EPA-RS72-002
August 1972
Socioeconomic Environmental Series
Information Retrieval Capabilities-
National Environmental Research  Center,
Cincinnati, Library
                                  ui
                                  CD
                               
-------
                                      EPA-RI-72-002
                                        August 1972
    Information Retrieval Capabilities-
National Environmental Research Center,
             Cincinnati, Library
                    M. Friedman
         National Environmental Research Center
                Cincinnati, Ohio 45268
               Project Element  1D2065

           NATIONAL ENVIRONMENTAL RESEARCH CENTER
             OFFICE OF RESEARCH AND MONITORING
            U.S. ENVIRONMENTAL PROTECTION AGENCY
                 CINCINNATI, OHIO  45268

-------
                         PREFACE








The library at the National Environmental Research Center,



Cincinnati, is the scientific and technical information



focal point for the entire Environmental Protection Agency



Library system.  As such, it provides technical information



back-up to all other EPA libraries, in addition to its



responsibilities to NERC-Cincinnati.








The Cincinnati Library has access to a number of information



centers which maintain over 25 scientific and technical



computerized data bases, and contains over 2.5 million doc-



uments.  We are now providing literature searches with our



Datapoint 3300 CRT computer terminals, using a variety of



on-line and off-line data base services.  These services



allow the library to provide retrospective searches on de-



mand as well as current awareness programs tailored to the



individual's requirements.







This manual describes these information services in more



detail.  Most of the services discussed herein are now



available for usage.  Some others will be installed and



operational in a few months.








Future developments in our information retrieval activities



will be designed to enhance the depth and breadth of coverage



and the variety of services provided to EPA staff members



requiring access to scientific and technical publications.

-------
                   ACKNOWLEDGEMENTS
We acknowledge with appreciation the permission of the



University of Georgia to use the sections on weighting and



logic, some of which is based on the material found in



"Profile Coding and Management Manual for the University of



Georgia Text Search System"







Preparation of the document has been aided by the significant



assistance provided by the entire staff of the NERC-Cincinnati



Library, in particular, Maxine M. Smith, Dottie F. Brofft, and



Mary L. Calkins.

-------
                         CONTENTS
SECTION
TITLE
PAGE
            PREFACE
I.
INTRODUCTION

The Search System
Data Bases
Types of Searches
                                                       1
                                                       1
                                                       4
II.
SEARCH STRATEGY                           12

STEP 1.  User Interview                   12

STEP 2.  Concept Identification           15

STEP 3.  Search Strategy Formulation      16
         A.  Subject Coverage             16
         B.  Data Elements                17

STEP 4.  Concept Expansion                17

STEP 5.  Profile Refinement               23
         a.  Truncation                   23
         b.  Weighting                    26

STEP 6.  Profile Coding                   30
         Forms                            32
         Group Number (GRP)               35
         Term Weight                      36
         DEM (Data Element Code)          36
         Search Terms                     37
III.
SUMMARY
42
Appendix A

Appendix B

Appendix C

Appendix D
              APPENDICES

         Data Base Description

         Price Information

         Sample Printouts

         Library Organization Chart
A-l

B-l

C-l

D-l

-------
                   LIST OF FIGURES
FIGURE     NAME                                   PAGE






  1        NERC-CINCINNATI INFORMATION NETWORK      2



  2        Boolean Logic                           14



  3        CAS Search Guide                        20



  4        Concept Expansion                       21



  5        Term Truncation Modes                   24



  6        Computer Search Information Request     32



  7        Profile Coding Logic                    33



  8        Typical Boolean Expressions             39



  9        User Evaluation Form                    41

-------
            U. S. ENVIRONMENTAL PROTECTION AGENCY


                  NERC - CINCINNATI LIBRARY


                MANUAL ON COMPUTER SEARCHES




I.   INTRODUCTION




The traditional library resources of EPA have been augmented


by access to various data bases for computer searching of


the literature.  With the ever-increasing amounts of published


literature, it is necessary to have a more efficient means


of access and comprehensive coverage of these sources of infor-

              i
mation.  Computers capable of rapid screening of material


answer this need and provide a necessary supplement to library


services.  The concept of an EPA Information Network was devel-


oped and expanded as exemplified in Figure 1.




This manual is designed mainly for use by librarians in under-


standing computerized searching, and discusses only currently


available or soon-to-be-implemented data bases.  Procedures


described herein will be applicable to all present and future


accessible data bases.




The data bases which the EPA library can or soon will be able


to access are located at the University of Cincinnati, Bio-


Sciences Information Service (Biological Abstracts), Battelle


Memorial Institute  (AMIC), Lehigh University  (Leadermart),


National Library of Medicine (MEDLINE), Informatics Inc.


(Toxicon and Environ), Ohio State University and the University


of Georgia.  These information centers and their systems are


described in more detail later in the manual.


                           1

-------
INFORMATION  REQUEST CENTERS
INFORMATION ACCESS CENTERS
                                      Battelle
             EPA Regional Offices
             EPA NERC's
        NERC-Cincinnati Librar
              EPA Labs
          EPA Headquarters
                 EPA Contractors
                                         University of Cincinnati
       Ohio State
         Informatics
         •«


       NLM
       s

   University of Georgia
           NERC-CINCINNATI INFORMATION NETWORK

                             FIGURE  1

-------
The University of Georgia has the most data bases, and is usually



accessed for retrospective searches.  Lehigh University, Battelle



Memorial Institute, and National Library of Medicine are searched



on-line from NERC-Cincinnati to their respective computers.  The



computer terminals for these are located in the EPA Library,



Columbia Parkway, Cincinnati.  Access to the terminal is available



by telephoning the EPA Library, using Commercial or FTS lines



(513/871-1820, ext. 492, 493, or 494), by Xerox telecopier  (auto-



matic) (513/871-1820, ext. 249), by TWX (810/461-2796), or by



mail addressed to the library.  Forms have been provided in order



for the librarian to clearly state the search strategy and selected



terms.  These forms should always be completed according to the



instructions provided and sent to the library.  Telephone conver-



sation should be limited to supplementing the information on these



forms or for unusual situations.





The NERC Library will supply hard copy or microfiche for retrieved



references if the requesting library cannot obtain them locally.





Appendix A. contains a brief overview of the major data bases



and a listing of the subject areas each data base covers.  This



appendix may be duplicated and distributed separately to libraries



for distribution to their subscribers.





This is not to be considered as the final selection.  As experi-



ence and changing requirements dictate, new data bases will be



added.  Additional systems will continue to be investigated for



possible inclusion.  This manual will be revised to show any



changes which may occur during this initial period of development.

-------
Appendix B. contains information on costs for all data bases
to be searched.  These costs are broken down to include charges
for each information center and each data base located at the
University of Georgia.
                         3A

-------
The NERC-Cincinnati library will notify all interested EPA



personnel of additional data bases, search services, price



changes, and new library services as they are developed.








The types of searches provided are:








     A.  ON-LINE SEARCHING



         On-line searching queries data bases instantly



         and provides immediate response between the user



         and the data base.  Data base coverage depends



         on the amount of information in the computer



         and usually is limited to the last few years.



         Besides quick response, on-line searching is



         recommended as a way of testing for the presence



         of relevant material, or for refining a search



         strategy before going further or deeper into a



         data base.







         MEDLINE is one example of our on-line retrieval system.



         Each article which enters the system is indexed under



         a number of subject terms.  In retrieval, Boolean logic



         is used, and terms are linked using "and", "or" or



         "and not".  Exact terms from MESH  (Medical Subject



         Headings) must be used.  Besides using subject terms,



         material may be searched by authors, languages, journal



         titles and year.

-------
LEADER is the name of the on-line retrieval system
at Lehigh University.  Each document which enters
the Lehigh University on-line system, with its
affiliated noun phrases, is processed by a series
of special programs which:

     1.  Isolate noun phrases for each document,
     2.  Isolate the component word stems of each
         phrase,
     3.  Create inverted files to link:  word-stems
         to noun phrases, phrases to documents,
         phrases to associated phrases.

The on-line system has various data bases, based on
the different document sources.  Currently available
are document corpora from:

     1.  Chemical Abstracts - This data base is de-
         rived from Chemical Abstracts.  Currently,
         coverage includes from v. 72, 1970 thru
         current issues.  Total documents of CA in
         system approximately 550,000  (Covers 12,000
         journals).

     2.  Engineering Index  (COMPENDEX) - Corresponds
         to the published Engineering  Index volumes,
         and contains mostly applied engineering
         references.  Currently, coverage includes
         from 1970 to date.  Total number of documents

-------
         approximately 120,000 with abstracts each
         10 lines (Covers 2,250 journals).

     3.   Civil Engineering - Recently initiated
         containing select articles pertinent to
         civil engineering.

     4.   MARC - A new on-line Library of Congress
         card catalog system is available for de-
         veloping computer-produced catalog records.
         An English language monograph can be cat-
         aloged by using the L.C. card number,
         author, or title.

This is  primarily a form of current awareness, on-line
system,  which provides quick retrieval and can be the
basis for developing a more comprehensive batch retro-
spective search.

Other data bases will be added as soon as feasible.
AMIC  (Analytical Methodology Information Center) is a
sample of a new data base.  Since it covers water
quality analysis it should prove to be quite valuable
to many EPA laboratories.

-------
B.  CURRENT AWARENESS
    Current awareness covers either the latest issue
    and/or each successive current issue from a data
    base and can be used to give a continuous flow of
    information from the computerized data base.

    The current awareness service available through
    the library consists of a profile developed for a
    researcher that will be matched against the latest
    editions of data bases.

    The profile is based on the researchers informa-
    tional needs and coded to reflect retrieval terms.
    On a bi-weekly schedule this profile is matched
    against multi-discipline data bases containing
    journal articles and government reports.  The output
    is printed on cards and sent to the requester.
    The cards consist of two parts - the smaller portion
    can be sent to the library as an order form and the
    larger portion (now a 3 x 5 card) can be filed by
    the researcher.  A profile can be adjusted at any
    time to reflect changing requirements or professional
    interests.

