Web of Science


Web of Science is a website which provides subscription-based access to multiple databases that provide comprehensive citation data for many different academic disciplines. It was originally produced by the Institute for Scientific Information and is currently maintained by Clarivate Analytics.

Background and history

A citation index is built on the fact that citations in science serve as linkages between similar research items, and lead to matching or related scientific literature, such as journal articles, conference proceedings, abstracts, etc. In addition, literature which shows the greatest impact in a particular field, or more than one discipline, can be easily located through a citation index. For example, a paper's influence can be determined by linking to all the papers that have cited it. In this way, current trends, patterns, and emerging fields of research can be assessed. Eugene Garfield, the "father of citation indexing of academic literature," who launched the Science Citation Index, which in turn led to the Web of Science, wrote:

Search and analysis

Web of Science is described as a unifying research tool which enables the user to acquire, analyze, and disseminate database information in a timely manner. This is accomplished because of the creation of a common vocabulary, called ontology, for varied search terms and varied data. Moreover, search terms generate related information across categories.
Acceptable content for Web of Science is determined by an evaluation and selection process based on the following criteria: impact, influence, timeliness, peer review, and geographic representation.
Web of Science employs various search and analysis capabilities. First, citation indexing is employed, which is enhanced by the capability to search for results across disciplines. The influence, impact, history, and methodology of an idea can be followed from its first instance, notice, or referral to the present day. This technology points to a deficiency with the keyword-only method of searching.
Second, subtle trends and patterns relevant to the literature or research of interest, become apparent. Broad trends indicate significant topics of the day, as well as the history relevant to both the work at hand, and particular areas of study.
Third, trends can be graphically represented.

Coverage

Expanding the coverage of Web of Science, in November 2009 Thomson Reuters introduced Century of Social Sciences. This service contains files which trace social science research back to the beginning of the 20th century, and Web of Science now has indexing coverage from the year 1900 to the present., the multidisciplinary coverage of the Web of Science encompasses 12,000 high impact journals and 160,000 conference proceedings. The selection is made on the basis of impact evaluations and comprise open-access journals, spanning multiple academic disciplines. The coverage includes: the sciences, social sciences, arts, and humanities, and goes across disciplines. However, Web of Science does not index all journals.
There is a significant and positive correlation between Impact Factor and CiteScore. However, an analysis by Elsevier, who created the journal evaluation metric CiteScore, has identified 216 journals from 70 publishers to be in the top 10 percent of the most-cited journals in their subject category based on the CiteScore while they did not have Impact Factor. It appears that Impact Factor does not provide a comprehensive and an unbiased coverage of high quality journals. Similar results can be observed by comparing Impact Factor with SCImago Journal Rank.
Furthermore, as of September 3, 2014 the total file count of the Web of Science was 90 million records, which included over a billion cited references. This citation service on average indexes around 65 million items per year, and it is described as the largest accessible citation database.
Titles of foreign-language publications are translated into English and so cannot be found by searches in the original language.
In 2018, Web of Science started embedding partial information about the open access status of works, using Unpaywall data.

Citation databases

The Web of Science Core Collection consists of six online databases:
Since 2008, the Web of Science hosts a number of regional citation indices.
The Chinese Science Citation Database, produced in partnership with the Chinese Academy of Sciences, was the first one in a language other than English.
It was followed in 2013 by the SciELO Citation Index, covering Brazil, Spain, Portugal, the Caribbean and South Africa, and more 12 countries of Latin America;
by the Korea Citation Index in 2014, with updates from the South Korean National Research Foundation; and by the Russian Science Citation index in 2015.

Contents

The seven citation indices listed above contain references which have been cited by other articles. One may use them to undertake cited reference search, that is, locating articles that cite an earlier, or current publication. One may search citation databases by topic, by author, by source title, and by location. Two chemistry databases, Index Chemicus and Current Chemical Reactions allow for the creation of structure drawings, thus enabling users to locate chemical compounds and reactions.

Abstracting and indexing

The following types of literature are indexed: scholarly books, peer reviewed journals, original research articles, reviews, editorials, chronologies, abstracts, as well as other items. Disciplines included in this index are agriculture, biological sciences, engineering, medical and life sciences, physical and chemical sciences, anthropology, law, library sciences, architecture, dance, music, film, and theater. Seven citation databases encompasses coverage of the above disciplines.

Limitations in the use of citation analysis

As with other scientific approaches, scientometrics and bibliometrics have their own limitations. In 2010, a criticism was voiced pointing toward certain deficiencies of the journal impact factor calculation process, based on Thomson Reuters Web of Science, such as: journal citation distributions usually are highly skewed towards established journals; journal impact factor properties are field-specific and can be easily manipulated by editors, or even by changing the editorial policies; this makes the entire process essentially non-transparent.
Regarding the more objective journal metrics, there is a growing view that for greater accuracy it must be supplemented with article-level metrics and peer-review. Thomson Reuters replied to criticism in general terms by stating that "no one metric can fully capture the complex contributions scholars make to their disciplines, and many forms of scholarly achievement should be considered."