Chemical Literature (Chem 184/284)

Lecture 17: The New Web and Chemical Information

What is the Internet...really?

• The Internet is a network of networks; a linkage of computer systems around the world using common protocols to facilitate communication and dissemination of information.
• The Internet grew out of previous networks designed for military and academic users (UCSB was involved early.) Eventually, the Internet became the world's data communication highway for all sorts of users.
• Now, the maintenance of the Internet backbone has been passed to commercial firms, and some academic and government users are developing new-generation systems for their own use.
• The protocols that define the World Wide Web, and new forms of "social software" are driving the information revolution forward.

World Wide Web: the hottest part of the Internet

• The World Wide Web is based on a protocol designed in 1989 at the CERN laboratories in Switzerland by Tim Berners-Lee (now at MIT). It allows for hypertext linkages among documents on the Internet and for transfer of text, binary files, graphics, audio, video, etc. The name of the protocol is HyperText Transfer Protocol. That's why you see http: at the beginning of all URL's (which stands for Uniform Resource Locator)
• Links are created using HTML: HyperText Markup Language.

HTML: What it does...

• HTML is an outgrowth of SGML (Standard Generalized Markup Language), a markup system developed for publishing.
• HTML lets the author label the content of the document so that it can be manipulated by the viewer's software. The labeling is done with tags, which use plain text words or abbreviations within brackets. Example: <strong> signifies that the text following it should be in boldface. </strong> ends the boldface text. Other types of tages can be used to label headings, titles, paragraphs, create outline form lists, designate rows and cells in tables, etc.
• Most importantly, it lets the author embed links to other documents on the World Wide Web.
• HTML is a constantly evolving standard, with official versions promulgated by the World Wide Web Consortium (W3C) organization. In recent years, a superset of HTML has been defined called XML (from Extensible Markup Language), which allows programmers to define formats for any type of data, e.g. MathML. Special versions of XML are being devised for various types of chemical information as well. See the Chemical Markup Language site at http://cml.sourceforge.net/ for more information.

Web Browsers

• To use the Web, you need a class of software called a Web browser.
• Early Web browsers were text based (e.g. Lynx). They could handled hyperlinks among documents, but the documets could only be plain text files.
• Mosaic, developed around 1993 at the National Center for Supercomputing Applications at the University of Illinois had two new important features which tremendously accelerated the growth of the World Wide Web:
  • It could handle graphics and other files.
  • It was distributed free of charge.
  Mosaic's creators founded Netsscape. The Netscape browser was, for a time, the world's most popular, until it was eclipsed by Microsoft's Internet Explorer
• There are now a wide variety of web browsers available. The two most popular are Microsoft Internet Explorer (IE) and Mozilla Firefox (MF), but there are a host of other browsers out there (e.g. Opera, and Safari, which is very popular in the Mac community.) Mozilla's software is "open source code"; the current version of Netscape is based on the Mozilla code.
• Both IE and MF come in a variety of forms for a variety of platforms (Windows, Mac, Unix, etc.), though there is some indication that Microsoft may limit further development of IE to the Windows Vista platform and its successors.
• Modern Internet browsers will allow the user to not only to display text files with hyperlinks (.htm or .html) but will handle a variety of graphic file (.gif, .jpg) Other types of files (audio, video, special chemical files) may require helper applications which are separate programs which the browser can open to handle other filetypes, or plug-ins, which operate "transparently" within the browser itself.

Locating Resources on the Internet

• Like print resources on the library shelves, Internet resources can be located either through classification and data collections or through indexes to the available files.
• As with other data collections and indexes, quality and usefulness can vary widely depending on your needs.

Data Collections: UCSB Libraries Website
http://www.library.ucsb.edu/

• The UCSB Libraries web site provides links to a wide variety of tools and subject resources.
• Librarian subject specialists have selected and organized the links for easy browsing.
• In many cases, the resource links are annotated to help you evaluate them.
• Help by Subject: Chemistry at http://www.library.ucsb.edu/subjects/chem/chem.html and its associated guides: So far, most links are to useful sites at remote locations. As time goes on, more pages developed at the Library will make original content available.

