COMMUNITY ON-LINE INTELLIGENCE SYSTEM (COINS) EXPERIMENT

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP80B01139A000100020007-4
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 26
• Document Creation Date: December 16, 2016
• Document Release Date: July 29, 2005
• Sequence Number: 7
• Publication Date: March 28, 1969
• Content Type: STUDY

'f43~-Y Y.-s ~`'i..Y. 7 'A C~l'' J V 41-( c. Y ! / l I / 1
Approved*- Release 2005/08/03 : CIA-RDP80BO89A000100020007-4
Community On-line Intelligence System (COINS)
Experiment
OBJECTIVES
1. Design, develop and implement a system for the
COINS Experiment which will operate at a Top Secret SI
security level and which will produce the necessary data
to determine the feasibility and permit an evaluation
of the utility of a secure interconnected computer based
information system with remote access capability.
a. How is the achievement of the objective
to be identified and measured?
(1) The first part of the objective has
already been partially achieved; i.e., the system has
been designed, the hardware installed and connected by
means of crypts/protected communication circuits,
and some intelligence information files are available
in some agencies for remote interrogation by users at
all participating agencies. This demonstrates that it
is feasible to exchange information from various computer
based sources. When all elements of the network are
functioning together as planned and users can efficiently
access any COINS file in reasonable time during a reason-
able period of each working day, the__first)part will have
been achieved.
(2) Experience and data relative to utility
cannot be acquired unless the system is used by valid
25X1
sers in
few
encies
s
bett
to
eet d
t
-
u
a
ag
a
a
er way
m
ay
or
25X1
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100021007-4
 Approved I + Release 2005/08/03: CIA-R?P80B019A000100020007-4
day tasks.
(3) A data collection plan has been
formulated to gather statistics on system utilization
and performance, and user opinions and attitudes towards
COINS.
(4) The feasibility of the system is very
dependent upon its ability to operate securely and to
demonstrate that inadvertent violations of need to know
can be limited to an acceptable level. A COINS Security
Officer has been appointed to whom all security matters,
including violations of need to know, will be referred.
b., What constitutes success or failure of this
objective?
(1) Success or failure of the objective
will depend first upon the technical capability of the
system to operate securely. Since such interconnection among
computers is being accomplished through communications
lines and a computer based communications switch, and no
serious breaches of security have occu this at least
denotes a measure of success in the employment
of hardware.
(2) The experiment will prove successful
if it evokes a favorable reaction among a crossection of
users in the participating agencies. This favorable
reaction need not necessarily be in terms that COINS
provided useful information in support of analysts' work,
Approved For Release 2005/08/03: CIA-RDROOB01139A000100020007-4
 Approved Release 2 0 08ib3 CIA-RDP80B01 9A000100020007-4
but it could include judgements that an efficient COINS-
like system containing more suitable data bases would
have high utility in support of the intelligence analytical
function.
(3) The experiment will prove a failure if
a sufficient number of users do not react favorably to their
exposure to COINS. Adverse reaction could result from a
variety of causes either singly or in combination. For
example:
(a) Inefficient operation of the net-
work or excessive down time.
(b) Differences in interrogation
methods and lack of data standards.
(c) Poor orientation and training.
(d) Lack of user interest in COINS
(e) Excessive interrogation response
(4) The number of interrogations and
responses can be tallied and are quantitative reflections
of both system activity and the users' participation.
Less easy to evaluate are the qualitative judgements
of the users as to the effectiveness of the system, the
pertinence of the data available, or the use which is
ma.de. of the data. Presumably, after a period of time,
increased outputs either in the form of finished reports,
Approved For Release ?L,O05108IQ3,:, QIA7RD,P80BO1139A000100020007-4
 Approved Release 21610 -ARBP80'30'A19A000100020007-4
improved files, and/or enhanced validity of data elements
could be counted and attributed to the system. However,
the basis for a determination of this kind has not been
established. Nor is it likely that such definition could
or should be attempted until the system has operated
for a sustained period of time. The system may have
considerable impact on traditional intelligence production
methodology regardless of the efficiency of the system
in its early stages. But for this phenomenon to be obvious
a considerable history of completed interrogations will
have to be accumulated.
(5) An excessive occurrence of accidental
spillage and violations of need to know will constitute
prima facie evidence of a failure to achieve the objective.
Approved For Release (2005/0$/03.: CIA-RDP80B01139A000100020007-4
 Approved.- Release f08/b3' ~IA R'8d~a0,a9A000100020007-4
2. To isolate, define and determine solutions to the
technical problems incident to designing, implementing
and maintaining a secure interagency computer based,
on-line intelligence information retrieval system.
a. How is achievement of the objective to be
identified and measured?