    A profile can be developed to combine terms, to
    weight them, or to eliminate irrelevant output.
    The following sample profile shows how terms can be
    selected, logic devised and terms weighted.  Sample

-------
    cards have been selected to show the kinds of



    retrieval that can be expected from a current



    awareness service.








Co  RETROSPECTIVE SEARCHING



    Retrospective searching can go back as far as the



    user desires or to whenever the data base was



    established.  Turnaround time for retrospective



    searches  (known as a "batch mode") may be two to



    three weeks due to the size of the files.  However,



    search results may be forwarded throughout this



    period as they are received rather than waiting



    until completion of the search.







    Search results are usually mailed within two days



    after receipt of the tape.  Retrospective searches



    are scheduled over a two-week processing cycle.



    Turnaround time for retrospective batch searches



    from receipt of initial profile  (search question) to



    delivery of final search results may run two to



    three weeks.  Search results are mailed throughout



    this period, however.








    Special demand searches on frequency less than this



    two-week cycle may be requested.  These are accommo-



    dated only at higher special rates for which the



    requester assumes the entire cost of computer runs.

-------
Each question should be a fully descriptive paragraph
describing the subject area to be searched and any
qualifications, such as particular applications
which are to be excluded.  Suggested index terms or
words frequently used in the research area under
consideration are especially helpful.  This step will
be explained later in this manual.  The search question,
or profile, may be modified by the user, in order to
obtain better search results.  Maximum number of
terms which can be processed in one search question,
or profile, is approximately 200, depending on the
length of the terms.

The search result is a bibliography for retrieved
documents.  The output includes the title, authors'
names, location of work (if available on the data base),
primary document citation, and all index codes or
index terms available on the data base.  Output on
8-1/2 x 11" one-part paper is standard output.  Output
on 4 x 6" card stock is available if desired, at
additional cost.

Any combination of these services can be requested as
well as any combination of data bases and years
covered.  These searches can be tailored to suit the
users' needs and are extremely flexible.  With exper-
ience these searches can produce a high rate of user
satisfaction and relevancy in retrieved information.

-------
In order to obtain maximum results from a computer search,
it is necessary to state the question clearly and to provide
additional information using synonyms and related terms.
Guidelines on developing profiles to serve as keys to unlock-
ing relevant information in the data bases will be discussed.
EPA forms have been provided for reporting this data and
this manual provides detailed instructions for completing
these forms.  These EPA forms need not be thoroughly com-
pleted but should include whatever information can be provided.

The library staff at NERC-Cincinnati will develop the profile
or will call the user for more data, if necessary, to complete
the profile.  It is important that care be taken in preparing
these profiles as they represent the most significant ingred-
ient in a successful computer search.  Each librarian should
maintain a close liason with the user and the Cincinnati
personnel in order to increase his/her knowledge on how best
to use computer searches.  The critical role played by the
librarian will be the ultimate test as to the potential worth
of the data base retrieval systems.

Another consideration is that more than one source has the
same data base.  Each librarian should decide on which source
and information center can be of advantage — based on years
covered, user's time constraints and costs — before report-
ing any preferences.  The NERC-Cincinnati staff will make
every effort to keep costs and turnaround time to a minimum.
                         10

-------
If new services become available which will offer an



advantage  (e.g. better price structure, faster turnaround,



more extensive coverage) then all participating locations



and organizations will be so informed.







Appendix C contains sample printouts that a user might



receive as a result of a computer search from the inform-



ation centers being used.







Appendix D contains a chart of the interim organization of



the NERC-Cincinnati Library and the names of staff members



and their responsibilities.







Study of this manual should lead to an awareness of the user's



and the librarian's role in properly accessing the vast amount



of information now available.  If there are any questions,



please do not hesitate in writing or calling NERC-Cincinnati



for assistance.
                         11

-------
II.  SEARCH STRATEGY








The following steps are designed to provide a logical



sequence of events for successful retrieval of information.



The summary provides a one-page outline of these procedures



for quick reference.








STEP 1:  USER INTERVIEW



To obtain pertinent references from the computer search, a



clear concise statement of the question is essential.  As



the person who will benefit most from the search, the user,



who will have the best understanding of his requirements,



should be consulted.  A discussion with him will help



formulate a clear statement of need.  He should also be



questioned for other terms related to the search and synonyms



used in the literature with which he is familiar.  The result



should be a well-conceived, written statement or paragraph



complete with synonyms arid related terms.  If the user does



not prepare this, then the librarian should write it with



the guidance of the user.







One specific point to be kept in mind while recording the



question is the use of the conjunction "and".  It is



frequently interpreted as meaning the same as logical "or"



as well as the connective "and".  Consider, for example,



the question, "I am interested in the effect of A on B and



C."  There are two potential questions represented:
                          12

-------
1)  the effect of A on B OR the effect of A on C, and 2) the
effect of A on B in the presence of C.  These questions
imply different logic strategy, so care must be exercised
in clearly stating the proper question.  As a general rule,
restrict the use of "and" in recording the question.  It
is better to develop short succinct statements than complex
sentences.  If possible, several abstracts or references
relevant to the search should be included.

Figure 2 shows some examples of how a search request can be
interpreted as a Boolean logic statement.  Although this
table shows only two (2) sets  (A and B), more complex logic
may be utilized.  For example:  A | B -» E signifies the request
for set A or B through E.  A & B & -» F signifies A and B but not
F.  Try to keep the logic as simple as possible and to use more
search statements.  This approach will be far more efficient
for computer operations.

At this time, a brief description of the various data bases
should be given to the user, so he will have an idea which
data base or bases will be most pertinent to his question.  The
scientists will be familiar with most of these in their field
such as Biological Abstracts, Chemical Abstracts, Engineering
Index, etc., and can decide which would have the best chance
of giving results.
                         13

-------
                        BOOLEAN LOGIC
  Logic Statement
 Boolean Statement
Venn Diagram
Search for a single

setj(sometimes[ called1
"descriptors" )(e.g., A
Search for the
presence of A and B
     A  &  B
     ^\^  "i
   AlB
Search for anything
excluding set A
Search for anything
containing set A
or set B .
        A/B
Searchjfor anything
containing set A orjset
B_but exclude) where
both  are present,
A  &7B//B &7A
 NOTE:  The above illustrates the most cotmonly used logic in a search statement.
       For best results do not develop more complex statements. In this illustra-
       tion the rectangle represents the universal set and the circle represents
       the logic set in question.  The shaded areas represent the specified set(s)
       for retrieval.  "AND" is shown as &; J'pR" is shown as |, and "NOT" is
       shown as ~i .
                                    FIGURE 2
                                     14

-------
As most data bases cover only recent years, the importance



of retrospective coverage should be decided.  Be sure all



searches are limited to the years covered by the data bases,



lest the librarian find she might be forced to commit her



own time for manually supplementing the computer output.



Cost may also be a factor as retrospective searches are more



expensive since charges are usually based on a per volume



or per year basis.  A current awareness search covering only



the most recent issue in a data base may be run to test the



profile before requesting a comprehensive search.








STEP 2:  CONCEPT IDENTIFICATION



Once stated, the question should be examined to identify



the main concepts.  In general, all nouns, including the



objects of prepositions, represent a concept.  For example,



the question, "What papers deal with the analysis of fatty



acids from tissues of mammals?"  contains four  (4) concepts.



A good test is to try to reconstruct the question by con-



verting adjectives to prepositional phrases.  If the result-



ing question is logically correct and noun forms of the



adjectives exist, the adjective represents another concept.



Adjectives modifying nouns should be treated as part of



the noun they modify.  For example, "All papers concerning



the use of acapella choirs in secondary schools" has true



modifying adjectives.
                         15

-------
STEP 3:  SEARCH STRATEGY FORMULATION
The term "search strategy" is used to mean the number of
search term groups and the Boolean logic relationships
between the search terms.  The strategy controls the way in
which search terms will be combined to determine whether
or not a given document is an answer to the profile.  Several
factors go into formulating the search strategy which will be
used in the profile.  The search strategy depends on the sub-
jective judgment of the profiler.  Experience and analysis of
the search results should lead to be best strategy.

     A.  Subject Coverage
     The nature and content of the data base itself will
     govern the strategy.  If one or more of the concepts
     covers all of the subject scope of the data base, then
     it can be eliminated from consideration.  For example,
     the concept "geology" in the question, "All papers
     dealing with the geology of the hilltop lake area of
     Ontario" is redundant with the subject scope of
     Bibliography of North American Geology: the data base
     contains only geology by definition.  A question which
     is to be coded for both CBAC and CA-Condensates may be
     coded with a different strategy for each since CBAC
     covers only biochemical papers while the CA-Condensates
     coverage is very broad.
                         16

-------
     B.  Data Elements




     The data elements available in the data base will also




     affect the precision recall balance; and therefore,




     the strategy to be used.  A data base such as Chemical




     Titles—which contains only titles—must have questions




     phrased much more broadly and with fewer concepts than




     does a data base which contains abstracts and/or index




     terms.  Profiles which are to be run against title-only




     data bases should seldom include more than two concept




     groups.  If the question contains more than two concepts,




     the librarian, in collaboration with the user, should




     select the two most important concepts.  With indexed




     data bases, especially when the vocabulary is thesaurus-




     controlled, the number of groups may be three or four,




     or higher, if necessary.  Precision of the search and



     recall of relevant documents are inversely proportional;



     if there are few groups, the recall is better, but the




     precision suffers.  The user should specify his require-



     ments of retrieval;it is then up to the librarian to




     construct the profile to meet the user's needs.
      *







STEP 4:  CONCEPT EXPANSION
                              *


Concept expansion is a translation of the question from the



language of the user to the language of the information file.



Each concept identified in the question must be expanded to



a list.of synonyms and related terms or replaced by thesaurus
                          17

-------
terms or codes.  Each concept will normally form one group in
the final logic expression with the terms in_ the group linked
by "or" logic.  Groups are usually linked by "and" or "not"
logic although there are no restrictions to these logic operators.

