Examining Influences of Publication Dates on Citation Recommendation Systems | |
Gao, Zequn | |
2015 | |
关键词 | component Digital library Recommender Systems Information retrieval LATENT SEMANTIC ANALYSIS |
英文摘要 | Citation recommendation is a research problem that is both interesting and challenging. Recent years, various methods [2] [4] [12] have been put forward. But these methods omit some features, which are also usable. Firstly, in a common citation recommendation system, each document has a value of rank, which estimates its relevance to the target paper. However, even if some papers have the same value of rank, they may not satisfy the needs of users to the same extent, as users always prefer to newer ones. In this paper, we also examine the effect of paper publication year in citation recommendation. Specifically, we use curve fitting methods to describe the distribution of citation publication year of each venue, and use it as a coefficient in citation recommendation. Secondly, an author may have done research for many years, and his research interests may have changed several times. So, authors may have different features in different years. When considering similarity of authors of two papers, authors' research interests about the paper should be taken into account, too. In this paper, we proposed a time-difference-dependent citation recommendation, which utilizes author research interests and candidate papers' publication dates. Experiments on both AAN [20] and Microsoft Academic show our system outperform the baseline systems.; CPCI-S(ISTP); 1400-1405 |
语种 | 英语 |
出处 | 12th International Conference on Fuzzy Systems and Knowledge Discovery (FSKD) |
内容类型 | 其他 |
源URL | [http://ir.pku.edu.cn/handle/20.500.11897/450184] |
专题 | 信息科学技术学院 |
推荐引用方式 GB/T 7714 | Gao, Zequn. Examining Influences of Publication Dates on Citation Recommendation Systems. 2015-01-01. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论