JOURNAL CITATION REPORTS(JCR)

Journal Citation Report简介 期刊引文分析报告(Journal Citation Reports，简称 JCR)是一个独特的多学科期刊评价工具； JCR是唯一提供基于引文数据统计信息的期刊评价 资源，可以显示出施引和被引期刊之间的相互关系； JCR中所包含的各种计量指标，为评价学科领域中 期刊的相对影响力提供了一套系统、客观的方法。 包含将近200个学科领域的7500多种，经过专家评议认定的高质量期刊

为不同目的设计的不同产品 JCR 和Web of Science都是涵盖书目信息bibliographic 和引文索引 citation indexing的数据库 两个产品对数据的分析和组织方式是为不同目的服务的 Observe animal behavior with binoculars Observe cellular behavior with a microscope WoS JCR

JCR和Web of Science: 计算期刊Impact Factor的分子 MATCH!

JCR 和 Web of Science 计算期刊Impact Factor的分母 NOT MATCHED? Why the difference?

为什么会不同? 数据年 vs 出版年 (WoK 5后会有所改变) Counted in WoS Counted in JCR

Journal Citation Report特点 JCR分为两个版本： JCR Science Edition—提供SCIE中所收录的170多个学科领域，共8,073多种期刊的引文分析信息； JCR Social Sciences Edition—提供SSCI中所收录的50多个学科领域，共2,731多种期刊的引文分析信息。

通过http://isiknowledge.com访问JCR数据库

选择JCR版本、出版年，以及分类方式

查看某学科领域内所有数据

对某学科领域内期刊进行排序

A QUICK LOOK AT GENERAL METRICS

期刊影响力指标 – 基于引文数据 总被引频次(Total Citations) 期刊影响因子(Journal Impact Factor) 5年影响因子(5 Year Impact Factor) 立即指数(Immediacy Index) 被引半衰期(Cited Half-life) 引用半衰期(Citing Half-life) Terminology matters here…a consistent message helps our market understand Total citations – each journal has a special algorithm built to aggregate citations – The JCR Editors are the hot-shots at that…I ask their help or I could aggregate my way out of a paper bag. Variants, abbreviations, homographs… Immediacy – one year, same year Journal Impact Factor and 5 Year Impact Factor – Let’s look at what this tells you…it’s a ratio – citations adjusted for size. A journal that consistently posts a high Journal Impact Factor and rank in category

Time Citations Journal Impact Factor Immediacy Index 5 year IF 2005 All Prior Years 2005 2006 2007 2008 2009 2010 2011 2012 5 year IF Citations There are two key things to note in this slide – That TIME, in this graphic, moves from left to right – from past to present – but the JCR moves from right to left. The JCR stands in the current time (more or less) and looks BACKWARD at previously published materials. It is history, not prediction. That All the citations for the JCR come from ONE year of processed source material. In this case, the year 2008. This refers to a 2008 COVER date. As you all know…2008 materials can be received as early as Autumn of 2007, or as late as yesterday. The JCR considers all materials with a publication year of 2008, processed from any point in 2004 on forward to any materials indexed in time for the JCR extraction. For the past 2 years, we have used an extraction close date of March 1. Generally, then, it means that JCR 2008 will include materials received as late as the second or third week in February. Nearly 2 million source items processed in 2008 – and all references from these articles to previously published material 42,878,592 million cited references for JCR 2008 施引文献 – 发表于2010年 被引文献 – 发表于2009, 08, 07, 06, 及05年

影响因子(Impact Factor) 图书馆员利用影响因子可以帮助他们决定应该订购哪些高质量的期刊； 研究人员通过影响因子可以决定将他们撰写的论文发表在哪本刊物上； 学生通过影响因子可以挑选出哪些刊物适合他们阅读； 编辑人员通过影响因子可以看出某项编辑策略出台后，刊物影响力的变化； 出版商通过影响因子可以长期跟踪自己以及对手的期刊评价情况； 科研管理人员可以看到他们研究机构的论文常发表在什么等级的期刊上。