(1) Routine, periodic reports from the COINS
Test and Analysis Panel will be used to reflect statistics
by lane on (a) average response time, (b) percentage of
successful interrogations, and (c) percentage of time
lane active during scheduled periods. In addition, reports
from users of spillage, improper response, or incomplete
answers will be used to identify and measure the achievement
of this objective.
(2) Isolation, definition, and resolution
of technical problems is a continual task engaging all
COINS participants and resulting in continual system and
software changes to improve network operating efficiency
and insure security. It is recognized that some problems
cannot be resolved in the immediate future (e.g., a common
network language, or interrogation language procedure) but
long term solutions are being sought by COINS participants,
panels and contractors.
(3) An accounting system to positively
identify system operational traffic loss must be developed.
Approved For Release 200,~/g8/,03, :,.CIA-RDP80B01139A000100020007-4
 Approvedfimpr Release12Oo51081a3': tIa `:- P8?0B0VE9A000100020007-4
b.- What constitutes success or failure of this
objective.
(1) This objective has largely been achieved.
Full achievement will be realized when all participating
agency computer systems operate simultaneously in the
COINS network during at least 75 percent of a scheduled
period each day.
(2) If, by the end of the experiment, at
least 75 percent of the queries levied by a COINS parti-
cipant on all other participants are correctly answered
in a reasonable time frame without accidental spillage,
and all messages can be accounted for then the technical/
security aspects of the Experiment will have been successfully
tested.
Approved For Release 21195108/03 C4A,-RDP80801139A000100020007-4
;;.. 9 a
 r .. .! ;sg Fjj
Approved ,@r- Release ti5I08103 CIA-k15P80B0W9A000100020007-4
3. To provide practical training and actual experience
for intelligence analysts in the use of an interagency
on-line intelligence retrieval system as a better way
to meet day to day tasks.
a. How is the achievement of the objective to
(1) Each COINS participating agency has
conducted, is conducting, and will conduct training
sessions for potential COINS users. Each agency has
produced written training materials describing how to
use various portions of the COINS network. The COINS
Environment Report, which the Test and Analysis Panel
has asked each COINS participant to prepare, will detail
the measures taken by each agency to interest analysts
in the use of COINS and the level and detail of training
provided.
(2) Periodic reports from the COINS Train-
ing Panel will provide information on training activities,
status and progress.
(3) In depth interviews of users which are
planned by the Test and Analysis Panel will provide insight
into the degree of training provided, experience gained
and additional training required.
b. What constitutes success or failure in the
Approved For Release 20Q5108i04 :,CIA-RD,P,B.OBC,139A000100020007-4
 Approved RQp-,Release 2005108103`: C1A-RDP80B0' A000100020007-4
achievement of the objective?
The objective will be successfully achieved
if a representative group of analysts in each COINS
participating agency are able to query COINS files with
a minimum of difficulty. In addition, those who are
so trained must exercise the system sufficiently to
gain a level of experience which would permit their
evaluation of system utility.
Approved For Releasee'2005108/03 : CIA-RDP80B01139A000100020007-4
 Approved FQp Release 20$J0`81O,3-_CiA-RDP$,OBQ A000100020007-4
4. To ascertain the essential characteristics of
files which make them suitable for inclusion in an inter-
agency on-line intelligence retrieval system including
user requirements for timely, accurate and complete
information.
a. How is the achievement of the objective to
be identified and measured?
(1) The COINS switch log will maintain data
on file usage, and, in addition, each analyst who queries
a COINS file will provide file usage satisfaction information
via the User Log Form. The in depth interviews which each
COINS participating agency is required to conduct to
obtain reactions from COINS users should be particularly
fruitful for obtaining data on user file preferences. In
addition, the IHC has requested that each of its members
provide data on the types of files recommended for inclusion
in an operational COINS.
