COMMITTEE ON DOCUMENTATION TASK TEAM V - BIOGRAPHICS FINAL REPORT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80B01139A000500020005-2
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 43
• Document Creation Date: December 22, 2016
• Document Release Date: May 3, 2007
• Sequence Number: 5
• Publication Date: February 1, 1966
• Content Type: REPORT

Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
U N I T E D S T A T E - S I N T E L L I G E N C E B OAR D
COMMITTEE ON DOCUMENTATION
TASK TEAM V -- BIOGRAPHICS
FINAL REPORT
T/V/R-1
1 February 1966
Group 1
Excluded from automatic
downgrading and
declassification. 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
SECRET 25X1
T/V/R-1
1' February '1966
U N I T E D S T A T E S I N T E L L I G E N C E B 0 A R D
COMMITTEE ON DOCUMENTATION
TASK TEAM V - BIOGRAPHICS
MEMORANDUM FOR: Chairman, Committee on Documentation
SUBJECT: Report of Task Team V
1. Attached is the report of Task Team V for your consideration.
2. The Team has attempted, in an evolving interpretation of its
Terms of Reference, to present realistic recommendations while
developing in some depth a substantive description of the problems for
the use of interested agencies. While the overall report is classified
SECRET Annex 2 has been given a lower 25X1
classification to permit wider distribution to U. S. Government officials.
3. A large file of information, monographs on various aspects of
the problem (National Agency Check System, search strategies, data con-
version techniques and experienced costs, SCIPS studies of PI files,"
etc.) is available in or through the CODIB Support Staff.
4. It is recommended that the Task Team be discharged on CODIB
acceptance of this report. A formal mechanism for continued exchange
on biographic problems and techniques is, however, contained in the
RECOMMENDATIONS.
support.
5. My thanks tol for his extensive and imaginative
Attachment:
Task Team V Report
Chairman, Task Team V
Group I
Excluded from automatic
downgrading and
~assif.ication.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2 25X1
SECRET
U N I T E D S T A T E S I N T E L L I G E N C E B OA R D
COMMITTEE ON DOCUMENTATION
TASK TEAM V -- BIOGRAPHICS
Table of Contents
Purpose
Summary of Findings
Recommendations
The Nature of the Problem
Counterintelligence and Security
Positive Intelligence
Annexes:
Page.
1. Glossary
2. Proposed Approach to the Lachine Recording of Personal Names
Attachment 1: Machine Recording Techniques for Personal Names
3. Biographic Index, Facts Summary
4. Data elements. in Team Member Agency Records
S. Examples of Name Variants
6. Terms of Reference
7. List of Task Team Members
SECRET,
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
SECRET/( 25X1
TN/R-1
1'. February 1966
U N I T E D S T A T E S I N T E L L I G E N C E B 0 A R D
COMMITTEE ON DOCUMENTATION
TASK TEAM V - BIOGRAPHICS
FINAL REPORT
PURPOSE
The objective of this Team was to "identify means for improving
the storage, retrieval and exchange of information from the major
name files and related data files in the Intelligence Community."
SUMMARY OF FINDINGS
1. Improvements in the speed and quality of biographic
information processing involving interagency exchange on U. S.
citizens and foreign nationals are necessary to further improve security,
and to afford policy makers and analysts better response from biogra-
phic intelligence files on foreign nationals of interest from a variety
of angles--military, subversive, political and scientific. The Team
finds that use of computer techniques and inter-agency telecommunica-
tions links may provide significant improvements.
.2. There are,'however, profound, complex problems and
significant costs in making major changes in the large biographic
holdings of community concern, particularly if the changes involve
conversion to computer systems.
3. There are three basically separate, but somewhat over-
lapping biographic areas: Counterintelligence* (CI), Positive
Intelligence* (PI), and Security*. Name finding* and name
searching* take place in all three. (See Annex 1, Glossary,for
definition of these and subsequent asterisked terms).
1E. The major indexes* considered by the Team ranged from
300,000 unit records (Secret Service) to 50,000,000 (FBI). These
now total about 170,000,000 unit records of interagency concern, and
are growing at the rate of over eleven million yearly. (See Annex 3).
5. An average of 30,000 requests concerning individuals are
made against these indexes daily. Of the 30,000 requests, about
SECRET (Excluded from automatic 25X1
downgrading and
deelassi fi eati nn _
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
one-half are made between agencies (see footnote) and the other half
are processed within the agencies where the requests originate. The
30,000 requests, plus file maintenance procedures, generate 155,000
name searches each day. About one-half of the 15,000 requests made
daily between agencies result in a no-record* response.
7. Agencies in the Washington area are answering security name check
requests from each other within two to eighteen days, portal-to-portal,
with an overall average response time of nine calendar days.
Considerable additional time and cost is involved in delivering the
results to the original requester within the requesting agency. The
timeliness of response is believed to vary widely owing to volume,
personnel costs, and a combination of many other factors unique to
each agency. It is difficult to measure the actual loss to the
government in terms of personnel not taken on board, personnel taken
on board waiting for appropriate clearances, personnel not utilized
in a contact or contractual sense because of the slowness of the
system. These are intangibles that only the various elements of the
respective agencies can weigh within the purview of their own
responsibilities and requirements.
8. In the area of name searching, significant quality and time
improvements may be obtained through automation and use of tele-
communication links. No major name index in the intelligence
community has yet been fully automated. Therefore, proof of
success has not been conclusively demonstrated. Several agencies
are at various stages in developing systems with practical appli-
cations anticipated in the near future.
9. The critical problem in any large name index used for
name searching is the way in which personal names are recorded,
filed, and searched. Any planning for index mechanization must
emphasize this aspect. The success of an improved interagency name
Note: Since these statistics were gathered, the number of inter-
agency name requests submitted by several agencies has increased on
the order of 50% during the last several months mainly as a result
of several new programs.
25X1
SECRET/ 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
SECRET/
check exchange system based on telecommunications coupled with
computer search requires a common approach to recording personal
names and certain additional basic identifying data.
10. Name Finding activities could be improved through increased
understanding resulting from the exchange between agencies (at both
the user and system planning levels) of information about the
nature and purpose of eachotherts specialized files as well as
the exchange of data files in certain cases and interchange of
information on manual and,ADP techniques for improving speed and
flexibility of response.
11. The team agreed that the professional interchange derived
from the Task Team effort was highly valuable to each member in
providing new insights in manual and machine techniques, inter-
agency channels, sources of information, and policies of other
agencies.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
RECOMMENDATIONS
IT IS RECOMMENDED THAT:
1. USIB urge those agencies with large name indexes used for
name searching in the National Agency Check system and in Positive
Intelligence applications of Community interest to continue to
strive within their organizations for index mechanization wherever
it is found to be feasible and practical (recognizing that several
agencies are already in various steps of development in this area).
The findings and report of,this Task Team should be used as a
point of departure.
