Prophet预测-修正的主题强度演化模型——以干细胞领域为实证

doi:10.13266/j.issn.0252-3116.2020.08.010

图书情报工作 ›› 2020, Vol. 64 ›› Issue (8): 78-92.DOI: 10.13266/j.issn.0252-3116.2020.08.010

Prophet预测-修正的主题强度演化模型——以干细胞领域为实证

张鑫, 文奕, 许海云, 刘忠禹

中国科学院成都文献情报中心成都 610041

收稿日期:2019-07-28 修回日期:2019-11-07 出版日期:2020-04-20 发布日期:2020-04-20
通讯作者: 文奕(ORCID:0000-0002-6520-2733),研究馆员,硕士,通讯作者,E-mail:wenyi@clas.ac.cn
作者简介:张鑫(ORCID:0000-0001-8784-3788),馆员,硕士;许海云(ORCID:0000-0002-7453-3331),副研究员,博士;刘忠禹(ORCID:0000-0003-3852-1947),助理馆员,硕士。
基金资助:
本文系国家自然科学基金项目"基于科学-技术主题关联分析的创新演化路径识别方法研究"（项目编号：71704170）和中国科学院信息化专项"面向干细胞领域知识发现的科研信息化应用"（项目编号：XXH13506-203）研究成果之一。

Prophet Prediction-Correction Topic Evolution Model——A Case Study in Stem Cell Field

Zhang Xin, Wen Yi, Xu Haiyun, Liu Zhongyu

Chengdu Library and Information Center, Chinese Academy of Sciences, Chengdu 610041

Received:2019-07-28 Revised:2019-11-07 Online:2020-04-20 Published:2020-04-20

摘要/Abstract

摘要： [目的/意义] 主题演化对科技前沿探测、创新战略部署具有十分重要的作用。[方法/过程] 将主题演化分析过程分解为主题的表示、相似性关联和强度演化计算几个步骤，提出一种主题强度演化与预测模型，使用LDA模型进行主题的表示，提出内容、共现和趋势相似度等维度进行主题关联计算，引入基于Prophet的预测-修正模型进行主题演化趋势预测。并以干细胞领域为例，进行演化的实证分析。[结果/结论] 实验表明，对每个研究主题采用Logistic增长模型进行预测R2Score都达到0.90以上，表明Prophet中的Logistic增长模型与该领域主题增长趋势规律相符合，能够较好地拟合主题强度的演化趋势。提出的主题演化模型对专业领域内主题分布与演化分析有一定的借鉴意义。

关键词: 主题演化, 主题相似性, 时间序列, Prophet

Abstract: [Purpose/significance] Topic evolution analysis plays an important role in detection the technology frontier detection and innovation strategy deployment. [Method/process] In this paper, the topic evolution analysis process was divided into several steps: topic representation, similarity correlation and intensity evolution calculation. The LDA model was used to represent the topic; content, co-occurrence, and trend similarity were proposed for topic correlation calculations, and the prophet-based pre-train fine-tuning model was used to predict the topic trends. An empirical analysis was conducted using the stem cell field as an example. [Result/conclusion] Experiments show that the Logistic growth model has a R2Score of more than 0.90 for each topic. It shows that the Logistic growth model in Prophet is consistent with the growth trend of topics, and can fit the evolution trend of the topic intensity. The topic evolution model proposed in this paper has certain reference to topic distribution and evolution analysis in specific fields.

Key words: topic evolution, topic similarity, time series, Prophet

中图分类号:

G251

张鑫, 文奕, 许海云, 刘忠禹. Prophet预测-修正的主题强度演化模型——以干细胞领域为实证[J]. 图书情报工作, 2020, 64(8): 78-92.

Zhang Xin, Wen Yi, Xu Haiyun, Liu Zhongyu. Prophet Prediction-Correction Topic Evolution Model——A Case Study in Stem Cell Field[J]. LIS, 2020, 64(8): 78-92.

