1985 ACCOMPLISHMENTS OFFICE OF RESEARCH AND DEVELOPMENT DIRECTORATE OF SCIENCE AND TECHNOLOGY

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP88G01332R000100070015-8
• Release Decision: RIPPUB
• Original Classification: S
• Document Page Count: 71
• Document Creation Date: December 23, 2016
• Document Release Date: May 5, 2011
• Sequence Number: 15
• Content Type: REPORT

Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
TRANSMITTAL SLIP
1 71.Z.2 a%
TLIU/9-
ROOM NO.
7.4 ol- 54
BUILDING
REMARKS:
?,1)- --L A
11/j-r
.
" im
-eJA
.......-
,11 I?f",4 -; ? :
3 1_71-t
aro
1----
FROM:
ORD/gUL
ROOM NO.
1225
BUILDING
Ames
EXTENSION
FORM NO. ? REPLACES FORM 36-8 (47)
L
1 FEB 56 4 I WHICH MAY BE USED.
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
? ? ?
S.
.1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
f.0
?42?--
u u u ,
c r`, 4- 61. I
3, 033 031 029 0 02
."'
qft---ti_ANZNYANAf I \11116
\,)114 t1W,
??? k?4;,\ - _
trA
1985
Accomplishments
SECRET
? - . ;:i fi
...0.;. :
....:_t'
e' 7-----:-
ir:11? N
'
4041-41?? 1 .4. - - ?* ?
'iit
\
? ? \1
I 1'
a
Tr" --- Office of Research and Development
Directorate of Science and Technology
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
413
STAT
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNCLASSIFIED
INTRODUCTION
The Office of Research and Development (ORD) is responsible for research and development of
technology that will improve the ability of the CIA to collect, process, analyze, and produce foreign intelligence.
In this document ORD presents brief descriptions of some of its significant accomplishments during Fiscal
Year 1985.
This document is structured along functional lines; each section is introduced with a listing of the
organization's principal programs, a description of its mission, and a list of its staff and their projects. Specific
inquiries regarding each project should be addressed to the key person indicated.
This document is the fifth of these annual publications; each edition is published shortly after the close
of the fiscal year. Comments and suggestions regarding content, format, and distribution for future
editions should be directed to the editor or to the Director of Research and Development.
(A BYEMAN appendix to this document is also available.)
UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
6 q t AM
? ?.;
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
T
C35 .733 1731 32 9
. . , . . '
WI WAWA"
\ ik\ I -I'l "a CAI V'k Are aCj.
t......f0V ?
_ AtChr.
1985
Accomplishments
4-
Office of Research and Development
Directorate of Science and Technology
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Office of Research & Development
DIRECTOR
DEPUTY DIRECTOR
Collection Technology
Group
Advanced Concepts
Staff
Security
& Materials
Research Division
Applied Physics
Research Division
Communications
Research Division
I
1
Processing & Analysis
Technology Group
SECRET
Analytic
Methodology
Research Division
Information Systems
Research Division
I
Processing &
Exploitation
Research Division
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
"
N oklatglekuli
vks,11_1011,Nttilingi-3
410 /gra tip-,
1985
Accomplishments
Advanced Concepts Staff
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
SECRET
ADVANCED CONCEPTS STAFF
PROGRAM AREAS
ADVANCED CONCEPTS
Novel Communication Concepts
Modeling Concepts
Political/Economic Studies
Novel Materials
Audio Surveillance Concepts
New Communications Collection Techniques
Parallel Processors
Data Base Research
SECRET
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNLLAJJ11-1tU
ADVANCED CONCEPTS STAFF
INTRODUCTION
The primary mission of the Advanced Concepts Staff (ACS) is to investigate new problems facing the
Intelligence Community, to study the application of emerging technology to these problems, and to foster the
advance of embryonic technologies to a status where they may be exploited for intelligence purposes.
Personnel in the Advanced Concepts Staff are from ORD, other Agency components, and the
Intelligence Community, industry, and academia. These individuals are mature, respected researchers who are
not bound by the formal "requirements process" in their pursuit of new ideas. The blend of diverse, motivated
experts contributes to the overall innovation of the Office of Research and Development.
UNCLASSIFIED AC-1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 5 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
J C l.- IN C I
Research
Center
ORD IR&D
Manager
w r?t
OD&E OTS OSO ORD Other
Appropriate
Offices
Evaluation
ORD IR&D
Manager
DOD Service
Managers
CY-85 Review Activity of 43 IR&D Centers
DS&T Distribution & Review Process
SECRET
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
AC-9
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
?::
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
1985
Accomplishments
Applied Physics Research Division
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
SECRET
APPLIED PHYSICS RESEARCH DIVISION
PROGRAM AREAS
MICROWAVE COMMUNICATIONS
Advanced Microwave and Millimeter Components
SIGNAL PROCESSING
Microprocessors
Signal Processing Chips
Signal Processing Architectures
ELECTRO-OPTICAL SENSORS
DATA STORAGE
Optical Disk Recorders
Optical Tape Recorders
SOLAR CELLS
Laser Gyroscopes
SECRET
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05 : CIA-RDP88G01332R000100070015-8
COLLECTION TECHNOLOGY GROUP
APPLIED PHYSICS RESEARCH DIVISION
INTRODUCTION
The Applied Physics Research Division (APRD) has been chartered to develop technology applicable to
future collection systems to be developed primarily by OD&E. The Division's efforts are grouped under two
broad areas: optics/electro-optics and electronics. Included in these areas are projects on fiber optics gryos;
solar cells; optical disk and tape recorders; advanced microwave and millimeter circuits; and high-speed, low-
power integrated circuits. These projects have been designed to support the technology needs of both Agency
and national level programs.