Depending on the data elements available in the data base
for search, two- conditions may exist.  The data element may
have a completely uncontrolled vocabulary.  That is, no
               i
restriction is placed on the words which may be used.  This
is commonly the case with titles and abstracts where the
choice of words is left up to the author or abstractor.
(Sometimes this approach is called "free-text searching.")
Vocabulary-controlled data elements, on the other hand,
imply use of some authority list or thesaurus for approved
words or terms.  Index terms and subject classification areas
or codes are frequently thesaurus-controlled.  Expansion of
the concept must be done differently for these two types of
data elements in order to effect retrieval.

For uncontrolled vocabulary data elements, such as titles
and abstracts,,it is necessary to construct a list of .
synonyms or related words for the concept term.  In essence,
the librarian with the user's help tries to predict all
possible ways the concept may have been expressed by authors.
In areas with which the librarian is unfamiliar, a basic
text can provide background and general knowledge of the concept.
                         18

-------
It may also suggest synonyms and related concepts which should



be included in the term list.  Search guides, such as the



CAS Search Guide (Figure 3) which contains narrower terms  (NT),



related terms (RT), synonyms (SY or S*), and broader terms



(BT), are valuable as memory joggers.
                           19

-------
January 1967
SEARCH GUIDE
    Amino Resins

A-Stage Resins
    BT Thermosetting Resins
    RT Phenol-Formaldehyde Resins
    SY Resols

Abandonment
    RT Depletion
    RT Life  (Non-Biological)

Abatement
    RT Damping
    RT Dilution
    RT Disposal
    RT Purification
    RT Reduction

Abdomen
    RT Gut

Abdominal Dropsy
    S* Ascite's

Aberration
    RT Distortion
    RT Images
    RT Interference
    RT Lens
    RT Reflection
    RT Refraction

Abies Sibirica
    BT Abis
    BT Evergreens
    BT Pinaceae
    SY Siberian Evergreen

Adietic Acid
    RT Rosins

Abiogenesis
    RT Biogenesis
    SY Spontaneous Generation
             i
Abiogenic Synthesis
    BT Synthesis  (Chemical)

Abiosis
    S* Death
                Abis
                  NT Abies Sibirica

                Ablating Polymers
                  BT Polymers
                  RT Ablation

                Ablation
                  RT Ablating Polymers
                  Burning
                  RT Combustion
                  RT Cooling
                  RT Melting
                  RT Vaporizing

                Abnormal Growth
                  RT Hypertrophy

                Abnormal Response
                  RT Sensitivity

                Abnormalities
                  RT Diseases

                Abnormalties
                  RT Irregularity
                  RT Malformation

                Abo Factor
                  RT Glood Groups

                Abo Incompatibility
                  RT Blood Platelet

                Abode
                  RT Habitat
                  RT Home
                  RT Residence

                Abrasion
                  RT Abrasive Blasting
                  RT Abrasives
                  RT Erosion
                  RT Friction
                  RT Gilling
                  RT Grinding
                  RT Metallography
                  RT Perforating
                  RT Polishing
                  RT Rubbing
                     Figure 3.  CAS Search Guide
                           20

-------
Figure 4 shows the concept "chroraatography" as it might be

expanded for a chemistry search in the uncontrolled vocabulary

data base CA-Condensates.   Abbreviations, acronyms, and alter-

nate spelling must all be taken into account if editorial

policies of the particular data base permit them.
                      Chromatography
                      Chromatographic
                      Chromatogram(s)
                      Chromatog
                      gas-liquid Chromatography
                      gas liq chromatog
                      gas Chromatography
                      vapour phase Chromatography
                      vapor phase Chromatography
                      thin layer Chromatography
                      paper Chromatography
                      V.P.C.
                      VPC
                      T.L.C.
                      TLC
                      G.L .C .
                      GLC
                      G.C.
                      GC

               Figure 4  Concept Expansion
The task of concept expansion is similar for thesaurus-

controlled data elements except that all terms must be taken

from the authority list.  Instead of predicting the author's

choice of words, the librarian must know the thesaurus and

have a working knowledge of the indexing policies used in the

assignment of index terms.  The hierarchy of index terms

must be taken into consideration in formulating the term list.

For example, all papers in ERIC reporting occupational educa-

tion of employees would not be retrieved by the broader term
                          21

-------
"personnel" since "employees" is also a term.  In other words,
there is no hierarchical posting to higher or lower level terms.

Data element types can be mixed during expansion of concepts.
A biological profile might include text terms (title, abstract,
and free index terms), biosystematic codes, cross codes, and
Registry Numbers, many as synonyms for each other.  This usually
occurs when a given profile is to be searched against two or
more data bases which have different data elements.  Librarians
should use some judgment as to whether it is better to write
separate profiles for each specific data base or to combine the
profiles using data base specific data elements as synonyms.

The search time is proportional to the number of search terms,
and the amount of core required for the profiles is determined
to a large extent by the number of different data elements.
Consequently, the economics of the computer search would dis-
suade the use of large numbers of non-productive terms and/or
data elements.  On the other hand, profile maintenance of a
question which is run in several data bases can be difficult
to control if many different profiles must be modified or
updated.  The Choice should be based on the degree of similarity
between the profiles which would be written independently.
Remember, too, that the data bases themselves may be dissimilar
enough to warrant different strategies for the same question.
                         22

-------
During the process of expanding the concepts, it is frequently
necessary to revise the search strategy formulated in the
preceding step.  Some concepts are almost impossible to expand
into any reasonable set of terms which will effect retrieval.
In such cases it is usually better to omit the concept from
the coding.  This may cause some irrelevant retrieval, but
there is less chance that relevant papers will be missed be-
cause of an inadequately expanded concept.  It may also be
better to revise the strategy to exclude irrelevant material
than to formulate the strategy to select only precisely
relevant documents.  This negative approach is particularly
ueful when the data base is vocabulary-controlled or has been
classified into subject areas.  Techniques for refining both
the profile terms and the search strategy are discussed in
the next step.

STEP 5:  PROFILE REFINEMENT
There are several techniques available for refining both the
search terms and the search strategy.  These include term
truncation and weighting.  It is also possible, where a data
base is available on line, to dp a short search to test the
validity of the selected terms before running a more expensive
and lengthy retrospective search.

     a.  Truncation
     Term truncation is a technique used to retrieve terms
     containing word fragments which are common to two or
                          23

-------
more different forms of the word.  For example, the

singular and plural or adjectival and adverbial forms

can be retrieved by a single term.  Thus, the term

*OXID* will retrieve DEOXIDATION, OXIDIZED, OXIDATION,

OXIDE, OXIDES, etc.  The asterisk (*) placed before

and/or after the word fragment is interpreted by the

search program to mean "accept any character, includ-

ing a blank, as valid in the position shown as asterisk."

Figure 5 illustrates the four possible modes of trunca-

tion and examples of each.


          i

    Term    Explanation

    CELL    Retrieves terms only if bounded by blank
            spaces or punctuation; e.g., CELL, CELL-FREE,
            STEM-CELL.

    CELL*   Retrieves the term followed by any character,
            including a blank space and preceded by
            blank spaces or punctuation; e.g., CELL,
            CELL-FREE, CELL., CELLP1/4JAME, CELLS,
            CELLULAR, etc.

    *CELL   Retrieves the term preceded by any character,
            including a blank space and followed by a
            blank space or punctuation; e.g., CELL,
            CELL-FREE, STEM-CELL, MICROCELL, etc.

    *CELL*  Retrieves the term preceded and/or followed
            by any character, including blank spaces and
            punctuation; e.g., CELL, CELL-FREE, ACELLULAR,
            MICELLES, CELLULOSE, etc.

             Figure 5  Term Truncation Modes
If carefully used, truncation can significantly decrease
the number of individual search terms which have to be
listed.
                      24

-------
It can also improve retrieval in that word forms which



might otherwise be overlooked will be automatically



retrieved.  It can also be dangerous if misused.  Word



roots carelessly truncated can cause large numbers of



irrelevant documents to be retrieved.  The choice of



appropriate truncated terms is aided by reference to



KWIC indexes in the printed publications or to word



frequency dictionaries available for some data bases.



A good rule-of-thumb is to avoid word roots of three



characters or less with both prefix and suffix trunca-



tion.  There are occasions when such small word roots



can be used effectively, but they are rare.








Although there are four possible modes of truncation,



not all modes can be used for all data elements.  Each



data element is classified as one of two types:  left-



anchored  (LA) or free text (FT).  The classification



determines the manner in which the data element is



searched.  For left-anchored elements, the match always



begins with the left-most character.  This implies that



the format and content of the data element are known.



Examples of this type include Registry Numbers, Cross



Codes, authors' names (inverted), and thesaurus-codes.



Only suffix  (right hand) truncation  (or no truncation)



can be used with left-anchored terms.  The only restric-



tion is that the truncation asterisk cannot immediately



follow any special character, including a blank.






                    25

-------
These are the first two modes shown in Figure 5.  The

second class of data elements character strings in the

data base.  Examples of this type data element are

titles, abstracts, and free index terms.  All four

modes of truncation are applicable to free text fields.



b.  Weighting

In a free text search system such as this, not all the

profile terms used in concept expansions are equally

significant with respect to the question.  Consequently,

a technique called "weighting" is available as an

optional feature to assign relative numeric values to

the terms.  There are two principal uses of weighting,

each of which will be discussed in greater detail.
           I
One use is to sort the answers into order of probable

importance.  The second use is an extension of the

search logic.



A suggested simplified approach to weighting can be

taken by ranking the selected terms.  Give each term

or rank a number, starting from 1 as the most important,

2 the next most, etc.  More than one term can have the

same rank of importance.  The Technical Reference

Librarian at Cincinnati receiving the request will

interpret the ranking for computer searching.  Initially,

a librarian may want to avoid weighting and should treat
                    26

-------
this aspect of a computer search only as an optional
feature.  With more experience the librarians can
prepare profiles involving the more complex approach
described below.

Weights for computer searching are positive or nega-
tive numerical values assigned to search terms.
Numbers must be in a positive or negative five figure
range.  In a weighted search, the terms within a group
may have different weights, and some terms may have no
weight assigned.  No specific order of weighted terms
is required.  A weight of zero (0) is assumed unless
otherwise specified.  A threshold weight can also'be
assigned to the question.  It is not necessary to code
weight with leading zeros.