(2) The proof of suitability of a file for
inclusion in an interagency system will be largely, but not
entirely, determined by the rate of usage by the agencies
other than the file sponsor. Files which are not used
frequently would not be very suitable unless a very short
response time is required whenever it is queried. Files
used only by the sponsor would not be suitable for an inter-
agency system. On the other hand, files which get frequent
interagency use but wherein most of the requirements could
be satisfied by overnight or 1-2 day service may not be
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
 Approved P'Release` ?65/O8/03-`: "CIA=RDlSOB01 A00010002000'7-4
very suitable for an on-line interagency system. The
question is whether, after determining by experience and
interviews of users that a given file is suitable for an
on-line interagency system, its characteristics can be
identified and assigned meaningful values against which
to measure the suitability of proposed new files without
resorting to actual on-line retrieval experience with the
new files. If possible, an attempt should be made about
midway of the T&A period to identify those files that
appear to be suitable for interagency use and to evaluate the
characteristics which determine that suitability. Some
of the factors to be examined are: subject and area
coverage, data elements included; size of each record;-
date span; update frequency; types of service materials;
number of users by agency (potential users for a proposed
file); response time required by users; alternative methods
of data exchange.
b. What constitutes success or failure in the
achievement of the objective?
(1) This objective will be successfully met
if'a set of files can be identified which would have high
utility in an interagency on-line system. This objective
would not be met if the results of the Experiment were
inconclusive regarding file usage and preferences.
Approved for Release ~-00510$/Q3 CIA-RDP80,B01139A000100020007-4
 Approved Release 2 Q /08/0 -CIA-RDP80B01 A000100020007-4
(2) Experience with other on-line information
retrieval systems has shown that with an increase in system
efficiency and experience gained by users increased confi-
dence is realized and more specific requirements come into
focus. If the mechanism provided by COINS management for
data base modification (COINS Operational Procedure No.
1-1968, 22 November 1968) is utilized a measure of success
will be realized.
Approved For Release 2005/08/03 ; CIA-RDP80B01139A000100020007-4
 Approved FjRelease 65L0.8/b` ::CIAAR, ?P.,8QB01 A000100020007-4
5. To determine those elements of the total system
design which encourage and facilitate analyst usage and
those elements which discourage or hinder usage.
a. How is the achievement of the objective to
be identified and measured?
(1) COINS system designers are already aware
of certain unavoidable problems implicit in the COINS
system which will inhibit usage, e.g., multi-query languages,
lack of data standards in COINS files, etc. The seriousness
of these problems and those aspects of COINS which are
viewed in a more favorable light will be fully brought
out by the reactions obtained from COINS users through
log forms and interviews.
(2) Each of the following factors must be
examined to determine the threshold between encouraging
and discouraging analyst use. For example; analysts
might be encouraged to use a system that was 90 percent
reliable but would be discouraged from using a system
that-was less reliable.
(a) Solution of data bases; number, type,
location and ease of operation of remote terminals.
(b) Query languages: one vs. many, abbreviations/
mnemonics vs. plain language; query procedures; preprogram-
med queries vs. on-line query programming.
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
 Approved W Releasd`i0'05(08/O3': CIA=RD1b80B019A000100020007-4
(c) Response time
(d) System reliability
(e) System availability (days and hours)
(f) Usability and readability of outputs
(g) Coded vs. plain language output
(h) Type and amount of training (and
motivating) of users
(i) On-line vs. off-line file update
procedures
(j) Interactive vs. non-interactive
systems
(k) Time sharing vs. batch processing
(1) Type and quality of analyst aids
(handbooks, on-line file and query descriptions, etc.)
b. What constitutes success or failure in the
achievement of the objective?
This objective will be successfully met if
each agency is conscientious in its collection of data on
user reactions to COINS. Low user paritcipation in COINS
could result in inadequate data collection.
Approved For Release 2005/08/03 : CIA-RDP80B0l139A000100020007-4
 ? Approved FQ,Release 2b05/08/03 CIA=RDP;0 01U,9A000100020007-4
6. To collect the necessary data on file usage,
system efficiency and reliability and user reaction to
permit an evaluation of the feasibility and utility of
operational implementation of a follow-on COINS-like
system. t
a. How is the achievement of this objective to
be identified and-measured?-
The Test and Analysis Panel has developed a
plan, procedures, and forms for collecting the desired data
and this plan has been approved by all COINS participants.
However, it should be constantly reviewed (monthly at
least) during the T&A period to see if additions, deletions,
or modifications are necessary.
b. What constitutes success or failure in the
achievement of the objective?
If the data collection plan is followed as
recommended,, and if there is sufficient usage of the net-
work to obtain the desired user reaction then this objective
will be partially or fully met. Partial failure could
result from gaps in the data collection plan which were
not anticipated during the planning stage.
Approved For Release 200510$/03 CIA-RD.P80B01139A000100020007-4
F i _ a
 1 !3'fl 1 i r .,~L
Approved F Releas ebb&68/03'% CCi4=13P>'0B01,A000100020007-4
7. Determine the elements and attributes which would
be critical to the success of a full system for continuing
community operations, and develop objectives and specifi-
cations for an operational system.
a.--How is the achievement of this objective to
If objectives 1-6 above are met, the data
will be available to develop a system plan for an operational
COINS network. In other words, this objective is unique
in that it is completely dependent on the satisfaction of
all. the other objectives. Development of objectives and
specifications for an operational system will be the
culmination of the total experience of the COINS Experiment.
