一种新的经典文献推荐机制的设计

摘要

Classical papers are of great help for beginners to get familiar with a new researching area. However,how to dig them out is a difficult problem. This paper proposes a novel classical papers recommendation mechanism, which is based on two disciplines: the download persistence discipline and the citation approaching discipline. The download persistence discipline indicates that the download frequencies of classical papers will not decrease even if they were published long ago; the citation approaching discipline indicates that when a classical paper is cited, other papers cited by it are also tend to be cited. The thesis has proven the above two disciplines with experiments. After that the classical papers for specific fields are dug out with actual data. Facts prove that the classical papers recommended by the mechanism designed in the thesis are of high quality and accuracy.%经典文献对于初学者进入某一新的研究领域大有帮助,然而,如何发掘出这些经典文献是一个难题.设计一种新的经典文献推荐机制,这一机制基于两条规律:下载持续律和引用传递律.下载持续律指出经典文献的下载频率不会因为其发表时间距今久远而有所降低；引用传递律指出如果某经典文献被引用,则它引用的那些文献也很可能被引用.首先用实验证明上述两条规律,然后进一步用真实数据来挖掘出特定领域的经典文献.事实证明该设计的机制所推荐的经典文献有着较高的质量和准确性.