Web of Science文章全记录页面

链接至JCR中期刊的影响因子趋势数据

5年影响因子

立即指数(Immediacy Index) 该指数用来评价哪些科技期刊发表了大量热点文章，进而能够衡量该期刊中发表的研究成果是否紧跟研究前沿的步伐。

被引半衰期(Cited Half-life) Citing Cited 被引半衰期(Cited Half-life) 使用期刊被引半衰期您可以查看很久以前出版的某本期刊是否仍然被引用,从而判断某刊过去的出版情况是否良好以及其中刊载文章的质量。 期刊被引半衰期有利于图书馆员有效的管理馆藏。如果某刊的被引半衰期比较短则意味着旧的论文没有新论文那样多次被引，因此可以将旧刊装订归档。

被引半衰期的定义 统计该刊不同年份发表的文章在当前JCR年被引用的次数； Citing Cited 被引半衰期的定义 2010年该刊发表的论文在2004年被引用的次数 总被引次数=2,534 被引次数的累计百分比 统计该刊不同年份发表的文章在当前JCR年被引用的次数； 计算出每一年以及该年以前的被引次数之和在总被引次数中所占的比例。这样随着年度的推移和比例的累积，直到累计百分比达到50%； 依据年度和年度的分段测定，便可以确定被引半衰期。 期刊被引半衰期的最大值是>10 年。

被引期刊图示 白色/灰色部分表示期刊的被引半衰期 淡蓝色柱状表示用来计算影响因子的数据 黄色表示自引次数，也是衡量期刊质量的重要指标之一 Citing Cited 被引期刊图示 白色/灰色部分表示期刊的被引半衰期 淡蓝色柱状表示用来计算影响因子的数据 黄色表示自引次数，也是衡量期刊质量的重要指标之一

被引期刊数据 Impact Factor 总被引次数 Eigenfactor & Art. Inf. 立即指数 被引半衰期= 8.2 yrs Citing Cited 被引期刊数据 Impact Factor 总被引次数 Eigenfactor & Art. Inf. 立即指数 被引半衰期= 8.2 yrs 期刊自引 The cited journal list is, in some ways, the single most important, and the single most overlooked set of data in the JCR. It is a tabular rendering of the cited journal - showing not just the journals citing it - but also how the citations from each journal are distributed over time. With careful use of the Web of Science, you can actually isolate the specific articles involved in these exchanges - that’s a somewhat esoteric exercise. At the journal-level, in the JCR, however, you can still get a look at the journals that are linked, scientifically, with your own. I referred to the Cited Journal list as the most important data in the JCR? It’s because the top-line (summary) data on this list is actually tied to four of the performance metrics: total citations, immediacy index, Impact Factor, Cited half-life, Eigenmetrics and cited half-life, and journal self-citations. This has always been The cited journal table has been part of the JCR since the product was published apart from the SCI - in the late 1970’s - and it has always contained data about the degree to which journal self-citation contributes to the performance metrics >….the interesting thing here is the reminder that performance metrics come out of the real interaction of the journal with the literature…

施引半衰期(Citing Half-life) Cited 施引半衰期(Citing Half-life) 了解期刊引用参考文献的时间范围，您可获知该刊是否正在引用最新的文献，从而可以确定该刊是否有利于您时刻紧跟研究领域的发展步伐； 总之，被引半衰期和引用半衰期的值比较接近。如果数目有很大差距则表明该刊引用了大量的最新文献，但是同时大部分旧文献也在被引用，反之亦可。

施引半衰期的定义 统计该刊在当前JCR年发表的文章对其他不同年份文章的施引次数； Citing Cited 施引半衰期的定义 该刊在2010年发表的论文对其他发表于2004年论文的施引次数 总施引次数=2670 施引次数的累计百分比 统计该刊在当前JCR年发表的文章对其他不同年份文章的施引次数； 计算出每一年以及该年以前的施引次数之和在总施引次数中所占的比例。这样随着年度的推移和比例的累积，直到累计百分比达到50%。 依据年度和年度的分段测定，便可以确定施引半衰期。 期刊施引半衰期的最大值是>10 年。