Web Indexes/Search Engines

• Numerous web searching tools have been made available. Most are free to use (advertiser-supported); some require subscription.
• The UCSB Libraries web site has links to indexing tools. Internet Search Tools at http://www.library.ucsb.edu/tools2.html
• Be aware of different approaches:
  • Some search the titles and/or headers and/or full text of items: Google, Ask.com, etc. Note that even "basic" search engines differ in exactly how they search, how big their collection of indexed pages is, how often they update, how they sort your results, etc. It is often worth trying out several search engines to see which ones seem to work best for your search.
  • Some specialize in hierarchical subject collections: WWW Virtual Library
  • Some have searchable subject collections: Yahoo
  • Some evaluate their sources: InfoMine.
  • Most of the above now combine different approaches to searching, or link to one another to allow the user greater flexibility.
  • Meta-search engines search multiple search engines at once.
• If you have a very specific term, search engines can be quicker...if they find it at all. But common words, or words which have multiple meanings may lead to large numbers of false hits, burying the items you wish to find.
• If you have a general term, hierarchical systems may be better in avoiding unrelated material. Hierarchical systems may also find related material of interest.
• See also Lecture 3 on Web search engines vs. library catalogs and article databases.
• IMPORTANT: Be aware that nearly all general Web search engines miss the "invisible" Web, which includes all sites which block web robots, and sites which required subscription or other payment for access, e.g. most scientific journals on the Web. However...

Google Scholar (http://scholar.google.com/)

• Google Scholar is a new initiative by the folks at Google to make scholarly information (jouranl articles, books, dissertations, preprints, etc.) available through the familiar Google interface. See About Google Scholar for their mission statement.
• Google is contracting with major publishers to allow the Google indexing robots access to their subscription websites, and with the OCLC WorldCat library catalog, in addition to sites on the publicly-accessible Web (e.g. PubMed, arXiv, government technical reports collections, open access journals.)
• In addition to the documents themselves, Google Scholar extracts the cited references from the documents and so can "find" documents which are not themselves available on the Web.
• Some studies have been conducted comparing the results of Google Scholar searches with those in more traditional broad-scope indexing services, such as Web of Science. See, for instance, Bauer, Kathleen and Bakkalbasi, Nina, "An Examination of Citation Counts in a New Scholarly Communication Environment", D-Lib Magazine, 11(9), Sept. 2005 at http://www.dlib.org/dlib/september05/bauer/09bauer.html. So far, results are mixed.
• Note that Google Scholar searches, like Google searches, can give variable results from one day to another, for reasons other than the growth of the database. Google Scholar is more stable than its general counterpart, but not so much as a traditional indexing database.
• The University of California is collaborating with Google to provide UCe-Links connections for Google Scholar searches done from UC IP addresses (though the familiar white-and-gold UCe-Links icon does not appear.)

Google Book Search (http://books.google.com/)

• Google Book Search is an ongoing project from Google (still considered in beta test) to index books and provide full-text access to books which are either out-of-copyright, or for which permission has been obtained from the copyright holders to make them available online.
• Here's an example of a Google Book Search on fullerenes. It retrieves over 1100 titles. Notice that not all of the books have "fullerenes" in the title; the indexing digs into the content of the books. Note, too, that since this is a fairly recent topic, the vast majority of the books do not provide full text online, but only a limited preview.
• Here is an example of a result for which the full text is available. Note that the link took us to a page on which the search term is used.
• Google has partnered with a number of academic libraries, including the University of California, to begin digitizing and indexing their out-of-copyright collections and make them available on the Web. These include some of the largest libraries in the world. Eventually, millions of books will be available through Google Book Search.
• Microsoft is engaged in a similar project, with many of the same institutions as partners, but it is not as far along.
• Google's digitization project has led to considerable controversy. Publishers distrust Google's willingness to respect copyright laws. Officials in non-English speaking nations have expressed fears that this project will marginalize non-English literature, and have moved to create their own national digitization projects.

Google Patent Search (http://www.google.com/patents?hl=en)

• Google has undertaken to digitize the full body of United States patents, creting PDF images of the patents, and using optical character recognition to index the text.
• Over 7 million patents have been digitized. The project is still formally in beta test, and there is no indication as yet that Google will expand its collection to non-US patents.
• Note that the OCR process is subject to errors, so searching older patents can be a hit-or-miss proposition, even over and above the challenges of dealing with patent language.

