专业领域高影响力机构与期刊大数据统计推荐算法与实现  

High impact professional institutions and journals recommendation： big data statistic algorithm and its implementation

作　　者：王云[1] 杨存榜[1] 粟强[1] 程功[1] 唐道润[1] 

机构地区：[1]中国工程物理研究院激光聚变研究中心信息中心,四川绵阳621900

出　　处：《教学与科技》2015年第1期1-5,共5页Teaching and Science Technology

摘　　要：针对激光聚变研究领域的机构和期刊，进行大规模数据统计评价，提供了一个“推荐关注机构排名表”和“推荐投稿期刊排名表”。具体工作包括：采集机构文章关键词；统计识别核心关键词；对关键词进行分类；生成组合检索式，批量化搜索与采集；统一机构变体名称；按相关度和用户量对文章进行评分；根据各个年度分值，利用最小二乘回归预测下年度估计分值，排序获得推荐结果。We conducted an evaluation on a large-scale statistical data concerning laser fusion institutions and journals, and provided a ＂recommended institutions ranking tables＂ and ＂recommended journal ranking table＂. The detail contents are as follows： collecting articles keywords of related research institutes; identify core statistical keywords; classify keywords; generate combined search strategy; batch search and acquisition; processing variant names of institution; rate articles by correlation and users data; according to scores of each year, get recommended results with forecast annual estimates score by using least squares regression and sorting.

关 键 词：激光聚变 研究机构 相关期刊 投稿 关键词识别:组合检索 相关性计算 统计推荐 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]