2. In conjunction with Recommendation 1, USIB request each
agency to study the feasibility of establishing telecommunications links
within the National Agency Check complex to facilitate the exchange
of requests and replies.
3. USIB request those agencies engaged principally in Positive
Intelligence activities to study the feasibility of tying into the
Washington area LDX system for the exchange of Positive biographic
intelligence.
4. Those agencies which plan to convert large manual
biographic indexes to computer-based name searching systems consider
the approach to the machine recording of personal names outlined in
Annex 2.
S. The CODIB Support Staff be directed to prepare and maintain
current publications to inform users of biographic information in
the community of the characteristics of each major collection, and
the procedures and channels for getting service from each, within the
limits of security classification and need-to-know prescribed by
each agency.
6. The CODIB Support Staff also serve as the vehicle for
informing those agencies developing new computer data files, par-
ticularly in the PI biographic area, of the format and coverage
requirements of others in the community to reduce unnecessary dupli-
cation and coverage gaps.
7. DIA expand its program for the processing of military
personality information to meet the needs of the PI community. This
should include the processing of open source material and should
provide for an EDP file of personality information as well as hard
copy backup for such a file. This can be coordinated by DIA with a
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
group composed of representatives of NSA,
service branches.
State and cognizant 25X6
8. The Task Team III (or its successor) be tasked to study
those various programs exploiting open source scientific and technical
information, which. generate personality information of positive
intelligence value as a by-product. In conjunction therewith, a
coordinated program should be developed using EDP methods to provide
machine indexes of the bibliographic data processed by any organiza-
tion in this field, so that the personality information is accessible
to a recipient in machine. form, with quick follow-up to the. translated
source. .
9. Two or three day seminars be held-semi-annually (with chairmen
rotating from the respective agencies) on the progress of the various
agencies in the biographic field, with working sessions for groups
with specific problems (such as CI, Security, PI, Communications,
the state of relevant technology, software, control techniques, and
other functional or technical aspects).
SECRET?
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
Secret Service, Immigration and Naturalization Service (INS), and
Civil Service Commission (CSC).' The r biographic records
are contained in the files.of the CIA 25X6
THE NATURE OF THE PROBLEM
1. The Intelligence Community has for many years collected an
ever-increasing amount of information about individuals from a great
diversity of sources through a large number of channels, and has
stored":this data in a variety of retrieval systems in diverse formats.
These have traditionally taken the form of index references, either
self-contained or leading to. dossier files or individual documents.
The Team decided, as a point of departure, that the relative pay-off
in system improvement would be higher in respect to the larger
biographic files in which there is a high degree of activity and
interagency communication. Thus, many of the smaller files studied
by SCIPS (the Staff for the Community Information Processing Study)
were not included.
2. There are three types of major biographic indexes and files
now in operation. They are the Positive Intelligence, Counterintelli-
gence and Security holdings. There is relatively little exchange of
requests between the PI biographic files and the Security files,
moderate exchange between the CI and PI communities and frequent
exchange between Security and CI. The Counterintelligence (CI)
biographic system centers around the foreign counterintelligence
repository of 0 the domestic counterintelligence holdings of
the FBI. The security and PI holdings of the agencies referred to 25X6
in this report also lead" to CI data in some degree. The interagency
exchange of Security data centers around the name search type
operations performed by = State, Army, Navy, Air Force, FBI,
both the Positive as well as the CI/Security activity. However, the
bulk of the requests in both areas involve name searching (above 95%
in the CI/Security area and about 80% in the PI area).,
and Air Force/Foreign Technology Division (FTD).
3. There are important and fundamental differences between, and
some similarities in, the basic operating procedures and kinds of
searches that are made in the PI systems versus the CI/Security
systems. The PI biographic systems are deeply intertwined with,
and in many cases actually part of, larger intelligence collection
and storage systems which are mission, subject or area oriented.
In contrast, the CI/Security systems are clearly oriented to the
heavy use of.name searching among alphabetically ordered biographic
indexes which, in most cases, lead to dossier files. The Team
determined that there is name searching and name finding going on in
DIA, NSA/Office of Central Re erence, Department of State
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
4. The critical problem in name searching large manual or
machine indexes involves the ways in which personal names are reported
and stored for retrieval. This is a spelling phenomenon, particularly
in PI and CI indexes,which may be classified in two parts:
a. Name Variants: Different spellings of the phonetically
same surname in the original language (SCHUKOW, CHOUKOV, DIUKOV,
DZHUGOV, JOUKOFF, YOUKOV, ZHJUKOV, ZHUKOV, etc.). Given name
equivalents, diminutives or abbreviations are also considered
part of the name variant problem (WILLIAM, WILHELM, WILL,BILL,
WM.).
b. Name Variations: Different conventions in recording
and using parts of names (name elements), for example: Fidel
CASTRO; CASTRO, Fidel A.; CASTRO y RUZ, Fidel Alejandro; John
Taylor BROWN; BROWN, J. Taylor; BROWN, John T.
S. The difficulties in handling the name variant/variation
combinations are particularly crucial in those systems in which the
preponderance of names are on foreign nationals, or U. S. citizens
where control of the source reporting (e.g., employee applications,
identification of individual by social security or other number,
etc.), is not available. The reasons for the corruption of name
spellings received by the majority of agencies considered in this
report reflect the real world of intelligence biographies - foreign
and domestic. The causes include different transliteration systems
between countries (and even within a given country),usage and custom,
mistranscription in rewriting names, typographical error, telegraphic
garble, and phonetic renditions of names overheard. Examples of
these problems are given in Annex 5.
6. Given this situation, the possible combinations and
permutations of name variants/variations are unlimited and, more to
the point, unpredictable. Thus no formal linguistically based
system for reducing name variants to a common denominator has been
found wholly adequate for reliable storage and search by those
agencies dealing primarily with uncontrolled sources. A pragmatic
approach to this problem - called name grouping - is being developed.
See Annex 5.
7. The problem is minimized for those agencies which have
numerical identifiers (such as social security number or date of
birth) in the large majority of their index records. The name
variant problem cannot be escaped even so, since these agencies are
recipients of name search requests on foreign nationals or U. S.
citizens on whom the requesting agency has no control number, and
quite possibly a different spelling?of the name.
SECRET 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
S. The high proportion of common names adds to the difficulties
in large indexes, foreign and domestic. For example, in one multi-
million card file on Soviets containing over 300,000 different surname
spellings, some 1,500 common surnames account for over 50% of the file.
In the case of Vietnam, 54% of the people in the Red River Delta area
have the surname NGUYEN; 85% of the Vietnamese population is represented
by twelve surnames, with the balance less than 300 clan names.
9. The lack of identifying data on named persons is intimately
related to the name variant and common name problems for those agencies
without source control. While Annex it shows the categories of iden-
tifying data recorded if available in the reporting, most foreign and
domestic reporting deals with vaguely identified personalities. It is
therefore impossible to develop rigid rules on what constitutes the
minimum identifying data required. Each agency, in recognizing these
problems and the nature of its own index, forms its own rules regarding
minimum identifying data for recording, and the depth of search according
to the nature of the request.
