[目的/意义]学术谱系以知识传承的方式助推科学发展。研究知识传承特征，探索学术谱系传承模式及其学术产出影响效能，为探索人才成长规律及人才政策制定提供参考。[方法/过程]基于LDA主题模型，以遗传学领域谈家桢为核心的学术谱系成员发表的期刊文献为研究对象，对该谱系成员的研究主题进行抽取，借鉴生物学"遗传"与"变异"的概念，根据主题相似度将谱系成员划分为"遗传学者""变异学者"和"非遗传非变异学者"，并对三种学者的学术绩效进行分析。[结果/结论]分析结果表明，谈家桢学术谱系内"遗传学者"和"变异学者"的学术绩效相对较高；"非遗传非变异学者"的数量占比最多，但学术绩效相对较低；"遗传学者"与"变异学者"在不同主题下的分布具有明显差异。
[Purpose/significance] Academic pedigree promotes science development by the way of knowledge inheritance. It is of great reference value to study the characteristics of knowledge transmission and explore the effect of inheritance model on academic output, and it is of great reference value for the relevant departments to find out the law of talent growth and formulate scientific and technological personnel training policy.[Method/process] By the method of LDA topic model, this paper took the journal literature of genetics published in CNKI database as research object, and quoted the concept of "hereditary" and "variation" in biology. Then, according to the topic similarity, we divided pedigree members into "hereditary scholars", "variation scholars" and "non-hereditary non-variation scholars", and analyzed the academic performance of these three kinds of scholars.[Result/conclusion] The results show that the academic performance of "hereditary scholars" and "variation scholars" in the academic pedigree of Tan Jiazhen is relatively high; The number of "non-hereditary non-variation scholars" is the largest, but their academic performance is relatively low; For different topics, the distribution of "variation scholars" and "hereditary scholars" is significantly different.
[1] 刘颖, 张燕蕾, 张大庆. 中国科学家学术谱系库的构建思路初探与实践[J]. 图书情报工作, 2014,58(S2):60-62.
[2] CRONIN B, SUGIMOTO C. Academic genealogy[C]//BLAISE C,CASSIDY R. Beyond bibliometrics.London:MIT Press, 2014:480.
[3] JACKSON D C. Academic genelogy and direct calorimetry:a personal account[J].Advances in physiology education, 2011, 35(2),120-128.
[4] MALMGREN R D, OTTINO J M, AMARAL L A N. The role of mentorship in protégé performance[J]. Nature, 2010, 465(7298):622-626.
[5] 常欢, 吕瑞花, 张佳静. 学术谱系内合作网络研究——以刘东生为核心的第四纪学术谱系为例[J]. 情报理论与实践, 2016, 39(4):14-19.
[6] BLEI D M, NG A Y, JORDAN M I. Latent Dirichlet allocation[J]. Journal of machine learning research, 2003,3(3):993-1022.
[7] ROSEN Z, MICHA L, GRIFFITH S, et al. The author-topic model for authors and documents[C]//Proceedings of the 20th conference on Uncertainty in artificial intelligence.New York:AUAI, 2004:487-494.
[8] DHILLON I S, MODHA D S. Concept decompositions for large sparse text data using clustering[J]. Machine learning, 2001, 42(1):143-175.
[9] 史庆伟, 乔晓东, 徐硕, 等. 作者主题演化模型及其在研究兴趣演化分析中的应用[J]. 情报学报, 2013, 32(9):912-919.
[10] WANG X, MCCALLUM A. Topics over time:a non-Markov continuous-time model of topical trends[C]//BACKSTROM L, HUTTENLOCHER D, KLEINBER G, et al. Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining.New York:ACM, 2006:424-433.
[11] 谢平. 生命的起源-进化理论之扬弃与革新[M]. 北京:科学出版社, 2014.
[12] 刘俊婉, 郑晓敏, 宿娜,等. 国内外情报学领域期刊发文时滞的计量分析——以Scientometrics和《情报学报》期刊为例[J]. 中国科技期刊研究, 2016, 27(12):1292-1299.
[13] 崔凯. 基于LDA的主题演化研究与实现[D]. 长沙:国防科学技术大学, 2010.