SECRET AP-1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 2 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
r, I
u
? r
? 025 ?)3 031 029
.er
4'24.14194
-
votirr: T \ "A\
Art r
1985
Accomplishments
Communications Research Division
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
SECRET
COMMUNICATIONS RESEARCH DIVISION
PROGRAM AREAS
TECHNICAL OBSERVABLES
Millimeter Wave Technology
SECRET
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 6 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
1985
Accomplishments
4'.
Security and Materials Research Division
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNCLASSIFIED
Rhf
Dial
Dia2
Skew 1
Skew2
Intercept
RESISTANCE. ohms
AC COMPLEX IMPEDANCE ANALYZER AND DATA
freq4
The left side of the photograph shows the hybrid HP-4792A LF Impedance Analyzer. The light brown box on the workbench next to the analyzer
holds the constant temperature bath where the batteries are placed during measurement. The analyzer measures the complete battery impedance
spectrum from 10-3 Hz to 100 kHz. The graphic representation of the reduced impedance data plotted as reactive vs. resistance illustrates the char-
acteristic parameters of battery impedance used in the decision tree model.
Where:
Rhf = the resistance of the electrolyte
Dia 1, Dia 2, Skew 1, and Skew 2 = measures of the lithium anode passivation
Freq 1 and Freq 2 = frequencies of maximum reactance due to lithium anode passivation
Intercept and Angle 4 = measures of the impedance contribution from the cathode, and
Freq 4 = frequency dependency of the cathode impedance.
UNCLASSIFIED SM-5
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 2 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
st:
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
? @
1 s.
--104 ets "
035 031 029
? ? ? ??? ".
\?yra \iv
`N\'-117W-Tailarg
-.4". r
-,N1N.0
7-\
\
1985
Accomplishments
-am"-
?
Analytic Methodology Research Division
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
SECRET
ANALYTIC METHODOLOGY RESEARCH DIVISION
PROGRAM AREAS
MILITARY ANALYSIS
POLITICAL ANALYSIS
GEO ECONOMIC RESOURCE ANALYSIS
PERSONAL SECURITY AND COUNTERMEASURES
COUNTERTERRORISM/COUNTERINTELLIGENCE
SECRET
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 6 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
:
?
si$1,3:3A
per
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
'13M
N 114,10ZPP911.,AluttIfiritio.
?,11rf4*N.A.116-11
it- gill 71;.:
-
\
1985
Accomplishments
Pt-
- -
,
Information Systems Research Division
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
ck_.Kc I
INFORMATION SYSTEMS RESEARCH DIVISION
PROGRAM AREAS
INFORMATION SYSTEMS
Computer and Network Architecture
Computer Security
Analyst Tools
Advanced Text Search and Retrieval
Data Base Management
Computer Aided Instruction
Artificial Intelligence
User Interfaces
Geographic Information Systems
SECRET
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
SECRET
PROCESSING AND ANALYSIS TECHNOLOGY GROUP
Information Systems Research Division
INTRODUCTION
The Information Systems Research Division (ISRD) conducts research and development in computer
science and information systems for the entire spectrum of the Agency's technical, analytical, and
administrative activities. The Division pursues projects in two general categories: projects to provide more
immediate product-oriented results in support of specific applications, and projects to investigate the potential
utility to the Agency of new or emerging information science technologies.
Currently the Division is pursuing projects addressing problems in computer security, advanced text
search and retrieval, and data base management. Other areas of activity include automated aids for language
training, user interfaces, fiber-optic bus technology, geographic information system development, Hypercube
computer architecture, and an Artificial Intelligence center for development of prototype Artificial Intelligence
systems.
SECRET IS-1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNCLASSIFIED
RULE-BASED RETRIEVAL OF INFORMATION BY COMPUTER (RUBRIC)
KEY PERSON:
CONTRACTOR:
PROBLEM
BENEFITING ORGANIZATIONS: FBIS, DI
Each day FBIS routinely processes massive amounts of unstructured narrative text documents from a variety of sources.
These documents are examined for content in a number of topic areas and are indexed according to formal guidelines. Although the
coding schemes used by FBIS ensure general consistency of indexing, the current procedures employed by FBIS are very labor
intensive and do not adequately meet all the requirements of the users. Substantial improvement is needed in the processing of
these large document databases.
Another problem confronting FBIS, and the Agency, is the lack of a text retrieval system that would enable analysts to peruse
large amounts of unstructured narrative text. Many commercial text retrieval systems are now available, but the retrieval languages
provided by these systems require that the documents be structured in a pre-defined format, a situation that is not currently cost
effective for either FBIS or the Agency.
PROBLEM SOLUTION
In an attempt to overcome the drawbacks of existing
information retrieval (IR) systems, ORD has developed a
research prototype system called Rule-Based Retrieval of Infor-
mation by Computer (RUBRIC). This system was designed to
help IR analysts gain access to large full-text databases. The
main goals in developing this prototype were that (I) the query
language should allow the user to express his or her needs in a
conceptually straightforward manner, and (2) the system should
be able to provide partial matching of queries to documents.