10. The above indicates what is involved in the quality of name
searching. In the past, many agencies have reduced their capability
for quality search in manual or machine systems,: (e.g., by restricting
the amount of data recorded). All involved in this Task Team recognize
the need to observe the following principles:
a. Preserve complete name spellings, and record name element
components in a consistent format for either manual or potentially
mechanized indexes. If an agency is planning the latter, the
methodology for the formatting of individual name elements as
explained in Annex 2 should be considered.
b. Retain in the index record all identifying data which
assists in distinguishing persons of the same or similar name
from one another. Such data elements as sex, date of birth,
place of birth, citizenship/nationality, occupation/profession,
location, social security number are generally agreed to be
desirable, if available, though additional amplifying data
further distinguishing the individual should be recorded -
regardless of the feasibility of machine search - for human
analysis.
c. Follow the progress of the TTname grouping" approach to
the name variant problem and, should it prove operationally
successful, take advantage of already developed computer
techniques to capitalize on the linguistic effort expended by
the Government and private agencies for this purpose.
SECRETS 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
11. It was also found that name finding requires substantially
more time and effort per search. This is true because a name finding
request generally must be structured in a more complex fashion and
requires a more involved search procedure.
12. The Team decided to consider the CI and Security systems
as one area and the PI biographic systems as a separate area for the
purposes of developing the facts, defining the problems, and making
recommendations in this report.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
25X6
25X6
COUNTERINTELLIGENCE AND SECURITY
1. The Security activity clearly stands out as a network of ten
large indexes which are heavily used. Name searches are conducted mainly
for granting security clearances for a variety of reasons such as employ-
ment, contact, association, contract, etc., and at a variety of security
levels. An agency's requirement to grant such a clearance results in the
selective checking by that agency of an average of s other agencies.
The major agencies involved in this program include State, Army, 25X6
Navy, Air Force, NSA, FBI, Immigration and Naturalization Service, Secret
Service, and the Civil Service Commission. The latter three listed are
not part of the USIB Community but, in formulating the Team, it was 25X6
recognized that these agencies are an integral and significant part of.
the National Agency Check (NAC) Program. Of the approximately 114
million unit records in the Security holdings, these three agencies
hold approximately 50 million (I&NS, 37 million; CSC, 12 million;
Secret Service, .3 million). Of the 28,000 requests generated daily in
the CI/Security System, approximately 8,000 are generated by these three
agencies.
2. Intertwined with the Security request activity are the foreign
and domestic Counterintelligence activities centered respectively in
CIA and FBI. There are, however, some CI functions in most of the
other agencies represented. The normal purpose of the Counterintelli-
gence biographic name check activity, as it takes place between the
agencies, is to determine the presence of information about: an
individual of interest to the requesting agency for some
counterintelligence reason (e.g., relating to hostile activities of
foreign intelligence services and the Communist Party),
orei.gn counterinte ligence responsibilities under NSCID 5/3. Security
indexes lead primarily to investigative cases and criminal records,
predominantly on U. S. citizens. In spite of the fact that requests are
made of the CI/Security holdings for different reasons, the nature of
the requests and the structure of the data bases involved are
substantially the same.
present time. This is true of the Office of Security
in the process of converting their indexes to machine lantauaize at the
3. The various contributing agencies are listed in Annex 3 along
with a set of facts about the respective size, type, growth, activity,.
etc.,of their CI/Security files. It can readily be seen that the size
of the various indexes ranges from 300,000 in the case of the Secret
Service to over 50 million in the case of the FBI. Most of the unit 25X6
records are still on 3 x 5 cards. Some of the individual agencies are
SECRETA 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
The Army and Navy indexes are already on IBM cards,
and the NbA e ity Records are on magnetic tape. As a result of
recent DoD action, the Army,.Navy, and Air Force are completing plans
to merge their three index holdings on punched cards by mid-1966.
Consequently, the Air Force will shortly convert its 3 x 5 index
cards to IBM cards for insertion into the common DoD index. This
DoD index, although to be in machine language (IBM cards), will, in
its initial phase of development, be searched manually. The Immigra-
tion and Naturalization Service is presently studying a program to
convert its index to machine language and prepare for a machine-based
system. This is likewise true of. the Secret Service, FBI, and the
Civil Service Commission.
4. The CI/Security indexes are growing at approximately 7%
per year. This means that they will double in size within ten years
at the present rate of growth. Of particular significance is the
fact that the 28,000 requests made per day in these indexes (along
with the daily maintenance) results in over 120,000 actual name
searches being made, mostly manually, in these indexes each day. Of
these 28,000 requests, approximately half are made between agencies.
From these 14,000 name checks flowing between the agencies, more
than half result in a no-record response by the responding agency.
5. The elements of the CI/Security search process considered
by the Team include the size and the activity between the agencies,
the accuracy and form of the requests and responses, as well as the
time that it takes the agencies to respond to each others' requests.
The Team noted the fact that there are literally dozens of name check
request forms now being utilized by the various agencies. In
observing some of these typical and most widely used forms, the Team
found that certain basic data such as name, place and date of birth,
service serial number, social security number, sex, etc. were included
on each form. The Team considered a study of the need for a single
name check form to be used by the various agencies. It was considered
more important, however, to examine the data elements used and what
rules should be applied to their control. These considerations become
increasingly critical as the agencies move toward greater use of machine
language.
6. To obtain a reasonably dependable determination of the kind
of response time in which the various agencies were providing informa-
tion to each other, a sample survey was made of 3,000 individual typical
routine requests. Emergency and priority requests are handled by
every agency in a matter of min or hours depending upon the results
of search. The FIB, I&NS, CSC, and Army participated in this25X6;t.
25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
These agencies tabulated the response times of requests from each'
other as well as from the Navy and the Air Force. The interagency
response time varies from two to eighteen days with the average of
all the agencies being nine calendar days. There were factors which
the Team recognized as causing possible aberration in these figures:
hand carrying of the requests by liaison personnel, the variations
in the depth of searches, (i.e., on the head* or checking different
possible spellings of the same name) and the researching of the
files by requesting agency personnel on the premises of the answering
agency. In spite of these, the Team feels that the nine-day figure
is a reasonably accurate estimate of the average time (within a day
or two) required for processing of the great bulk of the name checks
being made in this system.
7. It should be noted that the response time referred to
above does not include any internal processing time, in or out, by
the various requesting agencies. The time was measured in all cases
from the day the request left the requesting agency to the day that it
returned to the requesting agency. This time included the mail time
plus that required to make the index search by the responding agency
and the analysis of files in the case of possible identification.