That experience, properl,evaluated, will dictate the
objectives and specifications. The problem is to properly
evaluate the experience gained by the Experiment.
b. What constitutes success or failure in the
-achievement of this objective?
This objective could fail on two counts:
(1) It will fail if inadequate or inconclusive
data is collected in an area critical to the design of an
operational system.
(2) It will fail if the data collected during
the operational test conclusively proves that an operational
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
 Approved Fqoelease 2005/08/03 : CIA-RDP80B01 gA000100020007-4
COINS network is not required at this time.
If neither of the above conditions exist, this
objective can be successfully satisfied.
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
 Approved F%Release 2005/08/03: CIA-RDP80B01 QAA000100020007-4
8. To gain insights into COINS capability to (a)
improve utilization of resources of the participating
agencies by the centralization of selected information
files at one agency for access by all. (b) enhance the
value of intelligence information data bases.
a. How is achievement of the objective to be
identified and measured?
Bilateral efforts are currently underway between
DIA and NSA to establish two such intelligence information
files; a file on elite and other key Soviet military
personalities, and a file containing information on
Soviet airfields. Conclusion of a Technical Memorandum
of Agreement on these files and implementation of
procedures specified therein will signal partial achieve-
ment of the objective. The number of interrogations and
responses can be tallied and are quantitative reflections
of user access. In depth interviews of users to be
conducted under the auspices of the Test and Analysis
Panel will provide a measure of file utility.
b. What constitutes success or failure of the
objective.
If the two files mentioned previously are success-
fully implemented, if maintenance procedures specified
in the technical memorandum of agreement are strictly
adhered to, and if the files receive frequent use and
Approved For Release 2005/08/03 ; CIA-RDP80B01139A000100020007-4
 Approved F ,,Release 20 6/08/03` 'IA'-RDFg8OBO14Ji4000100020007-4
general user satisfaction is realized the objective can
be said to be successful.
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
,d J
 Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
USERS LOG FOR TEST OF COINS "EXPERIMENT" PAGE OF
TEST NO.
FROM TO
(DATE) (DATE)
DATE TO FM 'FILE TIME TIME TIME RECEIPT
(YYMMDD) ORGN ORGN INTG IN SWITCH RESULTS' NUMBER
RCPT
TYPE OF RESULTS REMARKS
(ANSR,ABRT,NONE)
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
 Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
ANALYSIS OF NETWORK ACTIVITY IN COINS "EXPERIMENT"
TEST NO.
FROM TO
(DATE) (DATE)
REPORTING ORGANIZATION: AAA
DATE
(YYMMDD)
INITIAL
TIME UP
TIME
DOWN
TIME
UP
Rs-
FINAL
TIME DOWN
REMARKS
690512
0810
0820
0830
compu
System went down and recovered
0845
0855
commo
Line to switch out
0958
690513
0805
0830
0845
compu
0910
0820
commo
0940
0950
compu
0955
UAIU
 Approved F AWase 2005/08/03: CIA-RDP80B01' 100020007-4
COMMUNITY ON-LINE INTELLIGENCE SYSTEM
COINS
OPERATIONAL PROCEDURES (FIRST DRAFT)
Number ?-1969
NEW COINS PARTICIPANTS
30 April 1969
Section
Authority . . . . . . . . . . . . . . I
Purpose . . . . . . . . . . . . . . . I I
Scope . . . . . . . . . . . . . . . III
Background . . . . . . . . . . . . . IV
Types of Participants . . . . . . . . V
Procedure for Submitting and . . . . VI
Handling Requests . . . . . . . . .
Criteria for Acceptance . . . . . VII
1)
References: (a) DIA Letter, C-15,117/MS-1, Subject:
"Release of COINS Documentation and Files
to Unified and Specified Commands," dated
27 August 1968.
(b) NSA Memorandum, Serial: COINS/161, Subject:
"same as above," dated 16 December 1968.
I - Authority
1. USIB-D-39.1/9, "Community On-line Intelligence
System," 11 April 1968.
2. USIB-D-71.4/1, "Management of the COINS Experiment,"
3 June 1968, paragraph 4k, page 3.
3. COINS Master Plan Part I FY 1969 - 1970, dated 2
December 1968. Part III, Para 3h, page 21.