To achieve these goals, RUBRIC draws upon the rule-
based (If...Then) paradigm used in Artificial Intelligence (Al).
The system supports hierarchies of rules that define the inter-
relationship between concepts (e.g., terrorism and car bomb-
ing) that the analyst expects to appear in a typical document.
The lowest level concepts (e.g., explosion) in a hierarchy of
rules define theme categories, such as terrorism, that may not
actually appear in a document.
The RUBRIC environment provides analysts a number of
tools that help with the construction and evaluation of rules and
with the assessment of retrieval results. Overall, RUBRIC pro-
vides a friendly interactive environment for the efficient and
effective development of queries that (I) give better precision
and recall than those based on commercial structured informa-
tion retrieval systems (e.g., LEXIS, DIALOG, or MEDLARS), (2)
are more easily understood by the user than statistical meth-
ods, and (3) do not require the complex processing associated
with natural language methods. ORD is currently performing a
requirements analysis for FBIS to integrate RUBRIC into their
working environment.
PAYOFF
The RUBRIC system will make it possible for IR analysts to
process the increasing amount of material available to them.
More consistent coding of the material will result in improved
productivity; at the same time, analysts will have improved
IS-4
access to archived material. In addition, when new topics of
interest arise, analysts will be able to retrospectively retrieve
information from databases.
imri A CCIPIFr)
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release,20,11/05/05 : CIA-RDP88G01332R000100070015-8
Terrorism
.8
Terrorist
Event Politician
1.0
Assassinati. 5
on'
Reason
Action
1.0
Actor
.6
Violent
Event
I1.0 Violent
-4-- Effect
Violent
.8 Act
General
Actor
Revolution
Killing
Specific
Actor
Killing Bombing Kidnapping
.5 .4 .3 .3
Shooting Slaying
.5
Device Explosion
.7
Ransom
-4-
Kidnap
Event
Government
Opposition
Encounter Takeover
EXAMPLE OF HIERARCHY OF RULES WITH PROBABILITY WEIGHTS NOTED
This figure shows a hierarchy of rules defined by an individual analyst. Each analyst constructs his or her own theme category (in this figure,
terrorism). The hierarchy of rules gives the analyst the advantage of being able to search large data bases with computer technology instead of with
the currently used manual search procedures. The numbers in the figure correspond to the amount of relevance (probability value) that an analyst
may attach to key words (e.g., explosion) that would define the theme category.
IS-5 UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
KNOWLEDGE-BASED GEOGRAPHIC INFORMATION SYSTEM
KEY PERSON:
CONTRACTOR:
PROBLEM
BENEFITING ORGANIZATIONS: OIT, CPAS, OGI, 01A
Current Agency geographic information systems lack the ability to answer complex queries. These systems can display areas
of the world using data from the World Data Bank (WDB), a computer-readable map of the world containing political and geographic
borders, rivers, roads, railroads, major islands, and major lakes. However, the systems do not know what the lines used by the WDB
mean, e.g., that one line represents the boundary between two countries, while another line represents a river. They can calculate the
distance between two selected points but cannot determine the distance via roads between two cities because they do not under-
stand what a road is. They can display the locations of a list of cities but cannot infer that European NATO cities are cities in coun-
tries that are in NATO and in Europe. This lack of understanding on the part of current geographic information systems places on the
user the burden of translating graphic representations into real-world features.
PROBLEM SOLUTION
During 1984, ORD restructured the World Data Bank to
meet Agency needs and determined what geographic informa-
tion was required by Agency analysts. A prototype knowledge-
based geographic information system, known as KBGIS, was
then developed that represented, stored, and utilized this
information to answer complex geographic queries and to
display the results graphically. The amount and type of data
available to the system was expanded by developing two other
data bases, one containing statistical information and the other
inferential information. Integrating these three data bases
made it possible to use them in combination to answer ques-
tions that one data base alone could not answer. This prototype
system demonstrated that KBGIS can analyze a query, retrieve
pertinent information, make appropriate inferences and calcu-
lations, and display the results.
A production KBGIS now being developed will reside on
the Agency's mainframe computer and will be accessible to
analysts through their PC workstations. A significant improve-
ment made to the operational system in 1985 was a totally
redesigned query language. A hierarchical menu provides an
environment in which it is easy for a user to understand the
essence of the data model and to develop queries. Once a
query has been completely specified, an estimate of the time
the system will take to respond is displayed?information which
should be useful to the analyst in scheduling his or her time.
With KBGIS, the geographic knowledge base (GKB) can be
augmented with information from a user's personal data base
that does not exist in the GKB; e.g., the course of a submarine
over a specified period of time at specific time intervals. The
production KBGIS will be delivered to OIT/DI in 1986.
PAYOFF
KBGIS will give analysts timely access to information that
is now difficult to obtain. Complex queries can be answered
quickly and conveniently. Questions that now require research
by geographers will require only a query of the system. In
addition, questions that are now not asked at all because of the
perceived difficulty of obtaining an answer will be answered in a
timely and understandable way.