Based on informal observations of the various Team members it appears
that, in the great majority of these cases, there is far more time
spent processing these requests within the requesting agencies
(i.e., from the time the original requester - e.g., analyst, investi-
gator, Ambassador, etc. - sends out his query to the point where it
re-enters the agency and is provided to the ultimate user) than the
nine-day figure of external processing time explained above. To
determine the extent of the internal processing lags and the reasons
therefor was a task far beyond the capability of the Team.
8. Many CI requests are answered from materials that are not
processed into the files, such as directories, working aids, etc.,
or from material too current to be in the file, such as today's
newspaper. Some files are restricted by security classification as
to what can be processed. Research in such a limited source file
often gives incomplete or out-dated information. It is doubtful
that any single file, whether it be computerized or manual, can
ever be considered a complete or sole source for biographic information.
9. It was not possible for the Team to consider specifically the
relative merits of: (a) the improvement of the manual systems within
each agency, (b) the potentials in automation of the index systems
within each agency, and (c) the system efficiency that might be
realized by the institution of a machine language communication system
between the various agencies. These are tasks requiring management
supported feasibility studies, dominated by the professionals within
each agency, in terms of the unique history and problems of each.
SECRET/ 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
POSITIVE INTELLIGENCE
1. The positive intelligence (PI) biographic files can be defined
as those files in the intelligence community that have been developed
to support the evaluation and production of foreign intelligence. The
files are used primarily by government reports officers, researchers
and policy makers in establishing or determining facts and reaching
decisions in the fields of foreign affairs and defense. The personalities
contained in the community's PI files are predominantly foreign nationals.
The team concentrated its review upon the major files of the PI community
(see Annex 3) on the assumption that the problems involved in the areas
of storage, retrieval and exchange would also exist in other PI files
and because a large number of the smaller subject-oriented PI files
contain the same source material. Development of these smaller files
may often be the result of the problems of size, immobility and acces-
sibility that have developed over the years in the large PI files.
2. The management of a PI file can be broken down into four
functional areas: collection of source material, selection of informa-
tion for the files from the source material collected, processing of
.information into the files, and dissemination of information from the
files. The task team concentrated mainly on the area of dissemination
and procedures for searching information requests. Since the other
three areas have a definite effect on dissemination,they were reviewed.
a. Collection - Literally hundreds of thousands of source
documents are received by a PI file system each year. They will
be in English or in a foreign language and each must be read and
evaluated. These sources will include the following: newspapers,
press services, foreign journals, books, government publications,
radio broadcast information and the entire intelligence output of
the US intelligence community. A portion of this material will
be of a very current nature, having been produced the same day or
the previous day.
b. Selection - The basic criterion of any agency for
selecting an item for a PI file is whether or not the item
supports the foreign intelligence effort on a particular
country or area. Every organization has its own standards
for selection based on the mission it is supporting and budgetary
limitations. The same source document is frequently processed
by different PI organizations. The amount of information that
is already available in authoritative sources such as military
registers, directories, etc.,will often determine what will be
25X1
25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2 25X1
SECRET/
selected for the files. On areas such as the USSR, China, etc.,
a great deal of open source and classified intelligence will be
processed because reliable directory type information is not
obtainable. There is an overlap of information in PI files
because the different file systems support the same requirements,
or because the personality mentioned in the source report meets
the selection criteria for two different requirements: e.g.,
CIA and State have an interest in military personalities who
are prominent in other fields such as politics, science, space,
etc., whereas DIA and NSA are interested in the same person be-
cause he is in the military field. There is no assurance,
however, that a personality mentioned in a source document will
necessarily be processed into a PI file.
c. Processing - Most PI organizations process an abstract,
page or the entire document into its file. The main file may
be in the form of a dossier or a structured alphabetical file
which can be approached directly or through a card or machine'
index. The file items may be photocopy, microfilm, multilith,
typed abstract, or the original document. Because of the
timeliness of some information (the same day or previous day)
and the current nature of some requests, it is necessary
either to process this information on?a priority basis and get
it into the file quickly or to arrange support files that will
give a researcher quick access to this information. The file
item may be indexed for a particular computer file at the
same time it is processed into a manual PI file system. The
personality name as it appears in a source document is often
either incomplete or misspelled and the name is researched
and corrected wherever possible. Routine processing time from
selection of an item to filing the item will range from an
average of seven to twenty days.
d. Dissemination - The dissemination of information from
a PI file will be usually one of two types: the ad hoc research
of a specific request for information on personalities or the
production of biographic intelligence by the PI element itself.
25X1
-11
25X1
typical research requests. These requests could be grouped into the
3. In order to analyze the biographic request activity, the
team members from DIA, CIA, NSA, and State each exchanged a group of
SECRG~
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BOl 139A000500020005-2
organizations. The requests involved either name searching, where the
identification or complete information on a named individual is requested,
or name finding, where the name of the person(s) is either missing or
so badly misspelled that research on the other data elements. available,
such as his position, location, organization or persons associated with
him, is required.
4. The group arrived at the following conclusions as a result of
its analysis of the requests and its discussion and review of the file
systems.
a. PI requests are basically 20% name finding and 80% name
searching. It takes more time to research a name finding request,
particularly if identifying data in the request is incomplete.
A name finding request may generate a list of hundreds of
personalities of possible relevance. Many name searching requests
require the analyst to use various name finding approaches. If
the requester wants a complete identification or biographic sketch
on a person holdin a overnment position or an organizational
position, (e.g.,
25X1
it is necessary to check the recor s y
organization. is will insure that any documents reflecting his
change in the organization by position but not name might provide
the desired information.
b. A computer system that is developed to process PI
information should provide the researcher with both name-searching
and name-finding approaches. In a manual system this is usually
accomplished by two file systems: a name file in which the
personality is searched by his name, and by files that are set
up by the other data elements such as organization, location,
occupation, etc. In a computer file of limited size, e.g., one
or two magnetic tapes, where the maximum search time is fixed,
a single file containing the name and all pertinent data elements
may be adequate. This will not be true of a file system containing
millions of personality records growing at the rate of a million
records per year. If name finding approaches are not provided in
a large system, the result may well be the development of a new
group of subject-oriented files, either manual or computerized,
similar to those that presently exist, to meet the needs of
specific components of an organization.
c. Many PI requests are answered from materials that are
not processed into the files, such as directories, working aids,
etc., or from material too current to be in the file, such as
today's newspaper. Some files are restricted by security classifi-
25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BOl 139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
cation as to what can be processed. Research in such a limited
source file often gives incomplete or out-dated information.
It is doubtful that any single file, whether it be computerized
or manual, can ever be considered a complete or sole source for
biographic information.
,I. "On the head" name search (i.e., researching the
name only as it is spelled in the request) cannot always be
considered adequate in the PI areas. This implies that
information will be found under the name spelling in the
g'(_equest; and since PI name spellings do not usually come from
g,l'Cieial sources, they are more likely to be incorrect than
names found in those indexes where source data is
controlled. As mentioned previously, an effort is usually
made to correct the spelling before an item is filed, and
the same effort is and must be made when performing research.