Use of weighting for sorting  (Threshold weight equal
00000)
Weighting may be used to sort the search results by
assigning relative term weights to the index terms.  The
most important words (according to the user) are assigned
the highest numeric values, and less important terms are
given lower values.  The sum of the term weights for a
document may be added together to give a question weight
for that document.  Output search results are sorted for
                     27

-------
each user in descending question weight.  Thus, the
documents which hit on the most important term(s) will
have the highest weight and will sort first, followed
by the remaining citations in descending order.  For
searches in which weighting is used only for sorting,
the threshold weight for the profile is left blank  (i.e.,
has a value of zero).

If the threshold weight for the question is zero, then
weighting is calculated non-cumulative.  That is, the
weight assigned to each term is counted only once for
each citation, regardless of the number of times the
term occurs in the citation.  The results are sorted in
descending order on the question weights calculated as
the sum of weights  (non-cumulative) for all terms which
"hit".,  The non-cumulative weight is printed on the out-
put.  This type of weighting (no threshold weight) is
used exclusively for ordering output.

Use of weighting as logic  (Threshold weight not equal
00000)
Weighting may also be used as an extension of the "and",
"or", and "not" logic.  "Not" logic, for example, is
absolute.  If a document hits on a "NOT" term, the docu-
ment will not be retrieved even if it also contains
desired terms.  By use of term weights and a threshold
                     28

-------
weight for the question, undesirable terms alone will
not cause retrieval, but the document will be retrieved
if desired terms are also present.  In other words, a
threshold weight may be assigned to a profile such that
the sum of the hit term weights must be equal to or
greater than the threshold weight.

It should be emphasized that weighting of terms is an
optional and not a necessary technique for developing
search strategies.  Only those with some experience or
knowledge of a data base should consider using weighted
terms to any extent.  Furthermore, weighted terms should
only be used for current awareness or retrospective
search and not for on-line systems.

For example, if the threshold weight for the question is
not equal to zero  (either positive or negative), weighting
is calculated cumulatively for all occurrences of all
terms in the data base.  This allows a simulation of
"ignore" logic by judicious use of positively and nega-
tively-weighted terms when truncation would otherwise cause
problems.  The cumulative weight is used to test against
the threshold weight to determine whether or not the
threshold weighting is satisfied.  The search results are
sorted in order of the non-cumulative weights for all
terms which "hit".  The printed weight for the question,
                    29

-------
     however, is the cumulative weight.  Consequently, the



     search output may not always appear to be in strict



     order by question weight.








     Negative weighting can be used in place of "NOT"



     terms, especially where a few words consistently cause



     problems in conjunction with truncation.  Consider the



     term *ASE which is frequently used in biochemistry for



     retrieving enzyme names.  However, it also retrieves



     words such as "decrease," "increase," "phase," etc.



     If the term *ASE is assigned a positive value (e.g., 100)



     and the terms DECREASE, INCREASE, PHASE, etc. , are



     assigned an equal but negative value (e.g., -100) with



     the threshold weight of 100, then a paper reporting an



     increase of oxidase still will be retrieved.  If only



     an undesired term matches, then the question weight



     would be negative and less than the threshold weight



     necessary to cause retrieval.








STEP 6:  PROFILE CODING



Once the search strategy has been formulated, the concepts



expanded to search terms and refinements such as truncation



and weighting noted, the profile is coded for computer search



using the "Computer Search Information Request," EPA 2170-1,



abd tge "Profile Coding Logic" form EPA 2170-2, (Figures 6 and 7).
                         30

-------
If possible, the current awareness files should be searched

first to test the profile.  The profile may need to be

modified before any retrospective searches are made.



If the computer search originates from a laboratory outside

Cincinnati, these two forms "Computer Search Information

Request" form and "Profile Coding Logic" form, together

with sample abstracts, sample references, or any other

material which will help the search are to be sent to:
                Environmental Protection Agency
                NERC-Cincinnati
                Cincinnati, Ohio 45226
                Attention:  Technical Reference Librarian
                         31

-------
                                                 Page 	 of

                  COMPUTER SEARCH INFORMATION REQUEST


                                                 Date
MAIL TO:
Environmental Protection Agency
NERC-Cincinnati
Cincinnati, Ohio 45226
Att:  Technical Reference Librarian
TO BE COMPLETED BY ORIGINATING OFFICE:

LAB NAME
USER'S NAME                           PHONE NO.
USER'S NO. 	 USER'S 1st REQUEST  ( ) YES     ( ) NO

LIBRARIAN                             PHONE NO.
ACCOUNTING SYMBOL

MAILING ADDRESS: _
(to which
 material should  •
be sent)
DETAILED STATEMENT OF REQUIREMENTS; Please describe, as specifically
as possible, the subject matter for which the search is to be con-
ducted.  Define any terms that may have special meaning.  If there
are points NOT to be included, please state these.  Include language
statement  (English and/or non-English).  Use additional page if
necessary.
THRESHOLD WEIGHT                     BOOLEAN EXPRESSION
ON-LINE SEARCH: YES 	 NO 	 BATCH SEARCH:  YES	  NO

DATA BASES TO BE USED AND YEARS COVERED:
AUTHORIZING SIGNATURE                                  DATE
TO BE COMPLETED BY NERC-CINCINNATI LIBRARY:

USER'S PROFILE NO. 	 DATE REC'D. 	 DATE COMP.
LIBRARIAN	
                        (SIGNATURE)


EPA Form 2170-1            32

-------
                                                Page 	 of


                  PROFILE CODING LOGIC
This form is to be completed whenever the originator desires to
develop his/her own coding logic as a means of expediting the
search.  Please attach this to the Search Information Request.

User Name 	

User No.
Librarian's Name

Date
Please rank the terms in the order of importance.

 GRP        WEIGHT       DEM         TERM
Request Reviewed:	 Date
                         Signature
EPA Form 2170-2         Figure 7
                          33

-------
The person preparing the forms should keep one copy of the

request at least until such time as a printout is received.


The other two copies should be sent to NERC-Cincinnati, using

the address printed on the "Computer Search Information Request"

form, (EPA Form No. 2170-1).




In preparing the "Computer Search Information Request" form,

the first part should be completed at the originating office.

The name of the laboratory or organization, the name of the

user, his/her phone number, the mailing address for the print-

out are self explanatory.  If a librarian is completing the

form, then the librarian's name and phone number are to be

entered.  User number is for the originating organization's

internal use.  The appropriate answer for first time usage

should be checked.  Accounting symbols refer to the account

number against which the search will be charged.  Each re-

questing organization will be required to pay for all external

expenses incurred in the course of answering a search request.




The detailed statement of requirements should be completed

next.  If desired, (and only as an option) the user or orig-
              i
inating librarian may enter a threshold weight and a Boolean

logic expression.  If this option is not exercised, the

NERC-Cincinnati staff will complete the form or will call the

originator for; clarification.
                         34

-------
The search questions and data base questions must be answered
as fully as possible.  The authorizing signature is for use by
the user's division chief, or equivalent, and is necessary
only to discourage ephemeral usage of the retrieval system.
The form can be folded so that the address of NERC-Cincinnati
can be seen through a window envelope.  The bottom portion will
be completed by the NERC-Cincinnati staff and one copy will be
returned with the printout.  If a series of printouts are to
be sent back, the form will be attached only to the last one to
indicate the completion of the requested search or profile.

The "Profile Coding Logic" form (EPA Form No. 2170-2) is optional
and may be used with discretion.  However, if it is completed,
it will greatly expedite the request, as it will be entered into
the computer system with only a check for accuracy and complete-
ness.  In addition to saving time and producing faster turnround,
the librarian or user who completes this form will most likely
obtain greater accuracy and less irrelevant output.  Thus, the
completion of this form to the extent possible is urged upon each
and every user.

Each search term entry consists of up to four types of data.
The Group Number, labeled GRP on the form, is a four character
string, consisting of the letter G and three digits  (e.g. GOOD.
Leading zeros must be coded in all cases.  Group numbers are
used to link all terms for a given concept.  All terms within
                         35

-------
a given group are linked by "or" logic by definition.
However, a group can consist of a single term, as well as
many terms.  There is no limitation on the number of terms
which may be included in any group.  However, there is a
maximum of 240 terms or 1920 characters (whichever occurs
first) which may be used for the entire profile.  Group
numbers must be assigned sequentially, starting with G001.
No group number may be omitted.

The term weight can be up to 6 characters.  This concept was
previously discussed in this manual.  The weight number can
contain both a plus and minus sign or be preceded by a minus
sign if appropriate.  The plus sign is assumed and is never
coded.  Unless weighting is indicated as being desirable,
the absence of any reference to this feature will result in
all terms being given equal value.  Weighting is usually not
necessary for most searches.

DEM represents the Data Element Code.  This is a three char-
acter acronym for the data element type which is being searched.
The valid acronyms and their corresponding data element names
will be furnished upon request.  The value of using the DEM can
be exploited only after gaining experience with profiling.

In addition to the specific data element codes given in the
chart, there is one general code, TXT, which may be used to
indicate that the title, abstract, and free index terms are
                          36

-------
to be searched.  This will undoubtedly be the most commonly



used DEM.  Unless otherwise indicated, TXT will be assumed.







The fourth element is the search term itself.  Truncation



is indicated by asterisks preceding and/or following the



characters of the search term.  Note that an asterisk can



never appear within a search term, only as the initial or



final character.  Ther term may consist of alphabetic,



numeric, or punctuation characters.  However, all punctuation



characters will be converted to blanks in the search, both



in search terms and the data base.  Punctuation should be



included where appropriate since these characters (hence the



corresponding blank spaces) are counted in the overall term



length.   Terms which are identical except for punctuation



(e.g. NMR and N.M.R.) need be coded only once.  The maximum



length of a term, including truncation asterisks and imbedded



blanks,  is 59 characters.  Be sure to use only one term per



line.  It is better to use more paper for describing a search



than take the chance of confusing the person entering the



request into the computer.







The Boolean Expression is a statement of the search logic.