IS-6 UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
AREA/PERIMETER MENU
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNLLAbblhltU
I LENGTH OF PERIMETER
2 AREA WITH
3 AREA
4 ARE
SELECT OP
ANALYSIS OF GEOGRAPHIC INFORMA
1 POINT/AREA TARGET
2 ? SPATIAL PROCESSING
3 THE
SELECT
SPATIAL PROCESSING MENU
1 ADJACENT
4
5
6
7
8
E XI
ANALYSIS OF GEOGRAPHIC INFORMATION MENU
1 POINT/AREA TARGET
1 SPATIAL PROCESSING
3 T
AREA TARGET FUNC
1 OVERLAY UNI
2 OVERLAY MERG
3 - ARE
4 FIL
NETWORK OVERLAY SELECTIONS MENU
1 RAILROAD
2 ROAD
3
SELECT
CIRCLE MENU
RADIUS - EITHER
1 CITY
2 ? PRO
3 ? C?
4 PL
5 PL
6 - SU
7 ? DIS
SELECT
Z . 300
CIRCLE MEN
CENTER
GEOGRAPHIC ENTITIES MENU
1 AREAS
2 - ORGANIZATIONS
3
4
5 -
1 ? CITY
2 PROVINCE
3 COUNTRY
4 ? PLACE CURSOR IN A
5 ? PLACE CURSOR INSID
6 ? SUPPLY LAT/LON COO
SELECT AN OPTION
TEGUCIAGALPA
PRE-BUILT AREA OF INTEREST SPECIFICATIONS MENU
1 OAS COUNTRIES
2 NATO COUNTRIES
3 BALKANS
4 EASTERN EUROPE
5 SCANOANAVIAN COUNTRIES
KNOWLEDGE-BASED GEOGRAPHIC INFORMATION SYSTEM
Figure 1: This is an example of a series of query building menus from the KBGIS. The system was asked to show, graphically, all the provinces
having an area greater than 100,000 acres that lie within a 300-kilometer radius of the Honduran provincial capital of Cuscatleu, Tegucigalpa.
Figure 2: All Central American countries other than Mexico are displayed and named in white. The provincial boundaries of these countries are
displayed in yellow-green. The location of Tegucigalpa is highlighted in orange. All provinces satisfying the specified query are shaded in black and
the 300-kilometer-radius circle is seen, faintly, in black.
UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
IS-7
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNCLASSIFIED
ADVANCED TEXT RETRIEVAL
KEY PERSON:
CONTRACTOR:
BENEFITING ORGANIZATIONS: FBIS, NPIC, OSO, ORD,
SOVA, OCR, OIT, others
PROBLEM
The Agency does not have a rapid text search and retrieval system that enables users to peruse very large volumes of relatively
unstructured narrative text. As the textual information and the number of documents being generated and collected continue to
expand, the magnitude of the problem of searching through all the data also expands. Furthermore, a lack of common architecture
for text retrieval systems forces the Agency to design each of its text retrieval applications individually.
PROBLEM SOLUTION
During 1984, ORD developed a general architecture for full
text management. The first prototype, implemented on the
Apollo workstation, integrated text processing and retrieval
tools to permit interactive browsing of unstructured narrative
text.
In 1985 the text retrieval system was converted to a truly
portable, efficient system by providing portable window man-
agement and communications systems. These systems provide
a uniform interface to the operating environment for applica-
tions programs by isolating the machine, display, and network
dependencies in special interface modules. The text retrieval
system can now be used on a variety of machines without
changes to most modules. Program maintenance has also been
simplified since there is a single version of most programs.
Process control and system-monitoring facilities have been
substantially expanded. The index, search, and document-
access functions have been respecified to aid in supporting
personal files, annotations, and gateways to other systems.
More extensive error detection and handling have also been
included. A preliminary test version of the system has been
delivered to SOVA/DA.
The testbed retrieval system provides an ideal environment
for human factors experimentation, and the initial design of a
methodology for human factors experiments is now being
investigated. In particular, work is under way to develop a set of
experiments comparing various indexing techniques such as
RUBRIC and N-Grams (both ORD programs), as well as aspects
of the window-based user interface. A number of new modules
are also being developed to support experiments, including an
experimental control module to select the system configuration
for a particular experiment; a calibrated delay module to
simulate system performance with considerably larger data-
bases; and a number of programs to collect, format, and
analyze the data from an experiment.
PAYOFF
The Text Information Management System will give the
Agency a common, powerful, and easily maintainable text
retrieval architecture that can be used in various system
development projects. The system will improve analysts' pro-
ductivity by providing an integrated set of text processing and
retrieval tools to permit interactive browsing of text databases.
Further, the text retrieval testbed will provide future system
planners with empirical data concerning alternative system
design issues and the relative merits of various text retrieval
hardware and software.
IS- 10 UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
USER
Window
Manager
H Query 1_
Reformulator
e.g. RUBRIC
HUser
Interface
H
1-1-1 j---
User
Interface
1 Displayer
_I Electronic
Mail
H Word
Processing
logical networks
Database
Manager
Database
Manager
Index
e.g.N-grams
Search
I Document
Access
Index
?1 Search
_I Document
Access
I
Database
Manager
Nontextual
Database
w/DBMS
Existing
Intelligent
Text Retrieval System
Gateway
e.g. SAFE
TEXT MANAGEMENT INFORMATION SYSTEM ARCHITECTURE
This figure shows a simplified schematic of the Text information Management System in modular form. The user works with the system through the
window manager by means of a keyboard and a mouse not shown). Two logical networks provide flexibility in both user services (on the left) and
data base servers (on the right) The user has the convenience of working with multiple windows (each with its own editor) to keep track of the text
retrieval process from original query to the final document -hit list.- The system also contains a system monitor and a system controller (not shown).