U. The PI request is often of a current and timely
nature, requiring an answer within an hour or even minutes if
it is to be useful to the requester. Routine requests are
normally answered within a day. Some extensive research
projects may involve thousands of names and require weeks or
months to complete. The need for rapid response is one of the
reasons a VI element often cannot rely on another agency to
answer its requests. This is one of the reasons for the
overlap found in the various PI files. The present
communications between agencies is not adequate for quick
exchange o' classified information.
g. The community could benefit from a coordinated effort
in the production of military biographic information from open
sources.
=) . it was not possible for the Team to consider specifically
[he relative mertis of: (a) the improvement of the manual systems
within each agency, (b) the potentials in automation of the index
systems within each agency, and (c) the system efficiency that
might be realized by the institution of a machine language communi-
cation system between the various agencies. These are tasks
requiring management supported feasibility studies, dominated by
the professionals within each agency, in terms of the unique history
and problems of each.
If CRE 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
Next 1 Page(s) In Document Denied
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
C-O-N-F-I-D-E-N-T-I-A-L
ANNEX 2
PROPOSED APPROACH TO THE
MACHINE RECORDING OF PERSONAL NAMES
INTRODUCTION
1. A USIB endorsed approach to the machine recording of personal
names is proposed, subject to qualifications outlined below. The pur-
pose in proposing the adoption of this approach is to insure that those
agencies automating their indexes for name searching purposes, where
continuing inter-agency exchange is involved, recognize the problems
of identifying the elements of personal names in machine recording,
and adopt similar, if not identical, logic in storing, maintaining and
searching these name elements. This is necessary if the agencies
concerned are to exchange, eventually, formatted queries via tele-
communications facilities, for input to automated biographic indexes
with little or no programmed format conversion and manual reprocessing.
2. In suggesting this approach, it is recognized that significant
problems could confront those now using or developing manual or. EAM
indexes. It also is not intended to preclude the immediate adoption
of electrical communications between agencies for speedier search
request response.
3. The proposed approach is subject to the following qualifica-
tions'and assumptions:
a. It is intended to apply only to those major PI, CI,
and Security indexes consulted regularly on an inter-agency
basis (e.g., Major NAC indexes, Biographic Register., NSA/CREF),
though the approach to personal name recording should be of
value as well to those developing internally-used index
systems.
b. The approach assumes computer data recording and
manipulation, as opposed to punched card systems (the rules
can only apply to variable length records and computer program-
ming techniques to manipulate data elements internally).
c. The proposal assumes that the rules would be applied
only at that point when an agency begins machine language
preparation of new input for eventual computer operation, and
is not intended to apply to existing punched card records
which, however imperfect, may be the only. means for converting
an existing file to a computer data base.
C-O-N-F-I-D-E-N-T-I-A-L
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
C-O-N-F-I-D-E-N-T-I-A-I,
ANNEX 2
- 2 -
4. It is felt that those agencies contemplating eventual
conversion to computer search systems should evaluate the desirability
of recording personal name and related identifying data in variable
length input format for computer processing. This will accomplish
the beginnings of a data base which will not require later keypunch
conversion, provide means for manipulating and editing index informa-
tion not possible in EAM or manual systems, and will provide also
the capability to print or punch index records as a byproduct to
keep up manual and EAM systems during the interim stages.
5. Attached hereto is a description of machine recording
techniques classified FOR OFFICIAL USE ONLY.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
FOR OFFICIAL USE ONLY
- h -
Annex 2
Attachment 1
MACHINE RECORDING TECHNIQUES FOR PERSONAL NAMES
1. Described below are some of the problems involved in the
recording, filing, and searching of personal names and suggested
solutions. The problems in the handling of personal names by
electronic data processing are dealt with specifically and considera-
tion is limited to large personal name indexes where (1) point of
retrieval is on name spelling, (2) the quality of name recording,
i. e., spelling and/or completeness of name, cannot adequately be
controlled, e.g., names recorded in newspaper articles, heard on
radio broadcasts, copied from documents, or obtained from second
or third hand sources whose knowledge of the name spelling and/or
completeness may not be reliable, and (3) where additional identifying
information such as date and place of birth, occupation, etc., may
not be consistently reported, and such specific numeric controls
as social security number, military service number, drivers registra-
tion number, etc., do not apply. These conditions are found not only
in the names recorded in an index, but also in the names received as
requests for information.
2. The first problem in recording personal names is to define
the basic order in which the name parts will be recorded. That is,
shall the name be recorded in the English signature style (given
names followed by family name) or in telephone book style (family
name followed by given names)? If the index in question stores
names of all nationalities (very'few do not), either style of
recording will require some rearrangement of name parts at the time
of recording. For example, Hungarian and Chinese name signatures
are quite different from the English signature style. That is,
the Hungarian or Chinese name is usually written with the family
name first, followed by the given names.
3. Regardless of the recording style selected, it is important
to define various elements within a name and to identify them in some
manner when they are recorded. The definition and identification of
various name elements is necessary to (1) adequately describe
recording rules to reporters and recorders as they apply to names of
various nationalities, (2) facilitate accurate filing of the name
records in the index, (3) permit accurate machine processing
(sorting) for alphabetic listings, etc., (4) and to facilitate storage
and retrieval (search) of name records by computer.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
FOR OFFICIAL USE ONLY
- 2 - Annex 2
Attachment 1
It. Many different codes, symbols, characters, or fielding
techniques may be used to identify various name elements. However,
if a printed version of the name is to be read by persons not
normally associated with the EDP environment, it is preferable to
use common punctuation which can easily be interpreted by the
customer, i.e., use a period after a single alphabetic; character to
identify an initial as opposed to a single character name or
particle.
5. Definitions of various name elements which should be
identified when recording the name follow:
a. NAME: That word or combination of words used to
identify a person.
(l) The minimum field length for recording
the name should be forty characters. Although many
names can be recorded'in less than 40 characters,
the truncation imposed upon lengthy names by, say, a 20
character limit, often eliminates the very elements
which provide discreteness. Such system-imposed
restraint increases the number of name records which
will be retrieved in a search. Additionally, it
often imposes pre-input editing to be sure that
critical elements of the name can be recorded in the
field size allotted. For example, the name
Evangelica Concepcion Rodriquez y Gonzalez contains
42 characters including spaces and without any
special characters to identify various name
elements. The usual pre-input edit of this name
would probably reduae it to RODRIQUEZ, EVANGELIC,
thus making it impossible to distinguish this
Evangelica Rodriquez from any other Evangelica
Rodriquez. It the name were not pre-input
edited, but merely truncated by the input
typist or arbitrarily by the machine, the
entry RODRIQUEZ Y GONZALEZ, EVANGELICA
CONCEPCION would be truncated to RODRIQUEZ Y
GONZALEZ which is even less discrete.. Forty
characters permits recording of the family
name and most of her given names, i.e.,
RODRIQUEZ Y GONZALEZ, EVANGELICA CONCEPC.
