摘要
当前使用的深度学习驱动、基于哈希算法的数据检索方式,容易受到原始数据集冗余信息和噪声影响,存在检索查准率和查全率不高的问题,因此提出了基于粒子群算法的科技创新数据检索系统设计。使用检索引擎构建索引库群,设计检索数据存储模块、关联导航模块、数据分词流程。利用粒子群算法求解分词最短路径,根据数据分词处理结果,获取查询关键词对应的查询分词向量,完成数据检索,以此作为分词结果,避免系统检索受到原始数据集冗余信息和噪声影响。由实验结果可知,该系统查准率最高为96%,查全率最高为97%,具有高效检索效果。
The current data retrieval method driven by deep learning and based on hash algorithm is easy to be affected by the redundant information and noise of the original data set,and the retrieval precision and recall are not high.Therefore,the design of scientific and technological innovation data retrieval system based on particle swarm optimization algorithm is proposed.Use the retrieval engine to build the index library group,design the retrieval data storage module,association navigation module and data word segmentation process,use the particle swarm optimization algorithm to solve the shortest path of word segmentation,obtain the query word segmentation vector corresponding to the query keyword according to the data word segmentation processing results,and complete the data retrieval.This is the result of word segmentation to avoid the influence of redundant information and noise of the original data set.The experimental results show that the system has the highest precision of 96% and the highest recall of 97%.It has efficient retrieval effect.
作者
马芳平
李林
郭金婷
柳玉兰
徐镭梦
MA Fangping;LI Lin;GUO Jinting;LIU Yulan;XU Leimeng(Guodian Dadu River Drainage Area Hydroelectricity Development Co.,Ltd.,Chengdu 610095,China)
出处
《电子设计工程》
2023年第15期66-69,74,共5页
Electronic Design Engineering
基金
国能大渡河流域水电开发有限公司科技创新项目(GJNY-DDH-2020-009)。
关键词
粒子群算法
科技创新数据
检索
查准率
查全率
particle swarm optimization
scientific and technological innovation data
retrieval
precision
recall