`RUBRIC: An ORD-developed system (Rule-Based Retrieval of Information by Computer).
CONFIDENTIAL
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
IS-11
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
-4 la
? el istito,
U u
? *
??,..0 , _ u , ,
035029
? ? ? .4 ..
? ?:
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
, ,
1
!Oar
sHi it" "
,
? ., ? ?
023 02
73N4
, I. ,;.7 II 17,24 IW'
It\ FP*.
. ? -
k
1985
Accomplishments
Processing and Exploitation Research Division
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 1 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
R
Next 6 Page(s) In Document Denied
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
1985
Accomplishments
Seminars and Symposia
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNCLASSIFIED
SEMINARS AND SYMPOSIA
INTRODUCTION
An important part of ORD's responsibility is the promotion and promulgation of relevant technology
throughout the Agency and the Intelligence Community. One mechanism for accomplishing this function is the
sponsorship and presentation of a growing number of technical seminars, meetings, and symposia. ORD has a
continuing calendar of such events; in this section we describe the most noteworthy of those which were held
in Fiscal Year 1985.
UNCLASSIFIED SS-1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
CONFIDENTIAL
VISITING SCHOLAR LECTURE SERIES
KEY PERSON:
CONTRACTOR: Various
BENEFITING ORGANIZATIONS: Agency-Wide, Intelligence
Community
PROBLEM
Current research in industry and academia often holds great promise for Agency applications. Agency professionals, however,
seldom have the time or opportunity to keep abreast of such research. Ways are needed to expose these professionals to advanced
research in their fields of expertise.
PROBLEM SOLUTION
In 1982 ORD originated the Visiting Scholar Lecture Series,
which brings outstanding scholars from industry and academia
to the Agency to lecture about exciting new research of
potential application to Agency participants. The 1984-85 Lec-
ture Series was organized as three five-day seminars on
aspects of signal processing and related disciplines, with pri-
mary speakers selected to represent both the best research in
the field and a balance of differing perspectives. Professor
Thomas Kailath of Stanford University was selected to assist in
the organization of the series. Kailath is internationally ac-
claimed for his ability to synthesize and exploit concepts from
automatic control, communications, and signal processing,
particularly as they relate to linear least-squares estimation?a
powerful collection of mathematical techniques.
The first seminar examined certain modeling trends devel-
oped originally within the automatic control and navigation
communities. Kailath presented exciting new work by which
these primarily linear techniques are being extended to such
fields as signal detection, data transmission, and antenna
design. Representing a different point of view, Harold Sorenson
of the University of California at San Diego discussed non-linear
modeling techniques and applications.
In the second seminar, Kailath discussed his current
research in architectural structures; Professor David Casasent
of Carnegie-Mellon University surveyed current promising
research in optical device implementations; while Earl
Swartzlander, a well-known digital designer from TRW, Inc.,
critiqued this work in light of developing electronic technology.
In the third seminar, Kailath?ably assisted by several
former students who have applied his work at IBM and
Hughes?discussed new algorithmic work in exact least-
squares methods for adaptive signal processing; Professor
David Messerschmidt of the University of California at Berkeley
and John Treichler, an Agency contractor, assessed these and
alternative techniques for a variety of applications.
Researchers from the Intelligence Community, industry,
and the national labs participated in these seminars. Response
was so positive that the program will continue under the joint
sponsorship of the NSA National Cryptographic School and
possibly other agencies.
PAYOFF
This program brings the best minds in a given field
together to discuss problems of particular interest to the
Intelligence Community (IC). Agency and IC professionals have
the opportunity to discuss their technical problems, within
security constraints, with leading researchers. Conversely,
these same researchers are exposed to (generally unclassified)
IC research interests in such a way as to peak their own interest
in similar problems. This, in turn, benefits the Agency's long-
term research program by stimulating related academic and
industrial research and by fostering cooperation and interac-
tion between research offices in different components of the
Intelligence Community.
SS-2 CONFIDENTIAL
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
25X1
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
LI N
CONCURRENT PROCESSING SYMPOSIUM
KEY PERSON:
CONTRACTOR:
BENEFITING ORGANIZATION: Intelligence Community
PROBLEM
The Agency has updated its computer center regularly, but the complexity of its analytical problems has grown even faster.
New technology has made huge amounts of data available for analysts, who need more powerful and faster computers to facilitate
the searching and analyzing of large textual databases in real time. In addition, industry is using supercomputers for analytic model-
ing, and our analysts must have comparable tools to use these models. This is especially true for analyzing similar models produced
by foreign countries.
PROBLEM SOLUTION
One solution to these problems is the use of supercom-
puters, where dramatic breakthroughs and improvements oc-
cur regularly. The first step in understanding how these break-
throughs can benefit the Intelligence Community is to
familiarize Intelligence Community scientists and managers
with the state of the art in concurrent processing machines and
supercomputers.
To do this, a Concurrent Processing Symposium was
sponsored by CIA and funded by DARPA; administrative sup-
port was provided by the BDM Corporation. The symposium
was held on 15-17 July at the West Park Hotel, Tysons Corner,
and on 18 July at CIA Headquarters. The two hundred fifty
attendees were primarily from CIA but also from other govern-
ment agencies in the Intelligence Community, including
DARPA, Navy (NRL, NSWC), DOE, NSA, and FBI, as well as
from the White House National Security Council, Air Force, and
the Al Steering Group of the Intelligence Research and Devel-
opment Council.
The three-day unclassified sessions were held to present
(1) the currently available hardware; (2) applications for which
the machines are intended and currently being used, including
unclassified applications of interest to the IC; and (3) architec-
tures and hardware that will be available within two years.
Speakers were drawn from major corporations involved
with the hardware and architectures, university professors
involved with the research, and government agencies. Topics
ranged from an overview of conventional and emerging archi-
tectures to a discussion of applications, modeling, and simula-
tions on supercomputers.
On the fourth day, panel discussions were held on classi-
fied applications of multiprocessors that would benefit the IC.
Panel members presented their problems; manufacturer repre-
sentatives gave a quick estimate of how their architectures
would handle the problems.
The conference was videotaped and a conference Pro-
ceedings was compiled. Both are available upon request.
PAYOFF
The symposium was highly successful. Managers and sci-
entists have been made aware of currently available hardware
and architectures and how supercomputers can assist them
with their particular problems. Enthusiasm has been generated
about the benefits to the Intelligence Community of both
current and future developments in supercomputers and con-
current processing machines. In addition, much beneficial inter-
action occurred among people from corporations, universities,
and the government. Responses to evaluation forms rated the
symposium as excellent. A follow-on symposium was suggested
to cover future developments in architectures and software.
SS-4 UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNCLASSIFIED
CONCURRENT PROCESSING SYMPOSIUM
AGENDA
MONDAY
Orntation
ORD, CIA
Symposium Chairperson
Welcoming Address
Mr. R. E. Hineman, Deputy Director for
Science & Technology, CIA
Overview of Conventional Supercomputer
Architecture
Dr. Signey Fembach, Consultant
Applications on Conventional Supercomputers
Dr. John Connelly, Director of Office of
Advanced Scientific Computing, NSF
Overview of Emerging Architectures
Dr. Stephen Squires, DARPA
TUESDAY
Cray Architecture
Dr. W. Jack Woriton
Modeling and Simulation on the Cray
Dr. Ralph Brickner, Los Alamos
ETA Systems
Dr. Charles Purcell
Denelcor HEP
Dr. Burton Smith
Convex C-1
Dr. Steve Wallach
Floating Point Systems FPS-264
Dr. John Gustafson
LUNCHEON SPEAKER: Dr. William 0. Baker,
Foreign Intelligence Advisory Board and Retired Chairman of
the Board, AT&T Bell Laboratories?"The Dimensions of the
Information Age"
Flex/32
Dr. Nicholas Matelan
Warp Architecture and Applications
Dr. H.T. Kung, Carnegie-Mellon University; Bruce Roberts,
GE; A. D. Toll, Honeywell
Hypercube Architecture
Dr. G.C. Fox, CalTech and JPL
Intel iPSC
Dr. Cleve Moler
IBM RP3
Dr. Greg Pfister
WEDNESDAY
Butterfly
Dr. Randy Rettberg, BBN
Teradata Database Machine
Dr. Phil Neches, Teradata
Fast Data Finder
Dr. Kwang-i Yu, TRW
Ultracomputer
Dr. Allen Gottlieb, NYU
Massive Memory Machine
Dr. Richard J. Lipton, Princeton
Connection Machine
Dr. Howard Resnikoff, Thinking Machines
LUNCHEON SPEAKER: Dr. Gary Demos,
Digital Productions
Concurrent Processing at TRW
Dr. Charles Childress, TRW
Vision and Data Fusion
Dr. Dan Dudgeon, MIT Lincoln Labs
Photointerpretation & Cartographic Data Bases
Dr. Azriel Rosenfeld, University of Maryland
Benchmarking and Performance Engineering
Dr. James Browne, University of Texas
WRAP-UP PANEL AND AUDIENCE DISCUSSION?TOPIC:
Factors Which Help Determine the Best Architecture for a
Given Application
THURSDAY
Panel #1: Imagery and Computer Graphics
Panel #2: Statistical Modeling and Fast Text Search
Panel #3: Past, Present, Future Needs in the
Intelligence Community
UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
SS-5
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
UNCLASSIFIED
TEXT RETRIEVAL CONFERENCE
KEY PERSONS:
CONTRACTOR: None
BENEFITING ORGANIZATIONS: CIA and other
Government Agencies
PROBLEM
ORD currently is supporting research and development into a number of information retrieval projects, including N-Grams,
RUBRIC, Advanced Text Retrieval, and NoteCards, all of which are described in this Accomplishment Book. There is much interest in
the strengths and limitations of each of these information retrieval techniques and how they compare. Are these projects all neces-
sary or are they duplicating the same work?
PROBLEM SOLUTION
To address these issues, a Text Retrieval Conference was
held to examine in detail the strengths and limitations of each
of the information retrieval techniques and to identify their
applications, concepts, and technologies.