Citing Cited 施引期刊图示 白色/灰色部分表示期刊的施引半衰期 黄色表示自引次数

Citing Cited 施引期刊数据 期刊自引 施引文献半衰期= 8.4 yrs

相关期刊 期刊间的相关性衡量了两本期刊的引文强度，相关期刊列表显示了一本期刊(一个学科类别)的”引文邻居”： 对于科学家，可以参考该列表选择发表论文的期刊 对于图书馆员，可以帮助他们决定应该订购哪些相关期刊 对于出版商，可以帮助他们发表竞争对手，以及潜在的稿源

相关期刊 其实也可以很简单……

期刊自引数据 New in 2009 Ideally, citation is an article-level phenomenon. When an author writes an article he/she will cite the works that are most relevant to the current work. This will usually have a small effect on the overall citation count – and will really be a part of how the journal defines its impact in the literature

期刊自引 期刊的自身引用本身并不是什么 “不端”行为。 一个作者当然会引用相关研究，这些研究成果发表的期刊 很可能成为作者投稿的选择项。 但是80% 在JCR Science Edition 中列出的期刊的自引率 都低于20%。基于学科的引用趋势我们可以了解什么是过 度自引，过度自引会减少期刊的影响因子的可靠性。 期刊的自身引用本身并不是什么 “不好的 ”行为。 一个作者当然会引用相关研究，这些 研究成果发表的期刊很可能成为作者 投稿的选择项。 但是-- 80% 在JCR Science Edition 中列出的期刊的自引率都低于20%。 基于学科的引用趋势我们可以了解什 么是过度自引，过度自引会减少期刊 的影响因子的可靠性。 31

期刊源文献数据 体现2010年该期刊发表文献的基本数据

学科领域整体的引文分析指标

中值影响因子&综合影响因子 中值影响因子 综合影响因子 将整个学科领域看作一本期刊，按照期刊影响因子的定义计算学科领域的影响因子 …… 0.122 0.535 1.047 1.152 2.461 3.911 4.767 …… 影响因子递增的顺序

综合立即指数 综合立即指数 将整个学科领域看作一本期刊，按照期刊立即指数的定义计算学科领域的立即指数

综合被引半衰期

综合施引半衰期

A DEEP DIVE ON THE NEW METRICS

学科列表的排名 许多期刊包含不止一个JCR 学科领域的内容范畴 。为了更好地揭示一个期刊在所有指定的类目中的影响，新推出的类目列表将会显示该期刊被标引的不同学科,同时根据影响因子计算给出该刊在对应学科中的排名以及期刊在学科所处的四分位值。 许多期刊包含不止一个JCR 学科领域的内容范畴 。为了更好地揭示一个期刊在所有指定的类目中的影响，新推出的类目列表将会显示该期刊被标引的不同学科,同时根据影响因子计算给出该刊在对应学科中的排名以及期刊在学科所处的四分位值。 。

类目的等级– 箱型图： 类目箱型图是伴随类目等级列表出现的。箱型中不同地方的空间、四分位点都表明了期刊在类目之中的影响力的分布程度。您可以通过图示的方式了解期刊的影响因子在不同学科中的位置– 想象在期刊的影响因子的位置有一条线可以有助于理解图中所示内容。我们以本图为例，代表影响因子的橘色线在类目C动物学顶部很高位置。我们来看一下箱型图的具体解释：箱型自身代表第25至第75个百分位，或中间的50%。箱型内的水平线代表中值影响因子，而加号表示 平均影响因子。箱型上方和下方的水平线表示终端，这些表示在特定百分位中的端点。圆圈代表端点之外的期刊。

相对于传统的2年影响因子，5年影响因子能更好的评估 那些发表的论文被引用周期很长的领域中的期刊影响力 生物化学及分子生物学 传统的影响因子基于前两年的期刊发表文章在当年的被引情况。 5年影响因子基于前五年的期刊内容被引情况在当年的被引用情况。 5年影响因子的介绍： 相对于传统的2年影响因子，5年影响因子能更好的评估那些发表的论文被引用周期很长的领域中特定期刊的影响。传统的影响因子基于两年的期刊内容被引情况：即本年度期刊对于该刊前两年发表文章的引用。 5年影响因子基于本年度期刊对于前五年发表该刊论文的引用。请注意此处显示的生物化学及分子生物学期刊的五年影响因子比传统的两年影响因子要低。