It consists of the group numbers, any combination of the



logic operator symbols "&", " ", and "S"1", and parentheses.



"Not" logic is indicated by the two symbols "&" without inter-



vening punctuation.  Another convention which can be used
                         37

-------
are the key words "and", "or", and "and not" enclosed within
periods (e.g., .and., .or., and .and. .not.).  Full Boolean
logic operations are permitted.  The maximum length of the
Boolean Expression is 180 characters.  The "or" operation is
the inclusive or; that is, A B is true if either A or B or
both are true.

Every profile must have a Boolean Expression, even if there
is only one group.  Parentheses can be used to improve
readability or to alter the order in which the logical oper-
ations are performed.  Without parentheses, the logic opera-
tions are performed in the order "not" equal to "and" greater
than "or", working from left to right for operations of equal
value.  For example, in expression (4) in Figure 8, the "and"
operation between G002 and G003 would be performed first,
followed by the "or" operation with G001.  Thus, the two ex-
pressions (3) and (4) are equivalent.  Expressions (7) and (8)
are different, however.  In (7) the "and" operation between
G001 and G002 is executed first followed by the "or" with G003.
Operations within parentheses are executed before all other
operations.   The parenthesis in (8) causes the (G002 j  G003)
part of the expression to be evaluated first, followed by the
"and" operation with G001.  If in doubt about the order in
which operations will be performed use parentheses.
                         38

-------
          (1)  G001



          (2)  G001 & G002



          (3)  G001  | (G002 & G003)



          (4)  G001  | G002 & G003



          (5)  G001 & G002 &  G003



          (6)  G001  | (G002 & G003) & G004



          (7)  G001 & G002   G003



          (8)  G001 & (G002   G003)



          (9)  G001  .and.  (G002 .or. G003) .and. .not. G004*



*Although available, this type of logic notation is not ordi-



narily used because of the time and space involved.



           Figure 8  Typical Boolean Expressions





The last line on the form labeled "Request reviewed" and date



is for use by the Cincinnati librarian who has reviewed the



request.  Feel free to photocopy or Xerox these forms until



the official version is available.  The library staff at



Cincinnati is available to answer any questions about enter-



ing a request or completing these forms.  Write or call if



help is desired.  In return, if it is found that incomplete



information or ambiguities will delay a request from being



answered, a librarian in Cincinnati will call for clarification



rather than simply returning the request.





Upon completion of the computer search, the requesting



librarian should complete the User Evaluation Form  (sometimes



called "feedback" as shown in Figure 9).  The purpose of this



is to allow NERC-Cincinnati to determine how to improve service



where necessary.
                         39

-------
NERC-Cincinnati will inform all participants in this network



of any changes that may occur.  If the users would like



NERC-Cincinnati to publish information such as successes, prob-



lems, suggestions, search results, etc., please let us know.



As of now the "Library Scene" (New Additions) will carry a list



of the successful searches and will supply copies upon request.







Free subscriptions and additional copies are available upon



request and are encouraged as a way of publicizing these



computer searches.
                         40

-------
                                              Page 	 of
                    USER EVALUATION
Lab Name
User Name
User Profile Number
Librarian's Name
Search Question
How do you rank the search?   (check one)
    (1)  Very Good 	
    (2)  Good      	
    (3)  Fair      	
    (4)  Poor      	
Suggestions and/or recommendations for improvement
                        Figure 9
                          41

-------
III.  SUMMARY



A typical search strategy should follow these ten (10)



procedures:



      1.  User Interview to obtain clear statement of search



          question with list of synonyms and related terms.



          Decide on data bases to be checked and years to be



          covered.  Include related abstracts or references



          if possible.



      2.  Concept identification.



      3.  Search strategy formulation.



      4.  Concept formulation.



      5.  Profile refinement.



          a.  Truncation



          b.  Weighting



      6.  Fill out profile coding form.  Keep copy.



      7.  Fill out search information form.  Keep copy.



      8.  Keep records by user name or own numbering system.



          Allow a maximum of two weeks for return printout



          when requesting large batch search services.



      9.  Check with uer when search is returned.



     10.  Send in user evaluation form.
                          42

-------
                    APPENDIX A

               DATA BASE DESCRIPTION




1.  Bibliography of North American Geology.

    Published since 1887, the B4.bi4.ogia.phy ofi  Noith

    Ge.otogy was first computer-produced in 1965  for  the  1961

    annual volume.  The kb&tia.c.t& o£ Noith Ame.i4.can  Geology
             i
    began publication in 1966.  Ge.ophy&4\c.a.£ Ab4£tac.£&, a

    related publication by the U.S. Geological Survey, was

    first published in 1929, and ceased publication  in

    December, 1971, the date through which the B4.bt4.ogia.phy

    ojj Month Amet'tca.n Gzotogy will also cease.   (For liter-

    ature coverage subsequent to 1971, use the GEO.REF data

    base)




    This bibliography lists publications concerning  the

    geology of the North American continent, Greenland,  the

    West Indies and adjacent islands, Hawaii,  Guam,  and  other

    island possessions, but not the trust territories of the

    United States.  Articles by American authors published

    in foreign journals are cited if they deal with  North

    American localities or are of a general nature,  but  not

    if they deal with foreign areas.  Articles on North

    America by foreign authors are included regardless of the

    place of publication, while those of a general nature are

    included only if they appeared in North American journals.
                        A-l

-------
2.   Biological Abstracts-Previews  (BA-P).
    The magnetic tape of B4.otog
-------
    assigness; bibliographic citations; and phrases from

    the CA issue Keyword Index.  There is no accompanying

    printed version of CA-Con.de.nAate.&.  The service is

    designed to provide the user with the capability to

    scan rapidly by title and key-word the pertinent con-

    tent of some 250,000 new articles and patents each year.




    Publication of the tapes began with Vol. 69 (July 1968).

    Two volumes of 26 issues each are published per year.




4.  Cataloging and Indexing (CAIN).

    The initial CAIN tapes  (3 reels) designated "CAIN 1969",

    contain 42,667 records.  The January, 1970, tape records

    are included in the CAIN 1969 volume.  Approximately

    5,500 tape records are produced monthly, depending on the

    number of publications received and processed by the

    National Agricultural Library.




    Approximately 11,000 publications are covered by CAIN.

    Tape records are produced monthly and are divided among

    18 broad subject areas providing bibliographic data con-

    cerning the literature of agriculture and related sciences.

    Journal articles, pamphlets, government documents, includ-

    ing USDA, state experiment stations, and extension service
              i
    publications, special reports, etc., received by the NAL

    are sources for entry of data on the CAIN tapes.


    CAIN contains about 100,000 records to date.


                         A-3

-------
5.  Chemical-Biological Activities  (CBAC).
    C/ie.mxtca£-8/to£og/cca£ Ac^u/c^e^  (CBAC) is a biweekly tape
    service supplied by Chemical Abstracts Service containing
    digests of current technical articles related to the field
    of biochemistry.  CBAC covers the scientific literature
    that reports the interactions of organic compounds with
    biological systems as well as related metabolism and
    in-vitro chemical studies.

    About 600 journals are referenced in CBAC per year.  Two
    volumes of 13 issues each plus Cumulative Indexes for
    each volume are published each year, dating from January,
    1965.  A total of 110,000 digests were publlished in the
    first 12 volumes (Jan. 1965 - Jan. 1972); an average of
    9100 digests/volume.

    The separate printed copy of CBAC ceased in December 1971.
    It is contained in the first five section groupings of
    Chemical Abstracts, the Biochemistry Sections.

6.  Chemical Titles (CT) .
    Che.mj.c.at Tj.tf.e.& (CT) , available from January 1962, pub-
    lished biweekly, is a current author and keyword index
    of recently published articles from approximately 700 of
    the world's most active chemical and chemical engineering
    journals.
                        A-4

-------
    Articles are input using only words in title, author(s)

    name(s) and journal reference.  There are 26 issues per
             !
    year with approximately 5,000 titles cited per issue.
             i


7.  Computerized Engineering Index (COMPENDEX).
             !
    C0MPEWPEX is the computerized version of the Eng-cneei/cng
             i
    Index.  The COMPEWPEX data base, containing the input to

    the E/t Monthly for each issue from January 1969 to the

    present, provides rapid access to over 7,000 abstracts

    monthly, reflecting the contents of 3,500 sources of

    worldwide! technical information and the work of 88,000

    authors, i It covers all fields of engineering and certain

    related fields of science and management, pertinent
             i
    quality research and applications literature; assembles

    the various engineering disciplines side by side with

    interconnecting cross references.
             i

             i
    Users have the option of searching the complete record

    (full text search) or searching selectively on chosen data

    elements:  Subject heading and subheading, document ID

    number, author(s) Ei abstract number, codes assigned to
             i
    the CARD-A-LERT service divisions, on most records the

    access words  (which are cross references chosen for the

    printed version of COMPENPEX), and on most records the
             i
    free language terms.

             i


    Abstracts are included.
                         A-5

-------
8.  Current Information Tapes for Engineers (CITE).



    The CITE tapes for Plastics and Electrical/Electronics



    were first issued in 1965.  The service originally called



    the U.P.P. (User Participation Program) became,  in 1968,



    the CITE service, or Current Information Tapes  for Engin-



    eering.  The Plastics Monthly and CITE service  were



    continued until December 1969.  The E/E Section was dis-



    continued in December 1967.






    The CITE tapes containing input in Plastics Engineering



    of some 30,000 articles covering the years 1965 through



    1969 and some 60,000 in Electrical and Electronics



    Engineering for the same time period, are for retro-



    spective searches only.





9.  Educational Resources Information Center (ERIC).



    The ERICTAPES comprise the magnetic tape copies of the



    Educational Resources Information Center (ERIC)  files,



    consisting of 30,000 report resumes from 1966 on,  12,000



    journal article citations from 1969 on, and a thesaurus.





    The Report Resume Files consist principally of  resumes of



    research reports filed by contractors and grantees on the



    results of funded educational research.  The Journal Article



    Resume Files contain resumes of journal articles on educa-



    tional research selected from over 500 education and educa-



    tion-related journals.  The Thesaurus File is a complete



    Thesaurus of ERIC Descriptors from which subject indexing



    terms are selected for both report and journal  article resumes.





