Plot Planning Of Animation Based On DBpedia

In the 1990 s,the technique of the Full-Life Cycle Automatic Animation Generation was proposed by Lu Ruqian who is an academician of Chinese Academy of Sciences.The technology uses limited language as input,introduces artificial intelligence technology and knowledge-based method,and automatically generates animation.In 2008,Zhang Songmao,a researcher of Chinese Academy of Sciences,proposed that apply the whole process of computer-aided animation automatic generation technology in mobile phone short message service,and led the research team to design and implement the automatic generation system of mobile phone short message 3D animation.The system uses the Chinese natural language sent by the sender of the short message as input.After information extraction,animation plot planning,animation quantitative calculation,real-time rendering and video synthesis processing,an animated short film that can express the short message content is generated and sent to the short message receiver.The plot planning module determines how the animation represents the content of the short message(SM),including the selection of the scene and model,the layout of the model,and the planning of the character's actions.It is an important part of the mobile animation system.The plot planning module takes the theme and template outputted by the information extraction module as input,and uses the animation knowledge base to match the theme and the template to obtain the animation entity corresponding to the short message,and uses the relationship between the animation entities as the knowledge base to plan the plot.Limited by the size of the data in the theme and template libraries and the number of relationships between entities in the animation knowledge base,the degree of planning that the story planning module can perform is very limited.DBpedia is one of the largest and most widely used cross-domain datasets in Linked Open Data(LOD).It is structured data extracted from Wikipedia and contains rich semantic information.This paper proposes to associate the animation knowledge base with DBpedia,and use DBpedia as the background knowledge in the plot planning process to improve the effectiveness,diversity and final expression of the animation.The main work of this paper includes:Firstly,the automatic association framework based on lexical and semantic for the animation knowledge base and DBpedia is designed and implemented.First,the statistical analysis of the animation knowledge base and the data in DBpedia shows the feasibility of using DBpedia as the background knowledge base of mobile animation system and the animation knowledge base.Then,based on wordNet,DBpedia entity is created.Name and synonym table of the entity name of the animation knowledge base;finally,construct lexical rules and semantic rules,establish the semantic association between the DBpedia entity and the animation entities such as scenes,models,background music,character actions,etc.Established a semantic association to optimize the animation knowledge base.Based on the automatic association framework of lexical and semantic,we have established six semantic associations between the animation knowledge base and DBpedia,including equivalence relations,parent-child relationship and entity type,and a total of more than 4,000,000 relationship triples.Secondly,the animation plot planning with DBpedia as the background knowledge is designed and implemented.First,the Chinese SM entity link is performed,that is,the corresponding entity is searched in DBpedia based on the result of the information extraction;then,through the semantic association between the animation knowledge base and DBpedia,the animation entity related to the DBpedia entity,including the scene,the model,Background music,character movements,etc.,and calculate selected scenes and models based on weighted random selection algorithm and similarity;finally,the selected scene is used as the basic framework for the planning of animation basics such as animation duration and background music,and based on DBpedia and The association between the animation knowledge base optimizes the planning of spatial layout and character movements.This article randomly selects 400 text messages received by the mobile animation system from January 1,2018 to December 31,2018.The design experiment evaluates the effectiveness,diversity,and improvement of animation performance plot planning based on DBpedia.Among the 400 text messages evaluated,383 SMs can generate valid plots,accounting for 95.75% of the total number of test SMs.Compared with the plot planning using only the animation knowledge base,the planning effectiveness is improved by 4.5%;Compared with the knowledge base,DBpedia-based plot planning has obtained more scenes for 267 text messages,accounting for 66.75% of the total number of test messages,and the average number of scenes per message has increased by 44.89%,indicating that the diversity of animation has been improved;The plot planning based on DBpedia,284 SMs got more models related to SM content,background music,actions and snowstorms,which accounted for 71.00% of the total number of SM and the average number of planned elements per SM increased by 44.30.%,indicating that the expressiveness of the animation has been improved.