地质学 但是，我们也会发现在地质学类目中，每一栏的五年影响因子都要比两年的影响因子高。这就与上一张幻灯片中看到的生物化学及分子生物学形成了对比。地质学研究发表的论文需要更多时间才能完全表现出它的影响力。

Journal Impact Factor captures peak citation activity in fast-moving fields

Journal Impact Factor captures rising citation activity in all fields

VIF reflects citation plateau in slow-moving fields

VIF always reflects a greater portion of total citation activity 174 Subject cats in SCI + 56 in SSCI = 230 Of 230, only 3 have at least 50% of their citation activity

Time Citations Eigenfactor™ metrics 2005 2006 2007 2008 2009 2010 2011 All Prior Years 2005 2006 2007 2008 2009 2010 2011 2012 Eigenfactor™ metrics Citations There are two key things to note in this slide – That TIME, in this graphic, moves from left to right – from past to present – but the JCR moves from right to left. The JCR stands in the current time (more or less) and looks BACKWARD at previously published materials. It is history, not prediction. That All the citations for the JCR come from ONE year of processed source material. In this case, the year 2008. This refers to a 2008 COVER date. As you all know…2008 materials can be received as early as Autumn of 2007, or as late as yesterday. The JCR considers all materials with a publication year of 2008, processed from any point in 2004 on forward to any materials indexed in time for the JCR extraction. For the past 2 years, we have used an extraction close date of March 1. Generally, then, it means that JCR 2008 will include materials received as late as the second or third week in February. Nearly 2 million source items processed in 2008 – and all references from these articles to previously published material 42,878,592 million cited references for JCR 2008 施引文献 – 发表于 2010年 被引文献 – 发表于2009, 08, 07, 06, 05年

特征因子(Eigenfactor Metrics) Eigenfactor MetricsTM通过特定方法测量期刊影响,提供了描述期刊影响力的更广泛的内容,现在已经被学术领域广泛接受。这些数据补充了期刊的影响因子及其他JCR计量数据, 这些计量数据是由Eigenfactor Project™ 开发的– Eigenfactor Project™是一个由华盛顿大学生物系的Bergstrom 实验室资助的非商业学术研究项目– www.eigenfactor.org 其它更新功能还包括添加了Eigenfactor metrics: Eigenfactor 和Article influence，Eigenfactor MetricsTM通过特定方法测量期刊影响,提供了描述期刊影响力的更广泛的内容,现在已经被学术领域广泛接受。这些数据补充了期刊的影响因子及其他JCR计量数据, 这些计量数据是由Eigenfactor Project™ 开发的–Eigenfactor Project™ 是一个由华盛顿大学生物系的Bergstrom 实验室资助的非商业学术研究项目– www.eigenfactor.org 48

特征因子(Eigenfactor Metrics) 一个期刊若经常被其他有影响力的期刊引用，则这个期刊也是有影响力的。 它跨越所有学科 并且在一定程度上去除了自引的影响。 The math is so thorny for these – that almost everyone is reduced to metaphorical explanations Eigenfactor considers the journal to journal citation network created by citations in the current year to any of the past 5 years.  For each journal the EF score is a reflection of the percentage of this citation network that is concentrated on this journal.   Operationally, this means that a citation FROM a highly influential journal will have more “weight” in determining the influence of other journal than a citation from a more modestly successful journal.  The “time” thing comes from one of the EF explanations…using random walk metaphor.  If you take a journal at random out of WoS and pick a reference at random, go to that cited item (journal) and continue to pick a cited ref at random.  Imagine you do this across the whole WoS collection in the past 5 years…  and do this for a really, really long time  - Millions of steps in this “random walk”  -  there will be some journals you are visiting more than others.  Lots of stuff cites NATURE – so you have a greater likelihood of randomly being directed back to NATURE.   So…the EF is a reflection  of the “time” you would spend with a  journal in this hypothetical random walk.  Several other metrics have used similar kinds of logic – though they differ in the fine points of the mathematical techniques

Eigenfactor™的利用 http://www.osa.org/Publications/ISI_Rankings.aspx Publishers are starting to use it, but they still don’t quite know what it means http://www.osa.org/Publications/ISI_Rankings.aspx