                         A-6

-------
     Under special authorization from the U.S. Office of
     Education, these tapes are available only by purchase
     from LEASCO Systems & Research Corporation.

10.  Environmental Protection Agency  (EPA) funded data bases;
     a)  Analytical Methodology In^oimoit4.on Co.nt.ifi   (AMIC) .
         The AMIC computerized data base was established in
         1971 by Battelle, Columbus Laboratories, Columbus,
         Ohio, under contract for the EPA, to inform the
         personnel of the Analytical Quality Control Laboratory
         (AQCL) and other segments of the National Analytical
         Methods Development Research Program of recent publi-
         cations on methodology.

         Each month AMIC adds to its data base approximately
         200 indexed and abstracted current reports on methods
         for determining the indentity, concentration, and
         ecological effects of pollutants, and measuring water
         quality.  Subject areas covered by AMIC include biology,
         chemistry, instrumentation, microbiology, and quality
         control .
         The current awareness bulletin, Review* oj
         L4.teJia.tu.tie. on knatytj.c.a.1 Methodology and Qu.alj.ty Control,
         containing the abstracts is published monthly.  There
         are now over 3200 documents in the data base.
                         A-7

-------
10. b)    Environmental Information Retrieval On-line (ENVIRON).

         ENVIRON is a new EPA system which replaces the

         Technical Information Management Planning System (TIMPS)

         established in 1970 by the Research Information Divi-

         sion, Office of Research & Monitoring, E.P.A.,

         Washington, D.C.



         This system utilizes a software package STIMS/RECON

         which was developed by Informatics, Inc., and will

         be run at the computer facility of COMNET (Computer

         Network), Washington, D.C.  EPA users throughout the

         country will have telephonic access to the data bases

         at the COMNET facility in an on-line real time

         environment.



         The data base contains pertinent information on all of

         the research and monitoring grants, proposals, projects

         and final reports generated by or for EPA.  This

         includes:

              Proposed R & M Projects
              Active R & D Projects
              EPA R & D Project Report System
              Solid Wastes Projects
              Water Quality Surveillance Network
              Marine Biological Literature
              Industrial Waste Literature
              Oil and Hazardous Materials   (Oil spill data:
                                            TADS  (Technical
                                            assistance data
                                            system)
                         A-8

-------
c)   EPA library files.



     The library files will consist of references to all



     EPA reports and publications, including predecessor



     agencies and the journal holdings of the EPA libraries.



     The book catalogs of EPA libraries/ as soon as they



     are available, will also be added to these files.







d)   Solid Waste Information Retrieval System (SWIRS) .



     The SWIRS data base covers the period from 1964, in-



     cluding both periodical and non -periodical literature



     of the world on solid waste management; for the period



     1941 through 1963, see Public Health Service Publica-



     tion No. 91 and Supplements A through F, which contain



     all-inclusive solid waste management bibliographies.
     The current awareness bulletin, Acce44>ton



     Soii.d Watte. In^otma-tton Re>fUeva£ Sy&te.m,  is published



     monthly to keep readers abreast of the solid waste



     management information being published.  There are



     over 13,000 abstracted items in this data base, con-



     taining a description of the contents of the documents



     being added to SWIRS and giving a complete biblio-



     graphic .citation.
                      A-9

-------
11.  Geological Reference File  (GEO.REF).
     GEO.REF is the data base originated by the Geological
     Society of America, in  cooperation with the American
     Geological Institute.   The  GEO.REF tapes are available
     beginning with 1967.  The bibliographic information
     dates from 1965, however.   The  printed publication
     corresponding to this data  base is B/cb-LtogAapfoj/ and
     Index OjJ Ge.ology.

     This data base covers world-wide literature on geology
     and related technologies selected from over 2,000
     serials and special publications.  Entries are classi-
     fied under 21 categories, which include such disciplines
     as Geochemistry, Geochronology, Paleontology, Stratig-
     raphy and Structural Geology.

12.  Geophysical Abstracts  (GPA).
     GPA data base covers the years  1966 through 1971, when
     Ge.oph.y&
-------
     For literature coverage subsequent to December 1971,  see

     GEOoREF data base.



13=  Institute for Scientific Information  (ISI) Science Citation
              i
     Index.

     The ISI Sc.-ce.ncfc Ci.ta.tion Index tapes  (source and citation)

     cover the current literature of science and technology

     (including the social sciences).  They contain biblio-

     graphic information about the source articles, and also

     about the references cited in those articles, which are

     covered in the five sections of Cu.JDie.nt Conte.nt&, approxi-

     mately 8,000 current items each week.



     Dating from 1964, coverage is comprehensive and includes

     all items (other than advertisements and ephemeral

     notices) in approximately 2,500 key scientific journals.



     File size (1971): 2,538,000 records  (source items)

     with approximately 410,000 items added per year.

     File size (1971); 32,000 items (cited references) with

     approximately 5,000,000 added each year.



14.  Machine Readable Catalog (MARC II).

     MARC is the production of LibJiaJiy o£ Congie-64 cataloging

     data in computer readable form.  The format was imple-

     mented in June 1968.  At present, MARC II cataloging  data

     is limited to English language monographs which also


                         A-ll

-------
includes government reports/documents, and conference



proceedings.  Plans for expansion to cover monographs



in French and German are being studied.  MARC is



available on-line and will be used as the basis for



NERC-Cincinnati library providing scientific and



technical cataloging service to any EPA library that



so requests.
                     A-12

-------
15.   MEDLINE
     MEP/.INE (acronym for MEPLARS On-Line) is a new service



     from the National Library of Medicine.  It provides



     on-line bibliographic searching of biomedical journal



     literature.  The data base includes more than 400,000



     citations with index terms from 1,138 of the most



     important periodicals indexed for Index Med hospitals and research institutions.





                          A-13

-------
16.  National Technical  Information Service (NTIS).
     The U.S. Gove.inme.nt Repot-ti  Announcement (USGRA) and
     U.S. Goveinmen-t Repo.t;t4  Index  (USGRI) , formerly called
     the U.S. Govei.nmen;t Reaea/tcrt and Ve.ve.topme.nt Repo-t/ta
     (USGRDR), tapes are issued twice monthly, 24 issues
     per year; approximately  50,000 document records are
     added per year.

     The National Technical Information Service {NTIS),
     formerly the Clearinghouse for Federal Scientific
     and Technical Information  (CFSTI), began making
     available magnetic  tapes of  its U.S.  Government
     Research and Development Reports (USGRDR) beginning
     with the January 10, 1970 issue.  With Volume 71,
     Issue No. 6, March  25, 1971, the name of the NTIS's
     announcement journal changed from U.S. Gove.Jin.me.nt
     Re.6eaic.rt and Ve.ve.lopme.nt Re.poit& to Gove.tinme.nt
     Repot^tA Announcement  (GRA) , and includes technical
     publications of the Department of Commerce which are
     not research and development reports,  as well as the
     research and development reports previously being
     covered.  A similar change was made in name to the
     U.S. Gove.finme.nt Re.6ea-t.ch and Ve.ve.topme.nt Repot;t6 Index
     (USGRDR-I) , which became Gove.fi.nme.nt Re.polt& Index (GRI)

     The USGRA tapes cover new U.S. Government-generated
     research and development reports and  translations of
                          A-14

-------
17.  Nuclear Science Abstracts  (NSA) .



     Nu.c.le.0.*. Science Ab&tia.c.tA  (NSA) is a semimonthly publi-



     cation of the USAEC Office of Information Services,



     published by the USAEC, Technical Information Center.



     There are 24 issues/year or volume.







     A total of 450,000 abstracts were published  in the  first



     21 volumes (1948-1967).  The annual volumes  now contain



     approximately 47,000 abstracts each.







     NSA, as a hard-copy publication, began in 1948.  The  first



     computer files available began with Volume 20, Issue  No.  12,



     (mid-1966) .  NSA provides the only comprehensive abstract-



     ing and indexing coverage of international nuclear  science



     literature.  It covers scientific and technical reports of



     the U.S. Atomic Energy Commission and its contractors,



     other U.S.1 Government agencies, other governments,  univer-



     sities, and industrial research organizations.  In  addition,



     books, conference proceedings, individual conference  papers,



     patents, and journal literature on a worldwide basis  are



     abstracted and indexed.







     Printed Publications and Aids for using  Nuc.te.cUi Science



               are:
     TID-4552  (Rev. 5), Su.bje.ct Scope. o&  Muclecu.  Science kb&t>ia.c.tt>



     TID-4577  (Rev. 3), Peac/Up^tcue Ca&tlog-cng  Guide.
                          A-15

-------
     TID-5001 (8th Rev.), Subject Heading* lUed  by  tku> and Ca.te.goiy  L^
      Nuclear
     IAEA-INIS-13 (Rev.)  IMIS:
                            the.
     the.  Indiana. U. Tape^ at  IRL  - Be^.fee£ej/, May  1970.
     TID-4578,  Sub^fcC^: Modsifae.*. Gu/ccte
     A/SA  Entiy  and Se.le.c.toi file.  fox.ma.tA
     TID-85,  Report Nu.mbe.ft. Se.Ji-ie.4 U^&d by the. VL kutkoiity L
-------
18.  PANDEX - CURRENT INDEX TO SCIENTIFIC & TECHNICAL LITERATURE.
     PAWPEX - Cu.tui<>.nt Indzx. to ScUen£c<^c 6
     provides a comprehensive data base from 2,100 scientific,
     technical and medical journals;  6,000 scientific technical
     books, 5,000 selected patents, and 50,000 U.S. Government
     technical reports annually.  It covers all areas of pure
     and applied science, with coverage beginning in 1967.  Each
     entry record contains manually indexed or manually edited
     thesaurus-generated subjects.  PANPEX serves research re-
     quirements not satisfied by single disciplinary indexes but
     provides access to all literature of interest to a disci-
     pline as well.