参考文献

[1] 罗文馨,王园园.技术主题演化研究方法综述[J].知识管理论坛,2018,3(5):255-265.
[2] HUMMON N P, DEREIAN P. Connectivity in a citation network:the development of DNA theory[J]. Social networks, 1989, 11(1):39-63.
[3] MARTINELLI A. An emerging paradigm or just another trajectory? Understanding the nature of technological changes using engineering heuristics in the telecommunications switching industry[J]. Research policy, 2012, 41(2):414-429.
[4] LU L Y Y, LIU J S. A survey of intellectual property rights literature from 1971 to 2012:the main path analysis[C]//Proceedings of PICMET'14 conference:portland international center for management of engineering and technology; infrastructure and service integration. Piscataway:IEEE, 2014:1274-1280.
[5] PILKINGTON A, MEREDITH J. The evolution of the intellectual structure of operations management-1980-2006:a citation/co-citation analysis[J]. Journal of operations management, 2009, 27(3):185-202.
[6] LAI R J, LI M F. Technology evolution of lower extremity exoskeleton from the patent perspective[J].Key engineering materials. 2014, 625:536-541.
[7] WANG Z Y, LI G, LI C Y, et al. Research on the semantic-based co-word analysis[J]. Scientometrics,2012,90(3):855-875.
[8] 胡正银,刘春江,隗玲,等.面向TRIZ的领域专利技术挖掘系统设计与实践[J].图书情报工作,2017,61(1):117-124.
[9] BLEI D M, NG A Y, JORDAN M I. Latent dirichlet allocation[J]. Journal of machine learning research, 2003, 3(1):993-1022.
[10] 范少萍,安新颖,单连慧,等.基于医学文献的主题演化类型与演化路径识别方法研究[J].情报理论与实践,2019,42(3):114-119.
[11] BLEI D M, LAFFERTY J D. Dynamic topic models[C]//Proceedings of the 23rd international conference on Machine learning. New York:ACM, 2006:113-120.
[12] WANG X, MCCALLUM A. Topics over time:a non-Markov continuous-time model of topical trends[C]//Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining. New York:ACM, 2006:424-433.
[13] PORTEOUS I, NEWMAN D, IHLER A, et al. Fast collapsed gibbs sampling for latent dirichlet allocation[C]//Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining. New York:ACM, 2008:569-577.
[14] HOFFMAN M, BACH F R, BLEI D M. Online learning for latent dirichlet allocation[C]//Advances in neural information processing systems 23.Vancouver:Curran Associates Inc., 2010:856-864.
[15] GRIFFITHS T L, JORDAN M I, TENENBAUM J B, et al. Hierarchical topic models and the nested Chinese restaurant process[C]//Advances in neural information processing systems. Vancouver:ACM,2004:17-24.
[16] MIMNO D, WALLACH H M, TALLEY E, et al. Optimizing semantic coherence in topic models[C]//Proceedings of the conference on empirical methods in natural language processing. Edinburgh:Association for Computational Linguistics, 2011:262-272.
[17] 王婷婷,韩满,王宇.LDA模型的优化及其主题数量选择研究——以科技文献为例[J].数据分析与知识发现,2018,2(1):29-40.
[18] WANG X, MCCALLUMA A, WEI X. Topical n-grams:Phrase and topic discovery, with an application to information retrieval[C]//IEEE International Conference on Data Mining. Piscataway:IEEE,2007:697-702.
[19] LI B, WANG B, ZHOU R, et al. CITPM:A cluster-based iterative topical phrase mining framework[C]//International conference on database systems for advanced applications. Dallas:Springer International Publishing,2016:197-213.
[20] 张琴,张智雄.基于PhraseLDA模型的主题短语挖掘方法研究[J].图书情报工作,2017,61(8):120-125.
[21] 刘自强,许海云,岳丽欣,等.基于Chunk-LDAvis的核心技术主题识别方法研究[J].图书情报工作,2019,63(9):73-84.
[22] 孙孟孟. 基于名词短语提取与词条权重分析的话题提取算法研究[D].杭州:浙江大学,2014.
[23] GRAVES A. Supervised sequence labelling[M]//Supervised Sequence Labelling with Recurrent Neural Networks. Berlin:Springer, 2012:5-13.
[24] TAYLOR S J, LETHAM B. Forecasting at scale[J]. The American statistician, 2018, 72(1):37-45.

Prophet预测-修正的主题强度演化模型——以干细胞领域为实证

Prophet Prediction-Correction Topic Evolution Model——A Case Study in Stem Cell Field

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

[1]	郭宇, 张传洋, 张海涛, 于文倩. 危机管理视角下突发事件舆情主题演化与治理分析[J]. 图书情报工作, 2022, 66(8): 113-121.
[2]	张吉玉, 张均胜. 考虑时序的单篇科技文献新颖性评估方法[J]. 图书情报工作, 2022, 66(17): 93-105.
[3]	梁爽, 刘小平. 基于文本挖掘的科技文献主题演化研究进展[J]. 图书情报工作, 2022, 66(13): 138-149.
[4]	丁敬达, 陈一帆, 刘超, 蔡微. 基于共词和Word2Vec加权向量的文献-主题语义匹配分析方法[J]. 图书情报工作, 2022, 66(12): 108-116.
[5]	翟姗姗, 王左戎, 陈欢, 潘港辉. 会话分析视角下的突发公共事件主题演化研究——以"新冠肺炎疫情"为例[J]. 图书情报工作, 2022, 66(11): 87-99.
[6]	马晓悦, 孟啸, 王镇, 刘益东. 网络突发事件中社交机器人情感的交互式影响机制研究[J]. 图书情报工作, 2021, 65(8): 74-84.
[7]	王晓, 马超, 翟姗姗. 会话分析视角下非正式信息交流主题与主题簇演化分析[J]. 图书情报工作, 2021, 65(17): 91-100.
[8]	马晓悦, 薛鹏珍, 陈忆金, 朱多刚. 社交媒体危机主题演化模型构建与趋势分析[J]. 图书情报工作, 2021, 65(13): 77-86.
[9]	王康, 高继平, 潘云涛, 陈悦. 多位态研究主题识别及其演化路径方法研究[J]. 图书情报工作, 2021, 65(11): 113-122.
[10]	何琳, 乔粤, 刘雪琪. 春秋时期社会发展的主题挖掘与演变分析——以《左传》为例[J]. 图书情报工作, 2020, 64(7): 30-38.
[11]	岳丽欣, 周晓英, 陈旖旎. 期刊论文核心研究主题识别及其演化路径可视化方法研究——以我国医疗健康信息领域期刊论文为例[J]. 图书情报工作, 2020, 64(5): 89-99.
[12]	徐涵, 刘小平. 作者主题模型及其改进的方法与应用研究综述[J]. 图书情报工作, 2019, 63(7): 135-145.
[13]	郝晋清, 王珺. 主题-时序视域下中国图书馆学会年会主题分布及演化分析[J]. 图书情报工作, 2019, 63(2): 107-119.
[14]	王军, 李子舰, 刘潇蔓. 不同文本长度的体验型产品在线评论时间序列研究——以电影评论为例[J]. 图书情报工作, 2019, 63(16): 103-111.
[15]	王园园, 赵亚娟. 基于非负矩阵分解的技术主题演化分析[J]. 图书情报工作, 2018, 62(10): 94-105.