How to Find Chemical Information, Part II

Another highlight from Maizell's How to Find Chemical Information...

Bad synthesis indexing (Ch. 7.7.K).  I knew it!  Searching SciFinder Scholar for synthesis methods can be frustrating because of the number of false hits.  Lots of articles may pop up when you do these searches, but often many never really discuss the synthesis.  I've wondered why this was.  Was I conducting bad searches?  Do I simply not understand the articles?  (I always have to admit the possibility that there's a conceptual gap between my chemistry understanding and the inferences of a chemistry research article.  Always.)  I'm feeling better about my librarian skills now, knowing that there's a reason for this: beginning in 1983 a 'P' designation for preparation methods was included on relevant article records, determined by a computer algorithm analyzing content.  Sounds good--if only it had worked.  A lot of articles would get the P for having something on preparation, and something on your chemical, but not necessarily both at the same time.  And sometimes it was even less accurate. 

The result is that, essentially, when you search for the synthesis of a compound, SciFinder returns a list of articles discussing the compound, some of which will be what you want.  You can avoid wasting time reading the wrong articles by first reviewing the abstract and indexing for returned articles, looking specifically for the chemical and its role in the study, but that has always seemed sort of ridiculous and not teachable.  Why not improve the algorithm?  In 1994, they did.  Literature added since then is indexed with preparatory terms only by the document analyst (i.e., a human scanning the article). 

Maizell reports that CAS staff believe the preparation indexing was 99% accurate.  How they arrive at this term is not mentioned, but based on my own experience I find it hard to believe.  Nor is there any indication they fixed all the false indexing created during the Reign of the Synthesis Algorithm.

Good to know.

Posted at 02:26PM Feb 28, 2007 by WILSON, JOSHUA in Maizell  | 

How to Find Chemical Information, Part I

I've been reading through Robert Maizell's How to Find Chemical Information to further my chemistry librarianness.  The book is nine years old at this point and therefore dated in many respects, but most of the concepts and history are valuable.  I'm still learning a lot about chemistry research myself, my own background being more astronomy, physics, computer science, and mathematics.  I figured I'd share some of the interesting tidbits I pick up as I proceed, for my edification and yours.

Translations (Ch. 5.7).  People in all sciences sometimes need foreign language articles.  Sometimes they optimistically think that we can always get an English copy of what's needed.  Having to break their spirit by explaining this isn't the case makes me wish this was true.  When you can't find a translation, there are still solutions to the problem, though, even if you're not friends with a bilingual colleague:

1. Read the abstract only.  Often this is in English, especially if indexed by CAS or another service.  Maybe this is enough to go on.  Or at least enough to know how much time you should devote to finding this article. 
2. Find equivalent journal articles, reviews, or books.  Try patents, too, especially if you're looking for a foreign patent.  Often an English version will appear later.
3. Scan tables, graphs, equations, and figures to pick out some of the relevant portions.  If nothing else, maybe that will help determine where to work with a foreign language dictionary on translation.
4. Purchase a custom translation.  There are services which do this for you for a fee, and it probably ensures more accuracy than doing it yourself with a dictionary.  [Maizell doesn't mention a similar suggestion I usually give people: hit up the foreign language department at your university.  Maybe you can get a student to translate it for you cheaply, or get it for free by having their instructor make them do it!]

Posted at 01:58PM Feb 28, 2007 by WILSON, JOSHUA in Maizell  |