CAJ | 학술논문

评价论文影响力是科学计量学领域的重要内容。科学引文网反应了论文之间的相互引用关系。本文基于 H指数的思想对于评价论文影响力提出“H 算法”，进而提出改进后的“H+算法”用以解决 H 算法简并度高的问题。把算法应用在美国物理学会(APS)提供的科学引文网中，实验结果表明，两种算法计算出的评价指标呈现高端幂律分布。这两种算法比PageRank算法的时间复杂度都要低；在评价优秀论文方面的表现上，“H 算法”准确率优于PageRank算法。
평개논문영향력시과학계량학영역적중요내용。과학인문망반응료논문지간적상호인용관계。본문기우 H지수적사상대우평개논문영향력제출“H 산법”，진이제출개진후적“H+산법”용이해결 H 산법간병도고적문제。파산법응용재미국물이학회(APS)제공적과학인문망중，실험결과표명，량충산법계산출적평개지표정현고단멱률분포。저량충산법비PageRank산법적시간복잡도도요저；재평개우수논문방면적표현상，“H 산법”준학솔우우PageRank산법。
Influence evaluation is important in scientific bibliometrics.The science citation network reflects reference relationship of papers published by scientists.We propose an “H algorithm”to evaluate paper influence.This is based on the H-index and we propose an improved “H+ algorithm”to solve the high degeneracy of “H algorithm”.Algorithm in science citation network offered by APS was used and it was found that evaluation index of the two algorithms show top power-law distribution.Time complexity of the two algorithms is lower than the PageRank algorithm;Accuracy rate to evaluate excellent papers of “H algorithm”is higher than PageRank algorithm.