Chemical Literature (Chem 184/284)

Lecture 1: Overview of the Organization of Information

Because the subject is HUGE...
For example: Chemical Abstracts
- Indexes 14,000 journal titles, plus patents, conferences, reports, dissertations, etc.
- Adds over 500,000 citations and 700,000 substance records per year.
- From 1907 to present, over 13 million citations and 14 million substances indexed

Information can be treated as a thermodynamic system, subject to entropy (described by Claude Shannon)
The organization of raw data turns it into information -- the better organized, the more value added.
Organization can be added at many levels...including the ultimate user.

End-user information processing puts the information in its final form for use. A new task each time -- stoichiometric
Information professionals try to create organization in ways that can be used by many people -- catalytic

PRIMARY -- The original publication of data: journals, patents, technical reports conferences, dissertations, some books.
SECONDARY -- Publications which provide access to the primary literature: reviews, indexes, abstracts, data collections, etc.

Classification and Data Collection -- physically grouping related data by some common element.
Indexing -- creating pointers to the original literature based on some piece of information in the original, e.g. author names or subject terms.

Libraries use classification schemes to group related books together for browsing by subject. In the Library of Congress system, chemistry materials fall under QD.
Data collections bring information from various primary sources for easier location, e.g. the CRC handbook series.

Some indexes use keywords from the original; others use standard subject vocabularies.
In US libraries, terms are assigned from the Library of Congress Subject Headings (LCSH).
MEDLINE uses the Medical Subject Headings (MESH).
Chemical Abstracts uses its Index Guide.

All information organization and retrieval involves tradeoffs
- Specificity vs. collation
- Relevance vs. recall
- Maximum precision vs. maximum quantity
Specific headings avoid the irrelevant
General headings bring like items together.
Searching narrowly avoids having to look at irrelevant items at the cost of missing some relevant material.
Searching broadly helps insure that nothing is missed, but may require later screening to eliminate irrelevancies.
The information professional has to decide how best to meet the needs of the intended audience with the available technology.
The information user has to set priorities based on the ultimate objective and the time, labor and money available for searching.
Together, they evolve the strategy needed for extracting needed information from the universe of scientific publication.