FOR OFFICIAL USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
FOR,OFFICIAL. ?USE ONLY
- 3 - Annex 2
Attachment 1
b. SURNAME: The word or words which comprise the element
of a name commonly referred to as the "last name" or "family
" including initials, abbreviations, and particles (defined
name
,
below) if reported'as part of the surname. The surname is that
element of the name which governs the primary position of a name
in an".alphabetic file. Surnames containing more than one word
are referred to as "compound" or "Multi-Word" surnames.
(1) Because surnames often contain more than one
word, and in view of its basic importance to the filing
and subsequent finding of the name record, it is necessary
to identify which part of the complete name is the surname.
In the examples which follow, surname is printed first
followed by a comma to show the end of the surname. If
some such method of surname identification is not used,
surnames which contain more than one word cannot be
distinguished from those with only one word followed by
first name.
Examples: BROWNE, T. R.
CESPEDA Y LOPEZ, JUAN
KAMAL AL DIN, MOHAMED
c. GIVEN NAME: The word or words in a name commonly
referred to as the "first," "baptismal," "Christian,"
"middle," or "patronymic," etc. Initials and abbreviations
are included. Given Names dictate the alphabetic position of
a name record within like surnames. Therefore, particles,
titles, and telecodes (defined below) are not included in the
definition of "Given Name."
(1) Whether the name parts being recorded are
called "Surname and Given Name" or "Clan Names"'or
whatever, is irrelevant. It is important, however,
to identify which word or words in a name are to be
used as the primary storage or search element (Surname)
and which are to be used secondarily, (Given Name).
(2) Note, in the following list of names recorded
without commas, that "compound" surnames cannot be
distinguished by a computer from non-compound surnames
and, therefore, the second word of the compound surname
is likely to be used as a given name.
FOR, OFFICIAL',- USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
FOR.OFFICIAL.USE ONLY
Annex 2
GARCIA LOPEZ JOSE should be GARCIA LOPEZ, JOSE
MAC DONALD HENRY MAC DONALD, HENRY
RODRIGUEZ L. JUAN RODRIGUEZ L., JUAN
ST. CLAIR ROMAN LUIS ST. CLAIR ROMAN, LUIS
STA. ANA RAUL STA. ANA, RAUL
d. PARTICLES: Particles include the articles (la, der,
etc.) prepositions (de, von, etc.) and conjunctions (und, etc.),,
foreign equivalents of the English the, of, and, etc., which
have..-not become an integrated part of the name.
(1) Particles are usually ignored in the filing of
names because they may be different each time a name is
reported and recorded or may at times be completely absent.
Therefore, if the particles were used in:detcrmining the
alphabetic file position of the name, the same name would
be filed in different places.
Examples: GARCIA LOPEZ,, JUAN
GARCIA (Y) LOPEZ, JUAN
GARCIA (E) LOPEZ, JUAN
(DE) GENNARO, GUISEPPE
(DI) GENNARO,, GUISEPPE
GENNARO, GUISEPPE
KAMAL (AL) DIN, MOHD
KAMAL (UD) DIN, MOHD
KAMAL (EL) DIN, MOHD
KAMAL (ED) DIN, MOHD
(2) For the above reasons, it''is important to
identify, those words in a name which, are particles.
When they have been properly identified, the computer.
FOR OFEIC-IAL ;USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
FOR OFFICIAL.USE ONLY
Annex 2
5 - Attachment 1
processing of these names will be able to facilitate
appropriate alphabetic sequence.
'(3) For name searching purposes, it is particularly
important that particles appearing in the given name field
be identified (for example, by enclosing in parenthesis)
so that they are not confused with given names.
Examples: NASSIR, GAMAL ABD (AL)
SHARIF, ABD (AL) MOHD
e. TITLES: A descriptive name or appellation which denotes
.rank, office, privilege, or is used as a mark of respect. The
terms Jr., III., 2nd, Mrs., Miss, Colonel, Prince, etc., are
included as titles.
Example: BROWN, JOHN /JR/
(1) In most files dealing with military personalities,
rank is normally fielded separately. If titles are included
in the name field, it is important that they be identified as
such, so that they do not become confused with given names.
Example: SCHEINHEIMER, BARON should be
SCHEINHEIMER, /BARON/
f. TELECODE: Numeric equivalent of ideographs used in
Chinese, Korean, and Japanese writings. Some Japanese ideographs
which have no numeric equivalent are represented phonetically,
i.e., "KATAKANA." When the ideograph is illegible and/or the
numeric equivalent is not known, the term, NTA (No Telecode
Available) is often used.
Examples: TOJIMA, FUSANOSUKE /2073/*02701/2075/0037/6534/
LEE, WON-LOU /NTA/0029/0283/
CHAN, LI-SHU /7115/0173/0209/
(1) Each numeric or alphabetic set in the telecode
should be separated from the other by some special character.
If the telecode is recorded in the name field, special
characters should be used to identify'it for potential special
processing by the computer.
FOR; O.F:FI;CIAL. ,USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
FOR OFFICIAL USE ONLY
Annex 2
Attachment 1
g. PREPARATION OF THE NAME FOR SORTING AND STORAGE:
(1) If characters other than alphabetic are used in
the name, certain special characters should be removed
for sorting purposes, creating a so called "Pure Name"
for sorting purposes. The internal creation of a sort
name is necessary to assure accurate sequencing of
names for alphabetic printing or storage. When the
name is printed, the original input name field is used.
(2) If characters such as hyphen or an apostrophe
were allowed to remain in the name during a sort, the
name HERNANDEZ-PELAGIO would be listed after the name
HERNANDEZ ZERTUCHE. A search for O'BRIEN would find it
listed before names beginning with OA and not in the OB
part of the list as would be expected.
(3) Characters and special elements to be removed
for sort purposes are:
(a) Particles - remove and left
justify the remainder of the name.
(b) Hyphen - remove and insert space.
(c) Period - remove and left justify
the remainder of the name.
(d),Comma - remove and insert an extra
space code.
(4) Titles and telecodes included in the name field
are sorted to numeric and/or alpha order. The virgules
or other special characters enclosing these characters
are also used in sorting and will provide the uniqueness
required to place names embodying titles or telecodes
after like names in the file, without a title or telecode.
(')) Upon the removal and substitution of the foregoing,
the name may be sorted accurately to alphabetic order. Note,
in the following examples, the effect of the foregoing rules,
especially with respect to compound names.
NAME AS PRINTED NAME FOR SORTING
'AZIM, MOHAMED (AL) AZIM MOHA.MED
FOR OFFICIAL-USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
Example: CHIANG, KAI-CHEK /1203/0009/7156/
b. Identify surname elements as opposed to given name
elements, i.e., by placing a comma between the two elements.
