Previous | Contents | Next | ||
Issues in Science and Technology
Librarianship |
Fall 2000 |
SciFinder Scholar gives students and faculty the ability to search the Chemical Abstracts family of databases for the years 1967 to the present. This search service has been available to academic institutions for a little over two years, whereas the commercial version, simply called SciFinder, is celebrating its fifth anniversary. Some commercial features were not included in SciFinder Scholar when it first became available, though some of these features have now been added. This review highlights ways that SciFinder Scholar distinguishes itself from other databases popular with chemists, and traces SciFinder Scholar's evolution over the past two and one-half years.
SciFinder Scholar currently provides seamless searching of the CAS family of databases: CAplus is an enhanced bibliographic file containing the entire Chemical Abstracts database from 1967 to present, and is updated weekly. The CAplus file includes all of printed CA as well as records for additional materials, such as American Chemical Society national meeting abstracts and citations from the 1,580 journals that CAS now abstracts cover-to-cover, but which do not meet the criteria for inclusion in the print index or the STN CA file. Registry is a database of chemical structures and names indexed by CAS since 1965 and partially retrospective before that time. CASREACT contains records of single and multistep chemical reactions from the Organic and Organometallic sections of CA, drawn from journals (1985-present) and patents (1991-present). CHEMCATS is a database of commercial suppliers for chemical compounds. CHEMLIST contains regulatory information records from 1979 forward. MEDLINE, added to SciFinder Scholar with the 2000 release, allows searching by Registry Number as well as keywords and subject headings.
Scholar is searchable by chemical structure, chemical name, CAS Registry Number, molecular formula, chemical reaction, research topic, author name, or specific reference (abstract number or patent number). Some enhancements were added with new releases of the software, while other enhancements took place without a new software release; these enhancements will be reviewed chronologically.
The first option, Chemical Substance or Reaction, gave the searcher the choice of drawing a chemical structure or entering a chemical name, CAS Registry Number, or molecular formula. After drawing the chemical structure, the Get Substances option searched the REGISTRY file for the exact structure as well as compounds in the same family, such as isotopes, stereoisomers, mixtures, salts, and polymers. Selecting Get Reactions executed a substructure search for the compound as a reactant or product in CASREACT. Version 1 did not include the option of searching chemical substructures in the REGISTRY file, and only one compound at a time could be searched in the CASREACT file.
Results from a chemical structure, chemical name, CAS registry number, or molecular formula search displayed the structures of all compounds matching the search. Clicking the Get References button displayed the CAplus references for the corresponding compounds. These references could be limited to preparations, reactions, and/or biological study.
The Research Topic option provides a box to enter a natural-language phrase to describe the topic. Scholar automatically combines terms, selects the appropriate database, and looks for variations (especially truncation), synonyms, and abbreviations of the terms. It is one of the few databases (chemical or otherwise) that processes natural language terms in a way that does not require the searcher to use special syntax. Depending on the number of terms and combinations it finds, Scholar provides the option of searching on some or all of the terms closely associated with one another, some or all of the terms present anywhere in the reference, or searching each term individually. The results can be refined further by adding an author name, document type, or publication year(s).
The third option, Author Name, asks for the author's last name and the first and middle names or initials, if known. Scholar automatically looks for spelling variations of the author's name.
The Specific Reference option allows a search on the CA accession number or patent number. While the user would not normally know these numbers ahead of time, this option comes in handy if more detailed information has to be retrieved later.
The references retrieved are displayed in reverse chronological order in a brief format that includes the author(s), title, source, CODEN, ISSN, accession number(s), and file name (CAplus). Clicking on a microscope icon to the right of each reference opens a window showing the full record, including the abstract and indexing terms and associated Registry Numbers. References may be printed or saved in the compact (title and accession number), standard (bibliographic information), summary (bibliographic information and abstract), or full (bibliographic information, abstract, and indexing) formats, although the summary format is the default.
Chakraborty, Manisha; McConville, David B.; Koser, Gerald F.; Tessier, Claire A.; Saito, Takeshi; Rinaldi, Peter L.; Youngs, Wiley J. A Novel High-Yield Synthesis of Substituted Isoindolequinones. J. Org. Chem. (1997), 62(23), 8193-8197. CODEN: JOCEAH ISSN:0022-3263. CAN 127:346265; AN 1997:724097 CAPLUS
References may be saved in a rich text, ASCII text, quoted, tagged, or answer keys (accession number) format. The various formats enable the user to view the references in a word processor (and retain all of the formatting, including the chemical structures), a spreadsheet, or bibliographic software such as EndNote, ProCite, or Reference Manager. Filters are available to aid in transferring the references to bibliographic software such as EndNote, ProCite, or Reference Manager.
In addition to refining results by author name, document type, and/or publication year(s), Scholar now allowed the searcher to refine by additional research topics and/or language, very welcome enhancements. Structure searches could be refined with an additional structure, allowing the user to specify the second component of a multicomponent substance. Similarly, a reaction search could be refined by drawing a second compound and searching them as two reactants, two products, or a reactant and a product simultaneously.
Scholar now provided links, through CAS' ChemPort web site, to the electronic full text of journals from participating publishers, such as AAAS, Academic Press, the American Chemical Society (ACS), the American Institute of Physics, the Institute of Physics, and the Royal Society of Chemistry. The list of publishers has grown since then, but noticeably absent are Elsevier, Wiley, and JSTOR. The "Full-Text" button in the Scholar window would launch a web browser and, using metadata, retrieve the PDF or HTML full text of the article in question, provided that the user's institution had an online subscription to that title.
