method according to claim 6 is characterized in that: The node information contents at the beginning and the end of the adjacent clusters are consistent with coherence, comprising the following steps: Obtaining the compactness corresponding to each set based on the collective quantity score corresponding to each set outline, and selecting a repetition mode or a continuous mode based on the compactness; Obtain the set outlines corresponding to two adjacent sets. When the adjacent set outlines meet the coherence, The repetition mode is characterized in that in the adjacent collection outlines, the beginning of the collection outline of the next collection repeats the content of the end of the collection outline of the previous collection; The continuity is characterized by that in adjacent episodes, if the scene elements and the environmental feature elements are the same, then based on the prompt word engineering, the plot coherence and logical coherence of the character elements, action elements, and dialogue elements at the beginning of the episode of the next episode and the character elements, action elements, and dialogue elements at the end of the episode of the previous episode are guaranteed; if the scene elements and the environmental feature elements are different, then it is guaranteed that the end of the episode of the previous episode corresponds to the conjunction point and the beginning of the episode of the next episode corresponds to the starting point.

9. The intelligent novel-script conversion method according to claim 6, characterized in that: Generating the script content corresponding to each episode based on the episode outline includes the following steps: Adding corresponding content to a preset script template based on the element information corresponding to the collection outline, wherein the script template includes a background section, a character section, and a plot section; The scene elements and the environmental feature elements are added to the background section, the character elements are added to the character section, and the action elements and dialogue elements are added to the plot section.

10. An intelligent novel-script conversion system, characterized in that: include: A text volume segmentation module, used for obtaining an input novel text, and splitting the novel text to obtain volumes, wherein the volumes contain at least one chapter; A chapter marking module, used for marking each of the chapters to obtain chapter marks, and performing reasoning on each of the chapters to obtain a chapter outline corresponding to each of the chapters; A character selection module, used for selecting characters appearing in each of the chapters to obtain a list of characters appearing; A role labeling module, used for labeling the role information of each character in the character list in the sub-volume based on each chapter outline; An outline analysis module, used for performing characteristic content analysis based on each of the chapter outlines to obtain the volume outline information corresponding to each of the volumes; A role merging module, used for merging the role annotation information corresponding to each of the sub-volumes to obtain a merged set, and performing multi-source content processing on the merged set to generate a complete role card for each of the novel texts; An outline merging module, used for merging the sub-volume outline information corresponding to each sub-volume to obtain a merged set, and performing multi-source content processing on the merged set to generate a full-text outline for each novel text; A script conversion module is used to obtain an episode request, and to allocate plot volume information and role allocation information corresponding to each episode based on the episode request combined with the complete role card and the full text outline; and is also used to merge corresponding multiple chapters based on the plot volume information and the role allocation information to obtain an episode outline, and to judge and adjust the content of the episode outline according to preset text rules; The script content corresponding to each episode is generated based on the collection outline and merged to output a complete script of the novel text.

Images