Annex 2
Attachment 1
NAME FOR SORTING
AZIM MOHAMED AL
GARCIA MARIA
GARCIA LOPEZ MARIA
GARCIA LOPEZ MARIA
OBRIEN JOHN
OBRIEN JOHN /DR./
SANTOS JOSE
,SMITH J X
SMITH J XAVIER
SMITH ZELAYA
SMITH CORONA JAMES
STE ANTON GREGOR
STE ANTON GREGOR
10. The following, in summary, is the approach the Team recommends
in the identification of name elements, with examples of the types-of
punctuation controls which may be used:
a. Record complete name elements in a consistent order,
i.e., surname followed by given names then by telecodes and/or
titles.
FOR.OFFICIAL'USE ONLY
NAME AS PRINTED
AZIM,? MOHAMED AL
GARCIA, MARIA
GARCIA-LOPEZ, MARIA
GARCIA (Y) LOPEZ, MARIA
O'BRIEN, JOHN
O'BRIEN, JOHN /DR./
(DE) SANTOS, JOSE
SMITH, J. X.
SMITH, J. XAVIER
SMITH, ZELAYA
,SMITH-CORONA, JAMES
STE. ANTON, GREGOR
STE-ANTON, GREGOR
Example : DOE, JOHN
Identify particles, i.e., by placing parenthesis around
them.
Example: GARCIA (y) LOPEZ, JOSE
FOR OFFICIAL -?USJ ONLY-
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
Annex 2
Attachment 1
d. Identify titles and/or telecodes, i.e., by placing
virgules around them.
Examples: CHAN, WON LI /0148/0029/0173/
ROBBINS, CHARLES A. /JR./
e. Identify initials from one character names, i.e., by
terminating them with a period.
Examples: SMITH, J. L. ARMAND
Y, LI CHU (one character surname)
SANCHEZ R., JUAN
f. Allow sufficient space for recording the entire name.
Forty (40) positions minimum are recommended.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2 25X1
SECRET
ANNEX 3
DEFINITION OF ELEMENTS ON THE BIOGRAPHIC INDEX FACTS
1. The index.size refers to the actual number of index records
(3 x 5 cards, IBM cards, logical records on magnetic tape, etc.).
2. The type of index record would include whether it is a 3 x 5
card, 5 x 8 card, IBM card, on magnetic tape (MT) in document form,
etc.
3. The increase per year is the best possible estimate of
the yearly change in the number of the index records during the
next three years.
4. A multiple reference card is one which leads to more than
one dossier, document, etc., by some reference mechanism such as a
number.
5. The emphasis in this definition is on the word "predominately"
with the understanding that probably all indexes being considered are
mixed to some degree. The purpose of this item is to indicate in
general terms whether an index mainly concerns U. S. citizens or
foreign nationals.
6. See Annex 1.
7. A "request" means a requirement levied on the index,
either by the organization internally or by another organization, for
the checking of a name of a person. If the request is in the form
of a list, for example, names of ten different individuals are
considered ten requests.
8. The average number of searches per request indicates how
many different ways on the average a request is searched. The searcher
may look for a variation in the name, for example, E. J. Jones, Ed
Jones, etc., or for the name variant in either the surname or other
name elements (for example Nicholas, Nichols, Nickols, Nickles, etc).
Some organizations may make one or both types of multiple searches
on a certain type or percentage of requests.
9. This is the product of column 7 times column 8.
10. Maintenance searches include such activities?as prechecks
SECRET/
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 25X1 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
SECRET/
ANNEX 3
25X1
for any reason, the filing of new cards, the refiling of cards for
any reason, activity involved in correction of cards, cards being
placed or removed for the purposes of opening new cases, purging
operations and any other index search or look-up which is not made
directly as a result of a normal request as defined under item 7.
11. This is the summation of items 9 and 10. This item
reflects the actual total number of searches performed by the
reporting organization per day.
12. This is the percentage of the requests (item 7) on which
no record or no identifiable information is obtained from a check of
the index. It was recognized by the Team that many possible
identifications made at the index level later result, after final
analysis, in a no record or a no identifiable information; but it
was agreed by the Team that since this figure was not readily
available, the best criterion for the purposes of this report would
be the no record at the index level.
13. This percentage figure represents that proportion of total
requests (item 7) which come from other agencies.
14. This represents the number of requests from other agencies
as calculated from the percentage figure in column 13 times the
request figure in column 7.
15. This percentage figure indicates the portion of requests
from other agencies for which no record is found at the index level.
The same criterion was used as for item 12.
16. This represents the number of external requests on which
column
no record ifigure ntcolumn level.. the
percentage
14.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
.FOR OFFICIAL USE ONLY
ANNEX 5
TILE NAME GROUPING APPROACH{
1.' The Name Grouping approach is designed to insure that a
search of a name brings together all references to an }ndividual
although his name may have been recorded in various spellings and
transliterations. This is accomplished by having linguists
(native speakers) examine the name spellings recorded in a
particular index in order to put names which belong together
phonetically in a group which is then identified by a:`number.
Thus, when the index is searched, references recorded on any
variant of a surname or given name are brought together through
the pre-analysis and grouping by the language expert,
2. The purpose of the technique is to build into a given index
system a one-time, professional linguistic analysis of each unique
name spelling related to other phonetically identical name spellings
on a purely pragmatic basis. That is, name grouping is concerned
with the name spellings actually received by an organization, not
by rules or theories on how names might have been, or ought to
be, spelled. The primary advantage is to avoid a variety of
search criteria by various index clerks.
3. Inherent in this technique is the logic for random access
storage of biographic records in a computer system. The surnames
and given names are used as computer dictionaries (tables) leading
to all group index records on a given name variant in one storage
area of a random access file.
