摘要
针对目前同名消歧方法只考虑单一策略且消歧外延过大的问题,通过工作单位这一具有高区分度的特征来缩小消歧外延,将同名消歧问题具体化到姓名相同、单位不同的科技人才歧义消解问题上,并利用科技人才的成果信息,组合了实体链接、成果时间窗、成果合著者和成果相似度四个消歧策略,提出了一种基于多策略组合模型的消歧方法。实验结果表明,该方法取得了较好的同名消歧效果。
Aiming at the problem that the current same-name disambiguation method only considers a single strategy and the disambiguation is too largely extended, the disambiguation extension is narrowed by the high-discrimination feature of the work unit. The problem of same-name disambiguation is embodied in the disambiguation of scientific and technological talents with the same name and different units. By using the achievement information of scientific and technological talents, the four disambiguation strategies of entity hnk, result time window, result co-authors and result similarity are combined, and a disambiguation method based on multi-strategy combination model is proposed. The experimental results indicate that this method could achieve better disambiguation effect of the same name.
作者
刘林
LIU Lin(School of Computer Science and Technology,Hangzhou Dianzi University,Hangzhou Zhejiang 310018,China)
出处
《通信技术》
2018年第8期1836-1843,共8页
Communications Technology
基金
浙江省重点研发计划项目(No.2017C01009)~~
关键词
科技人才
同名消歧
成果信息
多策略组合模型
scientific and technological talent
disambiguation by the same name
achievement information
combination model of multi-strategy