Approved For Release 2Q05f08/03I~-:, CIA-RDP38 B01139A000160020007-4
 i?.. 1 9 E.. ~r n 17
Approved Fc t se 2001/O8IU3 CIA=F4DFj$;13011 9M100020007-4
II - Purpose
These operational instructions set forth procedures
and criteria by which new organizations can be connected
into the COINS network with either a remote terminal or
central computer system.
III - Scope
Instructions in this operating procedure applies
only to the operations of the COINS network between parti-
cipating organizations. It does not preclude a participa-
ting organization from connecting other non-particpating
organizations to its computer system for non-COINS activity,
providing specified security requirements are satisfied.
IV - Background
1. On 27 August 1968, DIA requested authority to (a)
release COINS documentation to the Unified and Specified
(U & S) Commands and (b) permit them to have on-line access
to the COINS network through the store-and-forward switch
at DIA. (See reference a). In addition this letter set
forth a tentative schedule for the connection of the U & S
Commands to the DIA switch.
2. After consultation with the COINS sub-system managers
DIA. was informed to proceed with their plans to bring the
U & S Commands into the DIA system on the dates indicated.
Access to the full COINS network will be decided on a case-
by-case basis, after a period of successful operations with
the DIA systems (See reference b).
Approved For Release 200?/08/43,2 .CIA:R?P;8UBQ1139A000100020007-4
 Approved ease 2005/08/03 CIA-RDP80B01Adt0100020007-4
V - Types of Participants
There are at the present time two types of participants
in the COINS program, namely:
1. Contributors - A contributor is a participating
organization which has a computer system and files which it
is making available to the other participating organizations.
In the present COINS Experiment the contributors are DIA,
Hq CIA, CIA/NPIC and NSA.
2. Subscribers - A subscriber is a participating organ-
ization which does not have a computer systems in the network,
but does have access to the network through a remote terminal
connected to another contributor's computer systems. A
subscriber may or may not be sponsoring a file stored in a
contributor's system. In the present COINS Experiment the
subscribers are the Department of State and National Indica-
tions Center (NIC).
VI - Procedures for Submitting and Handling Request
1. Requests to connect into the COINS Experiment must
be forwarded in writing to the Director, National Security
Agency, Executive Agent for the COINS Experiment. Each
request will be handled on a case-by-case basis. At a
minimum each request must contain the following information:
a. Name and location of organization
b. Requirement for participation
c. Type of participant (i.e., subscriber or contributor)
Approved For Release 200kjQ$1Q3 :yQI,A FDP80?91139A000100020007-4
 Approved Foftekase 2005/08/03: CIA-RDP80B011%VW100020007-4
d. Date expected. to be connected into the COINS
rork
e. Statement on the willingness to accept the
costs to connect into the network, as well as current
operating procedures.
f. Statement certifying that the current security
procedures and constraints can be satisfied
g. The files to be made available in COINS by
the requesting organization.
2. Military services, except cryptologic agencies
(i.e., ASA, AFSS and NSG) and the Unified and Specified
Commands, must submit their requests through DIA. The
Cryptologic services must submit their requests through
NSA.
3. Upon receipt of a request the COINS Project Manager
will submit, copies of the request to all sub-system managers
for consideration. Sub-system managers will respond within
30 days, indicating whether or not their organizations concur
or non-concur in the request. Non-concurrence must be
accompanied, with a statement of the reasons for non-concurrence.
4. Requesting organizations will be notified within
60 days after their requests are received by NSA as to
whether or not they have been accepted as a COINS participant.
VII - Criteria for Acceptance
Acceptance will depend on the following points:
Approved For Release 2005108/9 -:~ A-I4bP8O8 1139A000100020007-4
 t r u'r y
Approved Fa , eiaase 2005/08/03: CIA-RDP80BO11 Q 100020007-4
1. The requirement of a:
a. Requesting organization to have access to
shared data bases in the COINS Experiment.
b. COINS participant to have access to data bases
in the requesting organization's system.
2. The willingness of the requesting organization
pay the costs to connect into the network and adapt to
existing COINS operating procedures and conventions.
3. The
organization
constraints.
their remote
must operate
in the COINS
willingness
and ability of the requesting
to operate within the prescribed security
Currently all. computer systems including
terminal and communications lines and facilities
at the TOP SECRET SI level while participating
network.
Approved For Release 2005108/Q3 : c'1A=Rb60B01139A000100020007-4
 Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4
Next 1 Page(s) In Document Exempt
Approved For Release 2005/08/03 : CIA-RDP80B01139A000100020007-4