江苏省科技期刊可视化阅读平台

导航条

江苏科技信息 ›› 2019, Vol. 36 ›› Issue (8): 47-49.doi: 10.1004-7530/2019-36-8-47

• 应用技术 • 上一篇    下一篇

科技文献个性化服务系统在高校中的应用

蒋晓玲,吴卫成   

  1. 淮阴工学院 计算机与软件工程学院,江苏 淮安 223001
  • 出版日期:2019-03-20 发布日期:2019-07-09
  • 通讯作者: 吴卫成
  • 作者简介:蒋晓玲（1981—）,女,江苏淮安人,讲师,硕士;研究方向：数据处理,物联网技术。
  • 基金资助:
    2018年度江苏省建设系统科技项目(2018ZD269);2018年度淮阴工学院课堂教学改革实践专项课题(2018021)

Application of individualized service system for scientific and technological documents in colleges and universities

Xiaoling Jiang,Weicheng Wu   

  1. Collge of Computer and Software Engineering, Huaiyin Institute of Technology, Huaian 223001, China
  • Online:2019-03-20 Published:2019-07-09
  • Contact: Weicheng Wu

RichHTML

632

PDF

1532

摘要：

为了提高文献推送的准确率,文章设计了一种基于Word2Vec的高校师生科技文献个性化服务系统,通过爬虫挖取以及基于Lucene的全文检索方案对数据进行检索,通过Word Embedding技术计算相似度,在一定程度上解决了系统数据稀疏性与冷启动问题。在120万篇文献中,实验取得了科技文献82%的推荐准确率,满足高校师生对文献检索的效率需求。

关键词: 个性化推荐, 信息过载, 用户建模

Abstract:

In order to improve the accuracy of document push, this paper designs a Word2Vec-based personalized service system for scientific and technological documents of teachers and students in colleges and universities. Through crawler digging and full-text retrieval based on Lucene, the similarity degree is calculated by Word Embedding technology, which solves the problem of system data sparsity and cold start to some extent. In 1.2 million articles, the experiment has acquired science and technology. The literature recommendation accuracy is 82%, which meets the demands of the teachers and students for the efficiency of document retrieval.

Key words: personalized recommendation, information overload, user modeling

中图分类号: