PROPOSAL FOR CENTRALIZED COMMUNITY BIBLIOGAPHIC AND DOCUMENT RETRIEVAL SYSTEM

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP83T00573R000100120022-7
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 9
• Document Creation Date: December 12, 2016
• Document Release Date: October 2, 2001
• Sequence Number: 22
• Publication Date: March 23, 1979
• Content Type: MF

Approved For Iea M $ : i P83T0057 00010012002
ODP-9-480
SAF-E059-79
23 March 1979
MEMORANDUM FOR : Associate Deputy Director of Administration
FROM 25X1 A
Director, consolidated SAFE Project Office/ODP
SUBJECT : Proposal for Centralized Community Bibliographic
and Document Retrieval System
1. As you requested on 28 February, I have reviewed the
proposal made to the Chairman of the Intelligence Handling
Committee to provide the subject service to the community. In
25X1A particular, I asked to review the security implica-
tions of the service as proposed.
2. As noted in the attached memorandum, the accessing of
the RECON data base by both the community and internal Agency
users presents security problems. It appears that some of the
data in this data base requires controlled access and must be
screened out for outside users. If open access is given by the
COINS network to this file, then about 164,000 records would
have to be eliminated from the file. This would mean that
internal users of that file would not have access to the com-
plete file. Alternatively, external requests could be screened
by a human intermediary such that the same files could be used
for all users. Given the small number of external queries pro-
jected, this would appear to be the preferred method of operation.
3. Given the security considerations attached and the un-
desirable cost of duplicating the service, it would appear that
an internal service for free access by the Agency, but providing
screened access to external users, would be the least expensive
alternative. It should also be capable of sustaining the ex-
ternal load as projected.
4. These bibliographic files are indeed the files to be
used on the SAFE system and hence any extension of current
capabilities should be coordinated with the SAFE program to
avoid developing this capability twice. If the IH C initiates
a study on this subject, we would be happy to participate.
Approved For Release 2 Oq / ~~Q` T00573R0db~#0 Ze7omes UNCLASS'.F'-.,
when separated from atra:nment.
 Approved ForW6teleas6 "' FN~F- tU1P83T00570001 00120022-7
SUBJECT: Proposal for Centralized Community Bibliographic
and Document Retrieval System
5. I trust that either you or Mr. Eisenbeiss is still
the representative for the IHC. I would be glad to discuss
this further if you would like.
25X1A
cc: Acting Director of ODP
Director of OCR
Approved For Relea A /4/ 8 T4 LP83T00573R000100120022-7
 Approved For44elease 2002/01/08 : CIA-RDP83T0057 00100120022-7
ODP-9-428
SAF-E046-79
14 March 1979
25X1A
MEMORANDUM FOR: Director, Consolidated SAFE Project Office/ODP
FROM
SAFE Security Advisor
SUBJECT Proposal for Community Access to the
Bibliographic Files Supported by RECON
1. Before addressing the specific questions from
D/ODP on the buckslip attached to ODP-8-2184, 25 January 1979,
I read the memo to gain necessary background. It presents
the subject proposal to the DCI Intelligence Information
Handling Committee and develops three options for implementing
the desired service. These are: 3a. Off-Line Service, 3b.
Direct On-Line Service and 3c. On-Line Service through
Intermediaries. The last, 3c. is impressive for the following
reasons: (U)
a. Provides a high degree of security without system
complications. (U)
b. Is economical because it places the "filters" on
the side of RECON where there is the least activity
and because it does not require duplication of
hardware or data. (U)
c. Is flexible and provides a smooth transition for
the non-CIA users when RECON-supported bibliographic
files become the Central Index File for SAFE. (U)
d. Provides a customized service for the diverse
requestors by knowledgeable "librarians" so that
much training and unwanted output can be eliminated.
Because only 10 queries per day are reported now
with an eventual growth projection to 50 per day
later on, this method should remain economically
feasible and responsive. (U)
2. Preparing to answer the buckslip questions (copy
attached), which are primarily associated with the 3b.
Direct On-Line option, I discussed the matter with OCR/SAS,
OCR/ISG/SAIO, OS/ISSG and various CSPO personnel to gather
ORIGINAL CL BY (a 7 (~t~ `f
DECL ^ REVV'd
Approved For Release 2002/01/08 : CIA-RDP8 f Ag~ OD 7~-`~
6 i i'u 3 S L.r ,~ 1 1A REASON
 Approved For'Iease 2002/01 0'8` tI1 2RbP83T0057 00100120022-7
information and statistics. All were most helpful. From a
security viewpoint, this option is the least desirable.
However, certain general minimum security requirements can
be identified. More might be necessary later on based on
further details of the plan. (A-IOU)
a. The RECON terminals and printers at the non-CIA
facilities would have to be dedicated and in areas
meeting the requirements of tzh Combined Minimum
Security Standards for Compartmented Information.