Google Images (http://images.google.com/)

• Google Images specializes in searching for images: photographs, but also tables, graphs, charts...and structure diagrams.
• Google Images depends on how the images have been labeled on the original HTML document, so don't expect a comprehensive search of all images of, say, ferrocene, on the Web. But it can be useful for finding structure diagrams, illustrations of laboratory procedures and the like.

Specialty search engines for science research

• ChemRefer (http://www.chemrefer.com/)
  ChemRefer is designed for text-based searching of chemical articles on the Web.
• ChemSpider (http://www.chemspider.com/)
  ChemSpider provides chemical substance-oriented searching of the Web. Structure input is via name, SMILES string or InChi string. Direct structure drawing input is not yet available.
• eMolecules (http://www.emolecules.com/)
  eMolecules is a search engine designed specifically to search for chemical substances on the Web. You can search using chemical names, CAS Registry Numbers, SMILES strings (see below), or, using a Java structure drawing editor, search structures or substructures (which are automatically conversted to SMILES strings.) It should not be used for text searches of chemical concepts (general search engines, e.g. Google are recommended for those searches.)

  SMILES is a code developed to represent chemical structures with a string of ASCII character symbols. A possible SMILES representation of aspirin, for example, is CC(=O)Oc1ccccc1C(=O)O For more information see: The SMILES Home Page at http://www.daylight.com/smiles/.

  At present, eMolecules mostly retrieves links to online chemical catalogs, but it is growing in scope.

• Scirus (http://www.scirus.com/)
  Scirus is owned by Elsevier Science Publishers, it searches not only the Web at large, but many of Elsevier's proprietary files, such as the e-journals in ScienceDirect, Beilstein Abstracts and others. Since the UC system subscribes to ScienceDirect, most of what you find will be accessible here.
• eScience (http://escience.cas.org/cgi-bin/escience)
  eScience is the Web search engine which SciFinder Scholar links to if you "Get Related" for Web resources. However, it may also be searched directly. It provides four search sources: Google, chemistry.org (the American Chemical Society website, see below), ChemIndustry.com, and ChemGuide. ChemIndustry.com (a database of over 42.000 chemistry-related Web sites), and ChemGuide (an index to full-text sources on the Web created by FIZ Chemie Berlin, and which seems to be strangely unreliable of late.) All of these sources can be accessed directly.
• For more examples, see the page Science Search Engines at Search Engine Guide http://www.searchengineguide.com/

Alerting services and RSS

• Most sources of information on the Web depend on the user going out and tracking them down. This is sometimes referred to as "pulling" information. Services which send desired information to the user automatically are said to "push" information. Many of the electronic databases we have examined have some form of alerting service which allows you to set up a search, and then automatically receive selected pieces of data on a regular basis without further effort. Some online journals allow you to receive tables of contents; some online newspapers have e-mail news features.
• Recently, a new approach to "pushing" information out to users has come into vogue: RSS The following description is borrowed from Engineering Village 2 which has an RSS feature for Compendex and its other databases:
  • "RSS stands for Really Simple Syndication or Rich Site Summary. It is an XML-based format for content syndication. RSS is a way of publishing and distributing content from one Web site to another. It's an easy way for you to keep updated automatically on web sources that you follow."
  • "You need to have an RSS reader or aggregator to make use of RSS. RSS reader will display information feeds from your selected sites on your computer without visiting each Web site. You will automatically receive the most current information whenever these sources are updated. It will also allow you to share information with others in your research or study group."
  • "With an RSS aggregator, you will be able to find technical information from Engineering Village 2, latest news from your library (if your library is providing an RSS feed), or latest technology news from the New York Times."
  • "There are several types of RSS readers. Some are Web-based, such as My Yahoo, Bloglines or NewsGator, some are extensions of Web browsers such as Mozilla Foxfire and some are desktop readers like FeedDemon or Awasu. You can find a list of RSS reader and other information about RSS at the RSS Compendium." [see: http://allrss.com/index.html]
• The main advantage for the user of RSS over more traditional alerting services is that your RSS aggregator can pool RSS feeds from a variety of sources into a single location for you.

Blogs and Wikis: What are they, and should I care?