     The index' generated by PAWPEX from this data base is
     divided into two parts:  (1) Subject-entries containing
     full title, primary author and periodical reference,
     arranged by all significant subject words and subarranged
     by all significant secondary words; (2) Author-entries
     containing all authors arranged alphabetically  (primary
     author entries containing full title, all other authors,
     and the periodical reference)  (secondary author entry
     containing name of primary author and periodical reference) ,

     The magnetic tape files are prepared by CCM Information
     Corporation and issued weekly.  The printed hard copy
     publication is issued bi-weekly.
                          A-17

-------
19.  Psychological Abstracts.



     P&yc.hotog
-------
     articles taken from the original papers.  SPIN  also
     contains the reel and frame number of the first page
     of each article published in full text in Cut/ien-t
     P/i£/4-tc4 M^.cA.OjJo^m (CPM) .  Information centers sub-
     scribing to both SPIN and CPM can provide a current
     awareness alerting service for material chosen  to
     meet individual interest profiles with the full text
     backup for AIP published journals from the CMP
     microfilm.

21.  Sociological Abstracts.
     Soc.4.otoQ4.c.OLt Ab&tJia.c.t&, published under the co-sponsor-
     ship of the Eastern Sociological Society and the Midwest
     Sociological Society, covers a broad range of sociological
     articles in periodicals in various languages.   Frequency
     varies, but since 1964, the abstracts have been published
     9 times a year.  The printed abstracts date from 1952.

     Currently, the data base tapes include Volumes  13, 14,
     and 15 (1963, 1964, and 1965).  Tapes are being prepared
     for Volumes 11 and 12 (1961 and 1962) and later, Volumes
     16, 17, and 18 (1966, 1967 and 1968) will be added.
                         A-19

-------
22.   TOXICON.
     The new toxicology information system, entitled TOXICOM,

     is an extensive collection of computerized toxicology

     information and data.  It originates from the Toxicology

     Information Program of the National Library of Medicine.



     TOXICON provides on-line literature searching and is

     designed to serve health professionals working in the

     areas of environmental pollution, industrial or occupa-

     tional health and safety, pharmacology, toxicology,

     medicine, agriculture and other bioscientific disciplines.



     The data base contains citations primarily, most includ-

     ing abstracts and/or indexing terms, full text state-of-

     the-art reports and toxicity data.



     Among the data bases included in TOXICOW are:

     -The. Haye.A f-lte. on
      Includes more than 10,000 citations of reports on
      health aspects of pesticides with a limited number
      of index terms and abstracts. (EPA, 1940 to 1966)
     -HAPAB (Hzatth A^pec^A orf Pe.&ticlde.A kb&ttiac.t Bu.tte.t-in)

      Includes more than 8,000 abstracts of reports of the
      health aspects of pesticides in humans and animals;
      poisoning treatment, pesticide residue analysis and
      monitoring.
      (EPA, 1966-current)
                         A-20

-------
      -KEEP (Ab&ti
-------
                              COMPUTER DATA BASES
     AGRICULTURE

     a.    CAIN

     BIO-MEDICAL

     a.    Biological Abstracts
     b.    BioResearch Index
     C.    MEDLINE/MEDLARS
     d.    *TOXICON
             .Hayes File
             .HAPAB
             .KEEP
     e.    TOXITAPES

     CHEMISTRY

     a.    *CA Condensates
     b.    *Chemical Biological Activities
     c.    Chemical Titles

     EDUCATION

     a.   ERIC
     b.      .Current Index to Journals in Educ
            .Research in Education

     ENGINEERING

     a.   CITE
     b.   *Compendex
10
11,
                                                12
ENVIRONMENT

a.   *AMIC  (Water quality)
b.   SWIRS  (Solid Waste)

GEOLOGY

a.   Bibliography of North American
     Geology
b.   GEO.REF.
c.   Geophysical Abstracts

LIBRARY SCIENCE

a.   EPA Libraries' Holdings
b.   *MARC

MULTI-DISCIPLINE

a.   ISI Source Index
b.   PANDEX

PHYSICS

a.   Nuclear Science Abstracts
b.   SPIN

RESEARCH REPORTS

a.   *NTIS

  SOCIAL SCIENCES

a.   Psychological Abstracts
b.   Sociological Abstracts
NOTE:  *A11 or partially available On-Line.

-------
                     APPENDIX B







This Appendix contains the price schedule based on present



University of Georgia schedules.  Lehigh University and



Informatics charge $45 per terminal hour and a small extra



charge for printing, handling, and mailing.  MEDLINE services,



both searching and off-line printing, ENVIRON, and AMIC are



free at the present time.
      GEORGIA INFORMATION DISSEMINATION CENTER PRICE LIST



                  July, 1972 - June, 1973







CURRENT AWARENESS SEARCHES



These searches keep the user up to date on the current liter-



ature.  Current awareness searches are run whenever current



issue or update tapes are received from the supplier.  Search



results are usually mailed within two days after receipt of



the tape.







RETROSPECTIVE SEARCHES



These searches are scheduled over a two-week processing cycle.



Turnaround time for retrospective batch searches from receipt



of initial profile to delivery of final search results will



normally run two to three weeks. Search results will be mailed



throughout this period, however.






                         B-l

-------
Special demand searches on frequency less than this two-week
cycle will be accommodated only at higher special rates for
which the requester assumes the entire cost of computer runs.
Estimates for special searches are available on request.

SEARCH PROFILES
Information Specialists are available in the Center to con-
struct the search profiles.  There is no charge for this
service.  The question should be a fully descriptive para-
graph describing the subject area to be searched and any
qualifications, such as particular applications which are
to be excluded.  Suggested index terms or words frequently
used and the names of authors who publish in the research
area under consideration are especially helpful.  Please
include a telephone number at which the person requesting
the search can be reached for additional information or
clarification if necessary.

Changes to the profile will be made immediately upon receipt
of a request by a user.  The changes are not retroactive,
however, for searches which have already been run.  There is
no charge for changes to the profile.  The maximum number of
terms which can be processed in one profile is approximately
200, depending on the length of the terms.
                          B-2

-------
SEARCH RESULTS

The search result is a bibliography of retrieved documents.

The output includes the title, authors' names, location of

work  (if available on the data base), primary document

citation, and all index codes or index terms available on

the data base.




Output on 8-1/2 x 11" one-part paper is the standard output.

Output on 4 x 6" card stock is available for 2<= per card in

addition to the prices quoted for the standard output on

paper.  Abstracts will be printed for those data bases in

which they occur for an additional charge of IOC per abstract.




SUBSCRIPTION INFORMATION

Searches are priced on a cost-recovery basis by the Georgia

Information Dissemination Center.  The prices are established

on a per-profile per-search run basis.  For current awareness

searches the price is per-profile per-issue of the data base.

For retrospective searches the prices are set on a per-profile

per-volume searched basis.  Each data base is searched

separately, so that one profile searched against two data bases

would be invoiced as the sum of the prices for the two files.
             i
Partially completed volumes within the current year are billed

at the current awareness price (i.e., the issue price times the

number of issues in the partial volume).  Postage and shipping

charges are billed extra for shipments outside the United States.

Users requesting foreign delivery should specify the preferred

method of shipment.

                          B-3

-------
GIDC Price List, July 1972 - .June 1973




Users are billed by  itemized  statement monthly.   Searches  may

be  entered or  discontinued at any time without penalty.



Subscription discounts, where applicable, will be passed on by

NERC-Cincinnati.   In order to obtain  the discount the  organiza-

tion must guarantee  the minimum subscription.  Unused  balances

are not  refundable.
                CURRENT AWARENESS  SEARCHES
 Data  Base
 CA-Condensates  (CA)*

 CAIN

 Compendex**

 Current  Index to
 Journals in  Ed. .(CUE)**

 GEO.REF**

 GRA  (USGRDR)**

 Nuclear  Sci  Absts.

 Psychological Abstracts**

 Research in  Ed.  (RIE)**

 SPIN**
racts (BA)
ex (BIORI)
(CA)*


Frequency
of Updates
semimonthly
monthly
weekly
monthly
monthly
Price per Profile
Per Issue
$"7
7
7
10
10





quarterly

monthly

semimonthly

semimonthly

quarterly

quarterly

monthly
10

10

 5

 5

10

10

10
 *  Odd-numbered  or  even-numbered  issues  only may be specified.
 ** Abstracts  available.
                           B-4

-------
GIDC Price List, July, 1972 - June, 1973
                  RETROSPECTIVE SEARCHES
Data Base No
Biological Abstracts (BA)
1969 -
Bibliography of North America**
Geology 1961-1968
BIORI (BRI)
1969 -
CA-Condensates-Even Issues
July 1968 (Vol. 69)
CA-Condensates-Odd Issues
July 1968 (Vol. 69)
CAIN
1970 - i
The file prior to 1970 is
treated as one volume
CBAC**
1965-1971
Chemical Titles
1962-1968
CUE **
1969 -
CITE E/E only
1968-1969
COMPENDEX **
1969 -
Geophysical Abstracts **
1966-1970
GEO.REF **
1967 -
GRA (formerly USGRDR) **
1970
NSA
June 1966 -
Psychological Abstracts**
1967 -
RIE **
1969 -
The file prior to 1969 is
treated as one volume
Sociological Abstracts **
1965 - 1967
SPIN**
June 1970 -
.Vols/Year
2
1
1
2
2
1
2
1
1
1
1
1
1
1
1
1
1
1
1
Price per Profile
Per Volume
$ 40
35
80
40
40
70
35
70
35
35
100
35
100
70
70
35
35
50
70
** Abstracts available
                          B-5

-------
It should be noted, however, that NERC-Cincinnati will be



testing and evaluating other services.   Some information



centers, government agencies, and other organizations may offer



their services at reduced prices from time to time, for trial



periods and special data bases.  We will endeavor to investi-



gate each and every situation to determine if changes in our



present operations are warranted, or if some of these new



services should be added.  For example, a limited number of



searches are available from BIOSIS whereby a search of their



files containing 12 years of references, and their abstracts,



are available for 20 cents a citation.   This arrangement is



limited, but if demand warrants, another contract may be



negotiated.
                         B-6

-------
                        APPENDIX C



                     SAMPLE PRINTOUTS







The following represent a typical printout that a user



might receive:







All output that is larger than 8*5 x 11 will be reduced



to that the user need not worry about oversized papers.