一个例子 REV BRAS FARMACOGN: 区域性期刊 2009年第一次影响因 3.462 以Impact Factor排序： 2009年第一次影响因 3.462 以Impact Factor排序： 8 of 46 in Chemistry, Medicinal 56 of 236 in Pharmacology & Pharmacy 以Eigenfactor™ 排序： 44 of 46 in Chemistry, Medicinal 217 of 236 in Pharmacology & Pharmacy High Impact Low/no influence in the citation network The JCR was enhanced by these additional metrics precisely because they can reveal very different things about a journal.   Taken in the context of the full JCR data, we have a high impact, a high level of self-citation that is shown to be the major component of impact, a low “influence” in the network because network structure doesn’t include self-citation.  Finally, we REPORT the self-citation practices so prominently that people can actually see a journal that is compromised this way.    This is exactly how the system is supposed to work.  The JCR is not only transparent itself, it is embarrassing transparent for the journals it contains.

论文影响力评分(Article Influence Score) 考虑了5年间的引用数据– 并将期刊发文量考虑在其中： Eigenfactor AI = (Journal items in prior 5 ys)/(all items in prior 5 yrs)

THE COMPETITION – SJR AND SNIP Logically similar to EF – but they use subtly different math to resolve the network dynamics. h-index is fundamentally misleading here…because of what is missing… EF, but not SJR offers its pseudocode for free. If you have a citation matrix, you can calculate the EF’s using the freeware provided But not SJR. Presents a lot of numbers – but no underlying data…

SJR – Scimago Journal Rank

SJR - in brief a minimum prestige value from simply being included in the database; a publication prestige given by the number of papers included in the database a citation prestige given by the number and “importance” of the citations received from other journals.

SNIP – Source Normalized Impact per Paper There are several “normalizations” involved in this… Normalize based on the number of cited references in the paper

SNIP – a simple ratio Okay – this is a bit glib. My goal here is to get you to grapple with this – both as a real thing and as a competitive threat. It is, in fact, a VERY clever metric and it very, very cleverly portrayed. What does it really do, and what does it mean? SNIP is a citation metric based on citation counts per paper – kind of like aJIF, that is adjusted every which way from Sunday. It adjusts impact for the citation potential of a subject – meaning it considers the mean number of cited references in all papers that cite a journal (so subjects with smaller numbers of references on average – like Math – will have a greater weight applied to each reference) – article or subject citation potential It adjusts impact based on how much of that subject is covered in Scopus (at the journal level) – so for each citing paper, it considers how many of the citations are to covered materials in Scopus. The less of that material Scopus covers, the more weight each citation has. (Database citation potential) The RIPj is, essentially, the 3 year Impact Factor…cites to the prior 3 years divided by “articles and reviews” published in those years – LINKED…all and only based on linked citations. The fact is we have to take this seriously – “Fractional citation counting” is a hot topic right now. People want a way to compare journals in different fields using a meaningful adjustment for the different citation behaviors and different coverage in those fields.

SOME TAKE AWAY MESSAGES

图书馆员 出版商 研究人员 编辑 学生 信息分析师 开发和管理期刊馆藏； 测定期刊的市场影响力； 协助其决定从馆藏中选择或删除期刊，并对期刊存档周期做出科学的计划。 出版商 测定期刊的市场影响力； 比较期刊的学术性，从而确定其竞争策略。 研究人员 识别最恰当、最有影响的期刊发表其文章； 确认所发表文章的期刊的学术地位。 编辑 评审编辑策略对期刊学术性指标的影响。 学生 发现与其研究领域相关的刊物； 挑选适合自己阅读的刊物。 信息分析师 跟踪文献计量学的发展趋势； 研究引文关系模式和分布特征。

JCR的核心价值 JCR数据库中出版的期刊影响因子(Impact Factor)已经成为全球公认的评价期刊学术影响力的重要指标；但是影响因子并不适用于对单篇科技文献和单个科研人员的评价。 任何希望在多学科平台中评价期刊重要性和影响力的用户，都能够在JCR数据库中找到他们需要的信息，包括：图书馆员、出版商、编辑、研究人员、信息分析师……； 唯一的期刊引文信息分析工具

THANKS 2012 Feb