• Both blogs and Wikis are part of the general class of social networking software, which includes phenomena like MySpace and FaceBook. While the latter two haven't made strong inroads into the sciences, the former two are gaining in importance, especially in science education. See, for example, this artilce from Chemical & Engineering News about a 2007 symposium in the ACS Division of Chemical Education.
• Blog is short for Weblog. It is a class of software to allow easy publication on the Web, readily organizable, and with the possibility of allowing other users to add comments to your publication. For a capsule description of blog basics see What Are Weblogs? at http://www.userland.com/whatAreWeblogs. For an extensive collection of information on blogs, see Weblogs Compendium at http://www.feedbus.com/weblogs/
• Most blogs are sort of personal diaries/newsletters/collections of random thoughts. For a chemistry-related example, see Prof. Mark Winter's . However, blogs are being used for some types of systematic scientific information, replacing more static lists of web sources. See, for example, About: Chemistry.
• Some journal publishers are launching blogs at their websites to facilitate communication among their authors, editors and readers. See, for example, the BioMedCentral, Chemistry Central, and Nature blogs.
• Here's a list of chemistry-related blogs found by searching at BLOG Search.com (http://www.blog-search.com/).
• Wikis are a concept which has been around for a surprisingly long time: the first was created in 1995. Wikis are designed to allow easy creation of web pages. Like blogs, other users can comment on the contents of a wiki, but unlike blogs, any user can edit the content of a page. They can be used for communities of common interest to build shared resources. For more information, see the article "Wiki" in the Wikipedia at http://en.wikipedia.org/wiki/Main_Page.
• The Wikipedia is a good example of a resource being built by its user community. Anyone can add articles or edit existing articles in this "free encyclopedia". However, an informal community of chemist-volunteers has been working hard to make sure that quality chemistry information is being added to Wikipedia. They have created the Wikipedia Chemistry Portal at http://en.wikipedia.org/wiki/Portal:Chemistry, a great place to start exploring the chemicstry-related resources to be found there. Recently, Chemical Abstracts Service agreed to begin verifying the CAS Registry Numbers listed for major chemicals in Wikipedia aprticles.
• As with blogs, some journal publishers are adding Wiki sites to their online journals. See, for example, the Wiki at ACS Chemical Biology.
• Some information scientists have proposed using wiki software as an alternative, "free", method of scientific publication, which would allow an author to publish, peer reviewers to critique, editors to edit and readers in general to comment, all easily and interactively. Such concepts have not yet been widely tested.
• Note that blogs and wikis, perhaps even more than the average Web site, are subject to clutter with garbage. A critical eye is needed to separate the gems and gold from the trash. Even for generally high-quality products such as Wikipedia, a good rule of thumb is "Wikis can be a good place to start a search, but they should never be where you finish your search."

Podcasts -- the audio/video Web

• Podcasts are the popular term for audio and video segments "broadcast" via the Web, and viewable/listenable on devices like the iPod, as well as more full-featured computers.
• In addition to the recreational uses of podcasting, there are now many sources of scientific information being released as podcasts:
  • ACS Chemical Biology has a podcast site for materials related to their articles.
  • The Royal Society of Chemistry's journal Chemistry World has a podcast version.
  • Nature has a series of subject-oriented podcasts, including chemistry@nature.com
  • Chemical Abstracts Service is making e-seminars on their various products available.
  • Distillations is a series of podcasts on the past present and future of chemistry presented by the Chemical Heritage Foundation
  • Some universities are making lectures and other course materials available as podcasts. For example, the chemistry department at Imperial College in Great Britain has a series of lectures by and about famous chemists.
  • The online site, Podcasting News, maintains a (less than complete) subject directory of podcasts, including one for chemistry
• Some conferences are recording their symposia and publishing them on the web as podcasts.

Social bookmarking -- sharing your favorites

• del.icio.us (http://del.icio.us/)
  The first of its type (2003), this site is probably the best known social bookmarking tool (in fact, they coined the term! It was so successful that it was acquired by Yahoo!
• Simpy (http://www.simpy.com/)
• Furl (http://www.furl.net/)
  Now owned by LookSmart.
• StumbleUpon (http://www.stumbleupon.com/)
  This service makes use of a browser plugin to enable members to find and recommend (or dis-recommend!) websites. Now owned by eBay.
• Diigo (http://www.diigo.com/)
  This service allows you to not only collect entire Web pages, but to link to and annotate specific portions of Web pages.
• Ma.gnolia (http://ma.gnolia.com/
• CiteULike (http://www.citeulike.org/
  This system creates a personal database of documents from electronic journals. The database is, however, visible to other users, so, in effect, you share your favorite research sources on a topic with the world.