Figure C-l is an example from the University of Georgia.







Figure C-2 is a page from Lehigh University.







Figure C-3 represents a current awareness profile; and







Figure C-4 is a sample of three cards that were produced.
                         C-l

-------
PROFILE # 019840-001 JUN. 02, 1972
ACCOUNT # THRESHOLD WT . 00000
DATA BASE SEARCHED: USGRDR V 72:01-09
GROUP
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
G001
GOO 2
GOO 2
GOO 2
GOO 2
G002
GOO 2
GOO 2
GOO 2
G002
GOO 2
G002
GOO 2
G002
GOO 2
G002
GOO 3
GOO 3
GOO 3
GOO 3
G004
G004
G004
TERM
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
(G001&G002I


TYPE
TXT
TXT
ESC
ESC
TXT
TXT
EIT
EIT
EIT
EIT
EIT
EIT
EIT
TXT
TXT
TXT
TXT
EIT
TXT
TXT
TXT
TXT
TXT
TXT
TXT
ESC
ESC
ESC
TXT
TXT
TXT
TXT
EIT
EIT
TXT
TXT
TXT
TXT
TXT
G003

WEIGHT
01000
01000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00002
00002
00002
00001
00001
00002
00001
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
00000
JE-.G004

TERM ;
1
WASTE WATER* '
WASTEWATER*
452
187
SEWAGE*
SLUDGE*
SEWAGE ANALYSIS*
SEWAGE BACTERIOLOGY*
SEWAGE FILTERS*
SEWAGE PUMPING PLANTS*
SEWAGE PUMPS*
SEWAGE TREATMENT*
SEWAGE TREATMENT PLANTS*
EXCESS FLOW PLANT*
URBAN RUNOFF PLANT*
WASTE AND WATER
WATER AND WASTE
AUTOMATIC CONTROL* ^
AUTOMAT* A
FEED-FORWARD*
INSTRUMENTATION
ANALOG PROCESS CONTROL*
DIGITAL PROCESS CONTROL*
INSTRUMENTED
COMPUTERIZED PROCESS CONTROL*
732
943
944
DETECTION AND CONTR01 *
COMPUTERIZED CONTROL*
COMPUTER CONTROL*
COMPUTER AIDED CONTROL*
SFWAGF TREATMENT PL ANTS. MODELS*
SEWAGE TREATMENT PLANTS, SI MULATORS*
SEWAGE TREATMENT PLANT MODEL*
SEWAGE TREATMENT PLANT SIMULATOR*
MINE
MINES
MINING

NUMBER OF ANSWERS 3

-------
                                                                PAGE   1
       DOCUMENTS FROM THE LEADER RETRIEVAL SYSTEM
                                                                07/1B/72
PAGE   1 OF   2!  DOC   1 OF  301  LEADER DOC  72575.

TITLE  CASE FOR CONTINUED STRONG SUPPORT OF BASIC RESEARCH BY THE
       FEDERAL GOVERNMENT
AUTH.  ANON
JOUR.  RES MANAGE V 13 N 4    P 273-9
DATE   JULY 1970
SRCE.  CMPDX7104 EI19544
PAGE   1 OF   21  DOC   2 OF  301  LEADER DOC  61679.

TITLE  CASE FOR CONTINUED STRONG SUPPORT OF BASIC RESEARCH BY THE
       FEDERAL GOVERNMENT
AUTH.  ANON
JOUR.  RES MANAGE V 13 N 4    P 273-9
DATE   JULY 1970
SRCE.  CMPOX7104 EI19544
PAGE   1 OF   21  DOC   3 OF  30!  LEADER DOC 100541.

TITLE  INDUSTRY-GOVERNMENT COOPERATIVE RESEARCH FOR WATER POLLUTION
       CONTROL
AUTH.  HARLIN JR CC
JOUR.  AICHE WORKSHOP. IND PROCESS DES WATER POLLUT CONTR V 2. HOUSTON.
       TEX    P 62-4
DATE   APR 24-25 1969
SRCE.  CMPDX7108 EI50855
PAGE   1 OF   2*  DOC   4 OF  30!  LEADER DOC  71966.

TITLE  HOW RESEARCH IS HELPING TO SOLVE ASIAvS FOOD PROBLEMS
AUTH.  MYERS WI
JOUR.  RES MANAGE V 13 N 3    P 229-38
DATE   MAY 1970
SRCE.  CMPDX7104 EI19008
PAGE   1 OF   2!  DOC   5 OF  30!  LEADER DOC  61070.

TITLE  HOW RESEARCH IS HELPING TO SOLVE ASIAvS FOOD PROBLEMS
AUTH.  MYERS WI
JOUR.  RES MANAGE V 13 N 3    P 229-38
DATE   MAY 1970
SRCE.  CMPDX7104 EI19008
                                   C-3

-------
*
*
          G01NERC0003
          W4 8ENOIT
DISPLAY PROFILE  G01NERC0003
OIR AOMIN           NAT ENV RCH CTR


        0».7J50*******************************
                                             •REPORT*
                                             •PROGRESS
                                             •REPORT*

                                             •SCIEN*
                                             •TECHNI*
                                             LABORATORY
                                             •PROGRESS
                                             •REPORT*
                                             •FACIUT*
                                             •PERSON*
                                             •RESEARCH*
                                             •SCIEN*
                                             CPNTRACUQN.
                                             LABORATORY
                                             PHYSICAL
                                0.750  2  •IN-HOUSE
                                0,000  |  *NEUR*
                                6.666  1  •REACT*
                                0.750  2  •REPORT*
                                0.750  2  •SYSTEM*
                                o.ooo  i  AIR*
                                0.750  2  COMMUNICATIO
                               i 0.75JD  2  COMMUNICATIO
                               '0.750  2  COMPUTER
                                0.000  1  DEFENSE
                                0.660  1  ELECTRONIC
                                0.750  2  FEEDBACK
                                6.666  1  GENETIC*
                                0*750  I  INFORMATION
                                0.750  2  INTERIOR
                                0.750  2  MANAGEMENT
                                0.750  2  MANAGEMENT
                                0.750  2  MANAGIHEM
                                0.750  2  MANAGEMENT
                                0.750  2  MANAGEMENT
                                6.750  2  MANAGEMENT
                                0.750  9  MANAGEMENT
                                6.666  1  SATELLITE
                                0,750  2  STORAGE
                                0.000  0
                             PLANT
                                             RETRIEVAL
                          36 SINGLE WORO EQUIVALENTS ARE IN PROFILE

                      MECHANIZED INFORMATION CENTER
                      THE OHIO STATE UNIVERSITY LIBRARIES
                               C-4

-------
FINN KRt MLLER H6

IS YCUK MIS FIT FOR HUMAN CONSUMPTION?
  « INFCRMATION).
                                                  ;WJ  6ENOIT
                                                  lOlk ADMIN
                                                  !NAT LNV  KCH CTR
INDUSTRIAL ENGINEERING
  VOL. 3 1*71 NU. 11
TERMS: INFORMATION
WJ BENCH
        INFORM AT!
        SERVICES
                                     PAGE   18
IFINN KR,  MILLER
IIS  YOUR MIS FIT; !
                                                   JLIEA
                                                  iISSUfc  il      71
                                 NERCOOQ32062C25  (VOL.3     P.18
                                          3-17-72

                           CURRtNT  AWAKtNESS
                         IRVlCfc OF OSU  LIBRARIES
 NERC00032062025
    DETACH
                                                      HERE
HARVEY P

LIBRARY FOR UNIVERSITY OF riANCl  (INFORMATION
  SYSTEMS).   :
                                                  iwj  BENCIT
                                                  ;D1R ADMIN
                                                  INAT ENV  RCH CTR
BRITISH MEDICAL JOURNAL
  VOL. 1 197<: NO. 5
                                     PAGE
                                                  I
,'HARVtY P
iLIBRARY FOK UHI
                                                  SHEA,  CHI,  VET
                                                  JBMJOA
                                                  I ISSUE 5791    72
                                 NERC00032062027  iVCL.l   P.54
                                          3-17-72  !
                                                  JNERC00032062027
        S N FO RM ATI D N   CURRENT  AWARENESS     i '   DETACH
        SERVICES  A SERVICE OF OSU LIBRARIES   !**
TERMS: INFORMATION
WJ JicNCiT
TISHLtR W                  .        .                WJ  BtfiMOIT
                                                  IDIR ADMIN
WHAT IISFCRMA1ICN UUES MtuICiNAL CHEMIST  REALLY    INAT tNV RCH
  IMEEC? PROJECTIONS FOR FUTURE.                   :    -
JOURNAL OF CHEMICAL DOCUMENTATION
  VOL. 11 IS?! NO. 3
TERMS: INFORMATION
WJ BcNCIT
                                                  |TISHL6R  M
                                     PAGE   134    iWHAT  INFORMATIO

                                                  iCHE,  PHA, ENR
                                                  iJCHQA
                                                  iISSUE 3       71
                                 NEKC00032C62C29  JVCL.11   P. 134
                                          3-17-72
                                                  'NeRC00032062029
        t h
        S
                           CUKRENT  AWARENESS
               tJ i^. ZZ
                     SAMPLE:  BI-WEEKLY PROFILE OUTPUT
                                 C-5

-------
                      APPENDIX D



The following staff members  at NERC-Cincinnati and their

responsibilities can be  used as a guide to the appropriate

person to contact  for help or advice:


    MORTON H. FRIEDMAN        General services and new systems

    DOTTIE F. BROFFT          General reference

    LILIAN S. BOSWORTH        Record keeping, billing and
                              distribution

    MARY LEE BURBAGE          Central cataloging services

    MARY L. CALKINS           On-line,  biomedical and profile
                              analysis

    MAXINE M. SMITH           Batch, current awareness and
             '                 profile analysis



Do not hesitate to contact any staff member whenever necessary.

This system will work only so long as  the librarians and

researchers are willing  to participate and make it work.
 US. GOVERNMENT PRINTINGOFFICE 1972- 759-546/1013    D —

-------