These standards include physical, personnel and
-technical security. The link to RECON would have
to be via an approved encrypted communication
channel. (U)
b. A "duplicate" data base of some sort will be
needed as the least problematic way to serve the
CIA users. A number of the ways you have suggested
to provide parallel or serial updating of the
bases sound promising. Because there are so many
more CIA users even now - about 300+ generating
nearly three times as many queries as non-CIA - it
is imperative to provide them direct access to a
data base without "security gateways" (if such
were available) or any other impedance to good
service. As stated above, the outside use for
which the dedicated host is suggested will start
with an average of ten queries per day with a
projected growth to fifty per day. (U)
c. Another reason for a "duplicate" data base (and
the reason for the quotation marks) is to retain
present day dissemination restrictions. In the
current off-line method of handling outside users,
the OCR personnel screen out some requests or
limit the response based on dissemination controls
imposed on certain files by the data owners. A
comprehensive, double layer set of codes is
associated with the records for this purpose.
About 164,000 records would currently be limited
or screened out from an NFIB requestor. For an
off-line or human intermediary system, spillage
would present no serious problem. But for an
interactive system for non-CIA users, a data base
should be provided devoid of the restricted items
so that the software does not have to be trusted
to provide the control. (C)
Approved For Releas 2y0d1: q& RQP83T00573R000100120022-7
 Approved For#MIea iIl hJi3t dIi DP83T0057 D00100120022-7
d. Depending on the file update method chosen for the
two versions of bibliographic files, provisions
would have to be made in the SAFE design to securely
accommodate the communications for update. A
twisted pair in the secure distribution grid could
later be replaced by a dedicated channel on the
WBCS when SAFE makes this available. (U)
e. The transition from the current RECON system to
the Central Index File of SAFE would also have to
be addressed in the SAFE plan. Would it be
possible, for instance, to install the "internal"
version of the RECON-supported files directly in
the SAFE Center-to-be? Or must a temporary secure
home for it be provided which would be phased out
later? The other transition would be in terms of
the growth in internal users from the 300 mentioned
above to the total SAFE user population. (U)
f. If any new kind of access is approved for non-CIA
users, thisc-iange would have to be submitted to
the data owners involved in the RECON-supported
files for their concurrence. Their willingness to
provide certain sensitive files to the present
system might have been based on trust in the
dissemination controls built into a human intermediary
system. DDO would be especially sensitive to this.
(C)
3. All questions on the buckslip have been addressed
in the above security requirements. In addition, other
restraints are mentioned which go beyond the questions but
which are pertinent. The guidance does not rely on security
technology still developmental because IOC for RECON community
use could be about a year after approval. Incidentally, the
choice of 3a. (the slight enhancement of the current method)
or the 3c. On-Line Human Intermediary options would allow
easier implementation, although the remote terminals of 3c.
would also have the requirements of Paragraph 2a above. (U)
25X1A
Attachment: buckslip
Approved For F,e,#ere?fs.1~d,~lA-RDP83T00573R000100120022-7
 Approved For '20Dl?A 1N8 : Al-RDP83T0057 +b00100120022-7
Distribution:
D/CSPO/ODP
C/ISSG
Al P.
Chrono
ODP Registry
Approved For Relea I2b
 2 8 FEB 1979
MEMORANDUM FOR:
Approved For Release 2002/01/08 : CIA-RDP83T00573R000100120022-7
 Approved For Release 2002/01/08 : CIA-RDP83T00573R00010012002
ODP-9--435
15 MAR 1979
"1 MORA-1DUM FOR: Chief, Information Systems Security
Group, Office of Security
FROM ~ 25X1 A
Chief, Management Staff, 01W
SUBJECT Proposal for a Centralized community
Bibliographic and Document Retrieval
System Operated by CIA
1. During our meeting on 12 march with the Director,
of Data Processing, we briefly discussed a proposal that
the Intelligence Information {andling Corn ittee study
the feasibility and desirability of adopting CIA's '?FCO11
bibliographic index and ADSPAR micrographic document
storage and retrieval system as a Centralized Intelligcric^
Community Bibliographic and Document Retrieval System,
managed and operated for the Community by CIA. A copy
of the proposal, that was made by the CIA ":er, I'C -
Clifford D. ,*ay, Jr., is attached for your information.
2. Inasmuch as considerable interest in the pro-
posal is being expressed by IUC Committee nember. s , this
matter is called to your attention because of the possible
curity implications.
25X1A
Attachment
ODP-8-2184
cc : SO/0011 - X7/o Attachment
NEMOX
DISTRIBUTION:
Original and 1 - Addressee
2 - O/D/ODP ) t m-r
1 - MS Chrono
25X1A ODp 2 - ODP Registry
~P$Po M4AP CIA-RDP83T00573R000100120022-7
 Al NO1 WILL C-)4rCRt`CLA5.lfflVCA%"VuN YOf ANO fgf",^'Y'r,xe
? OFFICIAL ROUTING SLIP
NAME ANO AOQI USS'
DAT'L
INIrfAL3
II
ICDA
Z
6
ACTION
AIRECT REPLY
PREPARE
REPLY
APPROYAL
DISPATCH
RECOMMENOATIOH
COMMENT
FILE
RETURN
ONCURRFNCE
_
NFORMATION
SIGNATURE
Remarks:
`/"6
FOLD MERC TO RETUf1N TO 8LNDER
oved Fir'~I~s~~2(~21~1+~?Ss?~Ql-Ar~F~$~00 -
r
7