论文部分内容阅读
针对已有科技文献查询方法往往会遗漏“特征字符不同、但语义相似”的推荐对象的问题,从特征词语义角度研究用户兴趣偏好、科技文献特征的表示方法;提出考虑语义的科技文献与用户兴趣偏好相似性度量方法。在分析个性化科技文献查询推荐业务流程的基础上,设计个性化科技文献查询推荐原型系统ALRS,介绍ALRS的设计思路、功能模块;分析该系统核心算法CGR算法的关键数据结构;详细讨论CGR算法中“基于概念泛化的相似性度量方法”的设计思路,并给出实现该原型系统的核心代码及执行示例。
In view of the existing scientific and technical literature inquiry method, the problem of “the characteristic characters are different but the semantic similarity” is often omitted, the preference preference of the user and the representation of the scientific literature feature are studied from the semantic meaning of the characteristic words; the scientific and technical documents Similarity measurement method with user interest preference. Based on the analysis of the recommended business process of personalized scientific literature search, we design the personalized prototype of ARRS, and introduce the design idea and function module of ALRS. The key data structure of CGR algorithm is analyzed. The CGR algorithm In the “similarity based on conceptualization of measurement methods ” design ideas, and gives the core of the prototype system code and implementation examples.