The conference, which was held on 15 and 16 January
1985 at the Arlington Hyatt Hotel, was attended by seventy
people. Most of the participants were from CIA; but the FBI,
NSA, DARPA, and other organizations in DoD were also repre-
sented. Corporations and universities were represented by their
employees who are currently doing text retrieval work for the
Agency.
Formal presentations were given on N-Gram Indexing,
Architectures for Text Retrieval and Handling, RUBRIC, and
NoteCards. Presentations were made by the principal investi-
gators from PAR Technology Corporation, Advanced Informa-
tion and Decision Systems, the University of Utah, and Xerox
SIS. The presentations were divided into two parts: nontechni-
cal to present the overall flavor of the project, and technical to
describe the project in detail.
To afford the opportunity for "hands on" experience,
demonstrations of the project systems were held in a separate
room. These demonstrations ran concurrently with the presen-
tations, during breaks, and during the evening.
On the last day presentations were made by ORD and
customers on the future needs and requirements of the Agency.
A technical exchange followed in which the principal investiga-
tors described where their research is heading and speculated
on the future. The meeting concluded with panel discussions on
topics previously covered by each of the contractors.
The luncheon speaker, Dr. Bruce Croft, Professor, Com-
puter and Decision Sciences Department, University of Massa-
chusetts, Amherst, gave an informative presentation relating
the past, present, and future of text retrieval. The conference
was videotaped and a Proceedings was compiled. Both are
available upon request.
PAYOFF
The objectives of the conference were achieved. Insight
was gained into the strengths and limitations of each of the text
retrieval projects, as well as into the differences and similarities
between them. The differences were justification that each
project has its own important contribution; the similarities
provided a basis for future and continued joint work between
contractors. The conference also generated additional interest
in these projects from other offices in the Agency and other
government agencies.
SS-6 UNCLASSIFIED
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
STAT
STAT
STAT
STAT
STAT
STAT
Sanitized Copy Approved for Release 2011/05/05 : CIA-RDP88G01332R000100070015-8
WIN
TEXT RETRIEVAL CONFERENCE
AGENDA
15 JANUARY (morning)?NONTECHNICAL PRESENTATIONS
Welcome by Dr. Philip K. Eckman. Director, ORD
Opening Remarks by C/PATG/ORD
Program Overview: Ten-minute introduction to projects by each
COTR, followed by a 25-minute overview by principal investigator.
DOTR, ISRD/PATG
Mr. Ray D'Amore, PAR Technology
COTR, ISRD/PATG
Dr. Richard M. Tong, Advanced Information & Decision Systems
COTR, ISRD/PATG
Dr. Lee Hollaar, University of Utah
COTR, AMRD/PATG
Mr. William Liles, Xerox SIS
15 January (afternoon)?TECHNICAL PRESENTATIONS, with questions and answers at the end of each
presentation.
Mr. Clinton P. Mah, PAR Technology
Dr. Richard M. Tong, Advanced Information & Decision Systems
Dr. Brian P. McCune, Al&DS
Dr. Lee Hollaar, University of Utah
Dr. Frank G. Halasz, Xerox PACC
DINNER SPEAKER: Dr. W. Bruce Croft, Professor, Computer & Decision Sciences Department, University of
Massachusetts
16 JANUARY
Presentation by ORD and Customers on Future Needs and Requirements of Agency
Senior Scientist, PATG/ORD
Technical Exchange
Technical presentations by principal investigators on where research is heading and speculations on the future.
Panel Discussions on Previous Topics by Each Contractor
UNCLASSIFIED SS-7
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
LUNNUCN I IHL
ARTIFICIAL INTELLIGENCE SYMPOSIUM
KEY PERSON:
CONTRACTOR:
PROBLEM
BENEFITING ORGANIZATION: Intelligence Community
Artificial intelligence (Al) is concerned with giving information systems more human-like abilities (such as seeing, hearing,
planning, reasoning, or explaining their own behavior). Recent progress in artificial intelligence has generated considerable interest
throughout the Intelligence Community in applying Al technology to the collection, processing, analysis, and production of
intelligence. However, developing successful applications will require both broad Community awareness of Al technology and in-
depth experience in the use of Al tools and techniques.
PROBLEM SOLUTION
In conjunction with the Artificial Intelligence Steering
Group of the Intelligence Research and Development Council,
ORD has sponsored annual symposia on Intelligence Applica-
tions for Artificial Intelligence. These symposia are intended to
raise the general level of awareness of the Community concern-
ing the science and technology of artificial intelligence and to
stimulate thinking about how Al might best be applied to the
Community's information problems.
The third Al Symposium was held on 19, 20, and 21 March
1985 at DIA Headquarters. Over 600 persons attended from the
Community, industry, and academia. The Symposium featured
technical presentations by Al researchers; summaries of major
PAYOFF
government programs, such as the Joint National Intelligence
Dissemination System and the Software Engineering Institute;
continuous product demonstrations by hardware and software
vendors; and classified problem-examination sessions called
Special Interest Working Groups. There were nine such working
groups: Data Support to the Analyst, Image Understanding,
Signal Processing/Analysis, Dissemination and Language Un-
derstanding, Strategic Defense, Indications and Warning, Col-
lection Management, Decisions Under Uncertainty, and Train-
ing. Results of the discussions within each group were reported
to the entire assemblage to assist in defining requirements
areas for Al applications.
