面向知识服务的多模态资源知识重组研究

Research on Knowledge Reorganization of Multi-modal Resources for Knowledge Service

[目的/意义]多模态资源虽类型丰富但分散无序,挖掘多模态资源间隐含的知识关联进而实现跨模态、多角度的知识重组,对于多模态资源的智能管理与深层服务都具有重要的价值。[方法/过程]文章立足知识服务的特点与需求,结合多模态资源的特征,探讨多模态资源的知识挖掘方案,并基于此方案探讨可通用的多维度知识重组方案。在此基础上,以多模态红色资源为代表,通过对红色人物陈毅相关多模态资源的获取与处理,实现相关文本、图像、视频资源的知识重组,开发了可视化知识服务平台“陈年毅事”,并从时间、地点、人物、主题等不同维度探讨了可视化知识服务的实现路径。[结果/结论]通过实证研究,验证所提出知识重组方案的适用性与可行性,可以为多模态资源的深度挖掘、多维组织与服务应用提供一定的借鉴与参考。[局限]图像及视频资源有待更深层的挖掘;知识关联方式还有待补充。

[Purpose/significance] Although multi-modal resources are rich in types, they are scattered and disordered.It is of great value for intelligent management and deep service of multi-modal resources to explore the implicit knowledge association among multi-modal resources and realize cross-modal and multi-angle knowledge reorganization.[Method/process] In this paper, based on the characteristics and the requirements of knowledge service, combined with the characteristics of the multi-modal resources, explore the multi-modal resources of knowledge mining scheme, and based on this scheme to explore general multi-dimensional knowledge restructuring scheme.On this basis, represented by multi-modal red resources, through the acquisition and processing of multi-modal resources related to red character Chen Yi, to achieve the knowledge reorganization of related text, image and video resources, and developed a visual knowledge service platform “old things of Chen Yi” project, it also discusses the realization path of visual knowledge service from different dimensions such as time, place, character and topic.[Result/conclusion] Through empirical research, the applicability and feasibility of the proposed knowledge reorganization scheme are verified, it can provide some reference for the deep mining, multi-dimensional organization and service application of multi-modal resources.[Limitations] Image and video resources need to be further explored;the way of knowledge association needs to be supplemented.