基于标签技术和群体绩效评价的多媒体检索

Multimedia Retrieval Based on Tagging Technology and Communities Performance Evaluation

多媒体文档的有效检索是现代企业实现知识管理和知识共享的关键,但目前对多媒体对象和内容的有效检索还没有令人满意的技术和解决方案。针对目前普遍采用的基于关键词的文档检索方法在多媒体文档检索中存在的问题,提出了基于标签技术和由企业中群体执行的绩效评价机制的多媒体检索问题解决方案,构建了引入绩效评价的文档管理平台标签系统的E-R模型,设计了文档管理平台标签系统的业务逻辑,给出了标签绩效评价指标测算方法。绩效评价和标签技术的有机结合有效保证了标签的数量和质量,检索的效率和质量会在系统的自学习、群体意识和群体知识库的知识发现过程中不断提高,可以满足绝大多数企业的日常需要。

Effective multimedia retrieval is the key to knowledge management and knowledge sharing for modern enterprises,but up to now the retrieval from multimedia documents is far from satisfactory for want of better technologies or technical solutions.With regard to the problems in keyword-based retrieval methods used on multimedia documents,this paper proposes a solution based on performance evaluation being executed by communities in a enterprise and tagging technology system and establishes an E-R model of the tagging system improved with performance evaluation for the document management platform.It presents a design of the business logic for the tagging system of the document management platform and provides an index assessment method for performance evaluation of the tagging.The purposeful integration of performance evaluation and tagging technology will guarantee the adequacy of the number and quality of the tags,and the efficiency and quality of the retrieval will be continually improved along with the system＇s self-learning,communities awaring and knowledge discovering on communities knowledge base,and this will surely satisfy the usual needs of most enterprises.