The three Al Symposia have been highly successful in
creating a broad awareness of the potential benefits of artificial
intelligence for the Intelligence Community. A fourth sympo-
sium is planned for 1986. This continuing exposure to the most
recent
limited
developments in Al will help the Community use its
resources most effectively in coping with an increasing
amount and variety of information.
SS-8 CONFIDENTIAL
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
...,,,,,L, N I Ir-% I-
ARTIFICIAL INTELLIGENCE SYMPOSIUM
AGENDA
Keynote Speakers
John D. Macartney, Colonel, USAF, Commandant,
Defense Intelligence College
James A. Williams, Lieutenant General, USA,
Director, Defense Intelligence Agency.
Technical
Joint National Intelligence Development Staff
Tom Bushbach, JNIDS
Architecture and Technology for a Strategic
Surveillance Analysis System
David S. Spain, Advanced Information &
Decision Systems, Inc.
Reporter
John Woods, TRW
A Knowledge-Based System for Space Threat
Warning
Christine A. Montgomery, Logicon, Inc.
Utility of Artificial Intelligence in Missile Typing
Benjamin R. Peek
LES: A Generic Expert System
Walton A. Perkins, Lockheed
Applications of Expert Systems to Training
David L. Young, Mystech Associates, Inc.
Expert System Training at DEC
Gerhard Friedrich, Digital Equipment
Corporation
Big Oz: An Intelligent Information Management
System for Analysts
25X1 CIA
Al Systems for Interpretation of Synthetic Aperture
Radar
Bobby Hunt, Science Applications International
Corporation
Dr. Philip K. Eckman, Chairman, Al Steering Group,
Office of Research and Development, Central Intelli-
gence Agency.
Dr. Saul Amarel, Rutgers University
Papers
Threat Assessment Tools for Operational Platforms
Glen R. Allgaier, Naval Ocean Systems Center
The Artificial Intelligence Enemy Course of Action
Evaluation Aid
Michael L. Donnell, Science Applications
International Corp.
Warning Analysis
John McCreary, National Warning Staff
Inference Theory and the Value of Evidence
David A. Schum, Rice University
Evidential Reasoning for C3I
John D. Lowrance, SRI International
Collection Management Enhancement
Richard Scott, GTE
A Weapon System Model Using Al Techniques
R. Allen Riley, Los Alamos National Laboratory
Distributed, Cooperating Expert Systems for Signal
Understanding
Mark A. Williams, DELFIN Systems
NoteCards: An Information Structuring Aid for
Analysts
Williams Liles, XEROX
An Evolving System for Image Understanding
Julius F. Bogdanowicz, Hughes Aircraft
CONFIDENTIAL
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
SS-9
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
LUNfrlUtN I IAL
ARTIFICIAL INTELLIGENCE MANAGEMENT SEMINARS
KEY PERSON:
CONTRACTOR:
PROBLEM
BENEFITING ORGANIZATION: Intelligence Community
The popular media has recently reported many successful applications of Artificial Intelligence (Al). Claims have been made
that Al programs have discovered mineral deposits, interpreted oil drilling charts, performed medical diagnoses as accurate as those
performed by physicians, correctly interpreted passages of text, and configured computer components and systems. Al appears to
be one solution to the Intelligence Community's problem of coping with both the increasing collection of data and the scarcity of
technically qualified analysts. However, the Community's managers and programmers lack the familiarity with Al technology they
must have before they can assess the actual prospects and limitations of Artificial Intelligence.
PROBLEM SOLUTION
In conjunction with the Artificial Intelligence Steering
Group of the Intelligence Research and Development Council,
ORD sponsored a series of instructional seminars to investigate
the potential applicability of Al technology to intelligence prob-
lems. These seminars were organized to bring managers and
programmers from the Agency and other Intelligence Commu-
nity organizations up to date on the prospects and limitations of
Al for the Community's information-handling problems. The
seminars consisted of a series of six three-day courses attended
by a total of 148 mid- and senior-level managers; a four-week
knowledge engineering course attended by 18 programmers; a
two-day software engineering course attended by 24
technical managers; and ten sessions on "Al Tools for Building
Expert Systems" (with demonstrations by vendors) attended by
a total of 52 different programmers and technical managers.
Leading Al researchers from Yale University, University of
Texas, University of Pittsburgh, and Ohio State University
taught the courses.
During 1986 ORD will sponsor two-week courses on LISP
programming, Logic programming, and knowledge engineer-
ing. The courses will focus on providing skill training in these Al
areas for Agency scientists and engineers.
PAYOFF
Evaluations from participants in the seminars indicated
that they perceived these seminars to be valuable and informa-
tive. Al was seen as showing promise of helping to resolve
some problems but was not seen as a solution or a "quick fix"
for all problems. Many participants suggested continuing such
seminars in order to inform new managers about the most
promising areas for detailed study and to help technical pro-
grammers focus on those areas. The knowledge gained from Al
seminars will also help in preparing statements of work and in
evaluating proposals pertaining to Al applications. Further-
more, an informed management will facilitate technology trans-
fer, which should benefit the entire Intelligence Community.
SS-10 CONFIDENTIAL
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8
Sanitized Copy Approved for Release 2011/05/05: CIA-RDP88G01332R000100070015-8