FOR OFFICIAL'USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500026005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
FOR OFFICIAL USE ONLY
2 -
EXAMPLES OF NAME VARIANTS
VARIANT SPELLINGS OCCURING FROM TRANSLITERATION
ARABIC ND L YHM
FAR EAST
MUHI-AL-DIN
MAHJOEDIN
MAHAYIDEEN
MAHYUDDIN
MHIDINE
MOHAYUDDIN
MOHHDIN
MOHIDEEN
MOYIDEEN
MOHIEDDIN
MUHY-AL-DIN
MUHYI-UD-DIN
plus 25 more
J` = Telecode 0491
LIU = Mandarin
LAU = Cantonese
YU = Korean
RYU = Japanese
ANNEX 5
WI?1GE = WOE.GE, WERGE
JANSEN = JAANSEN
NaNEN = NOONEN
IANOZZI = JANOZZI, YANOZZI
SNJDER = SNYDER, SNIDER
MENSKJ = MENSKY, MENSKIY
PETROW = PETROV, PETROF
FELDMAN = FELDMAN, FELTMAN, FELDTMAN
FOR OFFICIAL USE ONLY
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
FOR OFFICIAL USE ONLY
ANNEX 5
EXAMPLES OF SURNAME GROUPS
002914 CHLADEK
HLADEC
HLADIC
HLADIK
HLADK
HLADEK
008687 ABOURGELI
004739
RUJAYLAH
FOGELER
VOGELER
VOGLER
WOEGELER
001712 MATZGER
METZGER
MEZHER
MAETZCHKER
METZKER
MEZGER
002194 SCHUKOW
CHOUKHOV
DIUKOV
DZHUGOV
JOUCOFF
SCHUCHOW
SHUKHOV
YOU KOV
YOUKOVA
ZHJUKOV
ZHUKOV
ZHUKOVA
EXAMPLES OF GIVEN NAME GROUPS
GROUP. NAME
ABRAHAM
BRAHIM
'EBRAHIM
IBRAGIM
JBRAHIM
STEPHAN
STEVAN
STEVEN
ISTVAN
ETIENNE
ESTABAN
STEFAN
STEFA
STEVE
STEVO
STJEPAN
Z00086 EDWARD
EDVARD
EDOARD
EDUARD
EDUART
EDVART
SEE ALSO: ED. GROUP
#Z00002 '
EDWIN`
EDVIN
EDWINS
EDVINE
SEE ALSO:
FOR OFFICIAL USE ONLY
EDW. GROUP
#Z00018
ED. GROUP
#Z00002
EDW. GROUP
#Z00018
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80BO1139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
S-E-C-R-E-T
ANNEX 6
TERMS OF REFERENCE
E__CTIVE
A. OBJ_
To identify means for improving the storage; retrieval and
exchange of information from the major name files and related data
files in the Intelligence Community.
FACT FINDING
1. Identify those significant index and related systems leading
to biographic information collections in the government which are
routinely consulted by intelligence agencies for their security,
counterintelligence or foreign .(positive) intelligence content.
2. Establish the following facts concerning each of the above.
a. Size: Number of index records (i.e., extracts of
information, such as 3 x 5 cards, punched cards, magnetic tape
records, disk records, strip records, etc. normally leading
to documents and files), type and size of index records, single
or multiple reference.
b. Emphasis on types of personalities covered: e.g.,
percentage of foreign vs U. S. citizens, scientists, military
political, Communist Party, Maritime, foreign intelligence
services, agents, etc. This will include the "name finding"
as well as the "name searching" activity.
c. Number of names searched daily: Percentage of positive
and negative responses, depth of search on name variants.
d. Major requesters; proportion of requests from each.
e. Methods of communicating requests and responses:
Forms, memoranda, teletape, transceiver, data phone; security
classification of requests and responses.
f. Identifying data in conjunction with name normally
included'in index-reference.
g. General description of input, maintenance and search
processing.
h. Current requirements for submission of requests.
S-E-C-R-E-T
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
ANNEX 6
i. Classification of the index.
C. REVIEW
1. Examine costs, methodology and prospects for biographic systems
now undergoing mechanization.
2. Identify basic problems to be faced and areas where policy
decisions are required by each agency in planning for mechanization.
3. Identify those areas where format, methodology and equipment
compatibility are required or are highly desirable in name searching
or finding to obtain optimum speed, quality and economy in automating
query and response.
D. RECOMMENDATIONS
Formulate recommendations for CODIB and USIB approval outlining
policy objectives for the Community, with generalized projections of
cost, manpower and time required to meet these objectives. Include
specific guidelines for agencies to follow in systems planning and
development.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
ANNEX 6
TERMS OF REFERENCE
A. OBJECTIVE
To identify means for improving the storage; retrieval and
exchange of information from the major name files and related data
files in the Intelligence Community.
B. FACT FINDING
1. Identify 'those significant index and related systems leading
to biographic information collections in the government which are
routinely consulted by intelligence agencies for their security,
counterintelligence or foreign (positive) intelligence content.
2. Establish the following facts concerning each of the above.
a. Size: Number of index records (i.e., extracts of
information, such as 3 x 5 cards, punched cards, magnetic tape
records, disk records, strip records, etc. normally leading
to documents and files),. type and size of index records, single.
or multiple reference.
b. Emphasis on types of personalities covered: e.g.,
percentage of.foreign vs U. S. citizens, scientists, military
political, Communist Party, Maritime, foreign intelligence
services, agents, etc. This will include the "name finding"
as well as the "name searching" activity.
c. Number of names searched daily: Percentage of positive
and negative responses, depth of search on name variants.
d. Major requesters; proportion of requests from each.
e. Methods of communicating requests and responses:
Forms, memoranda, teletape, transceiver, data phone; security
classification of requests and responses."
f. Identifying data in conjunction with name normally
included in index reference.
g. General description of input, maintenance and search
processing.
h. Current requirements for submission of requests.
S-E-C-R-E-T
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
i. Classification of the index.
C. REVIEW
1. Examine costs, methodology and prospects for biographic systems
now undergoing mechanization.
2. Identify basic. problems to be faced and areas where policy
decisionsare required by each agency in planning for mechanization.
3. Identify those areas where format, methodology and equipment
compatibility are required or are highly desirable in name searching
or finding to obtain optimum speed, quality and economy in automating
query and response.
D. RECOMMENDATIONS
Formulate recommendations for CODIB and USIB approval outlining
policy objectives for the Community, with generalized projections of
cost, manpower and time required to meet these objectives. Include
specific guidelines for agencies to follow in systems planning and
development.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
S-E-C-R-E-T
ANNEX 7
MEMBERS OF CODIB TASK TEAM V - BIOGRAPHICS
25X1 CIA
STATE
Mr. Mitchell Stanley
Mr. Halvor Eckern (Alternate)
ARMY
Mr. Paul Anderson
NAVY
Mr. Marvin E. Van Dera
Mr. William Urick (Alternate)
AIR FORCE
Lt. Col. Edmund M. Manning
Maj. Russell S. Keen (Alternate)
Mr. John L. Keefe
Mr. Earl W. McCoy
SECRET SERVICE
Mr. Frank G. Stoner
CODIB Support Staff
Mr. Pearley G. Buck
CSC
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
 Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2
ANNEX .7
25X1
MEMBERS OF CODIB TASK TEAM V - BIOGRAPHICS
STATE
Mr. Mitchell Stanley
Mr. Halvor Eckern (Alternate)
ARMY
Mr. Paul Anderson
NAVY
Mr. Marvin E. Van Dera
Mr. William Urick (Alternate)
AIR FORCE
Lt. Col. Edmund M. Manning
Maj. Russell S. Keen (Alternate)
Mr. John L. Keefe
Mr. Earl W. McCoy
SECRET SERVICE.
Mr. Frank G. Stoner
Mr. Pearley G. Buck
CODIB Support Staff
S-E-C-R-E-T.
Declassified in Part - Sanitized Copy Approved for Release 2012/01/24: CIA-RDP80B01139A000500020005-2