At this point, Scholar cost $15,000 per year for off-peak access for the first user and $6,000 per year for off-peak access for each subsequent user. To add daytime access cost an additional $10,000 per year for the first user. Adding substructure searching cost an additional $7,500 per year for the first user during the off-peak time, and more during the daytime hours. These charges were in addition to the cost of a print subscription to CA, which cost at least $20,000 per year for schools granting a Ph.D. in chemistry and $10,000 per year for smaller schools. Very few schools could afford 24 hour-per-day access with substructure searching for enough concurrent users to satisfy the demand on their campuses. Most schools had to decide between adding 24 hour-per-day access or substructure searching, or adding one user with 24 hour-per-day access and another user with off-peak substructure searching. For the latter arrangement, two separate passwords were needed, which in turn required two separate installations of the software on each computer.
When searching for the more common chemical substances in the REGISTRY file portion of Scholar, a hyperlink labeled "Commercial Sources" indicates substances that appear in the CHEMCATS supplier file. Following this link displays a list of commercial suppliers of the substance. A separate hyperlink labeled "Regulated Chemicals Listing" indicates that the substance appears in the CHEMLIST file, which lists the national chemical regulatory inventories of the United States, Canada, Europe, Australia, Japan, Korea, or the Philippines.
The "analyze references" feature helps reveal trends in the literature retrieved and can help narrow large results sets. One can analyze by author name, CAS registry number, CA section title, company/organization, document type, index term, journal name, language, publication year, or supplementary term. Analyzing by journal name is one of the most useful choices, especially when verifying an incomplete or questionable reference. The Analyze tool can be applied to all of the references retrieved or just to selected ones; the resulting histogram can be sorted alphabetically or by frequency of occurrence.
Highest Chemistry Degree Offered | Annual Price of First User without the Substructure Module | Annual Price of First User with the Substructure Module |
---|---|---|
Doctoral | $26,250 | $38,050 |
Master's | $19,690 | $28,540 |
Bachelor's | $13,125 | $19,025 |
An academic institution or user can subscribe to the task-based, commercial version of SciFinder for $6,500 per year for 310 tasks. Each search is considered a task, and a substructure search counts as four tasks. However, this quota of tasks would likely be expended in a very short period of time.
SciFinder Scholar 2000 offers the following enhancements:
The primary purpose of SCI is to find cited references, although many users search it for authors and topics. The Web of Science indexes approximately 5,700 journals in all areas of science, engineering, and medicine; about 600 of those journals pertain to chemistry. CA indexes approximately 8,000 journals, many of which are in related disciplines, but certainly covers many more chemistry journals than SCI. CA also indexes patents, conference proceedings, dissertations, technical reports, and books. Scholar covers the years 1967 to the present whereas the online version of SCI covers the years 1974 to the present. The years covered by Web of Science depend on an institution's subscription.
Since SCI does not use controlled vocabulary to index its references, it is difficult to predict how a chemical substance may be cited in the title or abstract. Which synonym did the author use to describe the compound? SCI recently added a reaction module to the Web of Science for an extra fee, but few schools have added this feature. While Scholar now displays cited references in records from 1999 forward, they are not searchable.
Beilstein is used to find properties, preparations, and reactions for organic compounds, and Gmelin is used to finding similar information for inorganic and organometallic compounds. Each of these handbooks/databases goes back to the 1770's and therefore covers a much larger time span than Scholar. Finding a melting point or spectral reference in these sources is much easier than in Scholar, even with the latter's additional searching capabilities such as properties or spectral properties. Each of these sources has its own indexing criteria, and often CA does not index something unless it is covered significantly in the original document. Furthermore, Beilstein and Gmelin index far fewer publications than CA. But, because of the vast difference in time coverage and indexing philosophy, these databases should be considered complementary.
Another enhancement would be to more clearly distinguish the journal title from other parts of the reference, especially the preceding organization address. One library patron recently asked if the library had the journal Canadian Inorganic Chemistry, abbreviated "Can. Inorg. Chem." When this entry was not found in CASSI, we realized that "Can." had nothing to do with the name of the journal, but rather was the location of the author's organization. Another example is confusing "Chem. Comm." with "Coll. Czech. Chem. Comm.", when a user does not realize that "Coll. Czech." is the beginning of this journal's abbreviated title.
The commercial version of SciFinder provides an option of setting the software's Preferences to limit retrieval in substructure searches, eliminating isotopes, mixtures, polymers, metal-containing substances, and incompletely described substances. Scholar users do not have this option. However, being able to "refine" substructure searches with these limits would greatly help reduce the number of hits substructure searches now retrieve. CAS has advised librarians that they plan to address this issue in the next release of the software. We would also like to have the option of limiting all structure searches, exact or substructure, to single-component substances.
To make full-text retrieval more comprehensive, CAS should redouble its efforts to add full-text links to JSTOR, Wiley, and Elsevier journals.
Adding the CAOLD file to Scholar would be helpful, but may also be complicated for users since the older file has far fewer access points, and the full references are displayable only in TIFF and PDF formats. CAOLD is a partial retrospective file covering the pre-1967 indexing period.
CAS has made vast improvements to Scholar over the past two and one-half years, and now offers more attractive pricing. Academic librarians are very pleased that they now have most of the features offered to their commercial SciFinder customers, and there is every reason to believe that Scholar will continue to be enhanced in the future.
We welcome your comments about this article. Please fill out this form for possible inclusion in a future issue. |
Would you like to be notified about new issues of ISTL? Join our mailing list. |
Previous | Contents | Next |