Types of information on the Web

• Publishers & societies
  Information for authors, members
  Example: American Chemical Society operates the a portal for members and other chemists at http://portal.acs.org/portal/acs/corg/content.
• Career and job information
  Besides general interest job-finder sites like Monster.com at http://www.monster.com, there are some specialty sites. One relatively new one is ACS Careers at http://portal.acs.org/portal/PublicWebSite/careers/index.htm. It provides job listings in chemistry and chemical engineering, with priority access given to American Chemical Society members. The Royal Society of Chemistry provides similar services at its RSC Careers at http://www.rsc.org/Gateway/Subject/Careers/.
• Academic institutions
  Faculty, contact information
  Example: UCSB Chemistry Dept. at http://www.chem.ucsb.edu
• Educational materials
  Example: ChemDB from UC-Irvine.
• Government-produced information
  Lots of NASA data, Dept. of Commerce, US patent info. Some is on US govt. servers, but some is on other sites maintained by academic institutions or even private companies.
  Examples: NIST Chemistry Webbook at http://webbook.nist.gov
  Data tends to be "plain vanilla"; sites which add value to government data tend to charge for their services. One exception to this is genome related data; many sources have put very useful information up free on the web.
• Corporate pages
  • Product catalogs (e.g. the Sigma-Aldrich catalog) & advertising
  • Contact info; sometimes job listings
  • Technical support information
  • Almost every major company has some sort of "presence" on the Web nowadays, but the quality of Web sites varies dramatically.
• Listservers, discussion groups and other e-communities
  • The Internet's original function -- facilitating communication among scientists -- is still one of its most important ones. Many areas of study have thriving electronic communities, through which members can share information, seek expert help with problems and generally network.
  • Listservers act as e-mail distribution points. You send your message to the central server and it gets reflected out to all subscribers. Some lists are moderated, i.e. someone reviews all messages for appropriate content before they are posted to the mailing list. Many lists have the option to receive mailings in digest form (all of a given time periods messages packed into one long message) and many have archives. You can find lists of chemistry-related listservs at L-SOFT at http://www.lsoft.com/scripts/wl.exe?qL=Chemistry&F=L&F=T, Links for Chemists at http://www.liv.ac.uk/Chemistry/Links/maillists.html, and CHEMINFO at http://www.indiana.edu/~cheminfo/listserv.html.
  • User groups take advantage of "Usenet" for threaded discussions. To locate Usenet groups, go to Google Groups at http://groups-beta.google.com/, click on the "Science and Technology" link and search the Group Directory for chem, chemistry or biochem or other appropriate search string. Note that not all hits will be true science sites.
• Electronic journals: See the discussion in Lecture 7
• Downloadable software (some free!)
  • Many software providers use the Web to distribute their software; one notable example is Mozilla.
  • Chemistry software manufacturers have also gotten into the act. Symyx at http://www.mdli.com/ makes available the Chime software for interactive chemical structures on the Web., for example. Similarly, CambridgeSoft, the makers of ChemFinder have a site FreeChem.com (http://www.freechem.com where you can get demo versions of their products and free plugins.
  • Note that free software on the Web is often a beta release version of the software - in exchange for free software, you get to help find bugs in the product!
• Web interfaces to standard databases
  • Many commercial database providers are now providing Web access to their products. These interfaces are usually menu-driven and aimed more at end-users than professional searchers.
  • Most require subscriptions (Cambridge, SilverPlatter) or an account with the database provider (STNEasy, Dialog Web). As electronic commerce on the Web becomes more widely available, users may be able to do individual searches on such sites by charging them to a credit card number or using electronic cash.
  • A few, mainly government databases, are free. PubMed at http://www.ncbi.nlm.nih.gov/PubMed/, the free public version of Medline, ERIC (the education research database) at http://eric.ed.gov/, and the TOXNET cluster of databases at http://toxnet.nlm.nih.gov/ are good examples.

    PubChem at http://pubchem.ncbi.nlm.nih.gov/ is a relatively new project from the National Institutes of Health to create a chemical substance database to facilitate searching of NIH funded research. It has been the subject of some controversy in the chemistry community as to whether it is a government-funded rival to the CAS Registry database, and, if so, whether or not this is a good thing.