摘要
提出一种基于增强遗传算法的对多媒体数据的查询优化算法。将查询种群组织成多个小生境,一个小生境用于查询文档空间的一个区域,设计相应的基于项权重和相似项的交叉算子、自适应变异算子,通过引入局部搜索机制来增强算法的搜索能力,最后依据相关性次序将查询结果进行合并,返回查询结果。实验结果表明,该算法在查询精度和查询速度上均能获得比较满意的效果。
In this paper, a novel query optimization algorithm based on genetic algorithm for multimedia data is proposed. The population is organized into query niches and each niche is used to explore an area of the potential document space. The fitness function is computed based on query similarity with relevant documents and the number of query niche. Crossover operators based on term weight and similar terms are adopted for reproduction of new query population. Experimental results show that this algorithm has good precision of document retrieval and faster query speed.
出处
《计算机与数字工程》
2007年第6期28-31,共4页
Computer & Digital Engineering
