TIMP270001-9 Approved For Release 2000/08/07 :C0ftF*D1Nr 140 Final Report DATABASE MANAGEMENT FEASIBILITY (U) Prepared for: DEFENSE INTELLIGENCE AGENCY WASHINGTON, D.C. 20301 Attention: SG1J CONTRACT MDA903-81-C-0292 January 1982 SPECIAL ACCESS PROGRAM FOR GRILL FLAME. RESTRICT DISSEMINATION TO ONLY INDIVIDUALS WITH VERIFIED ACCESS. 333 Ravenswood Avenue Menlo Park, California 94025 U.S.A. (415) 326-6200 Cable: SRI INTL MPK TWX: 910-373-2046 Intermi 3onal go Mved For Release 2000/08/07:CGkiFdEDEN;EbAiD270001-9 pib 40 a SLELF.2212.1d For Release 2000/08/07 : C0FNMFyNrUA6270001-9 UID Final Report Covering the period October 1980 to September 1981 DATABASE MANAGEMENT FEASIBILITY (U) 00 W By: EDWIN C. MAY Prepared for: DEFENSE INTELLIGENCE January 1982 AGENCY Contract No: MDA903-81-C-0292 WASHINGTON, D.C. 20301 SRI Project 3279-7 Approved by: ROBERT S. LEONARD, Director Radio Physics Laboratory DAVID D. ELLIOTT, Vice President Research and Analysis Division CLASSIFIED BY: DT-1A REVIEW ON: 31 January 2002 FIN 333 Ravenswood Avenue International (415) 326-6200 ' Cable: '-*~ N, x proved For Release 2000/08/07 CONFIDENTIAL I Menlo Park, California 94025 * U.S.A. SRI INTL MPK * TWX: 910-373-2046 : CIA-RDP96-0078MO1300270001-9 Copy No. ..14') ..... This document consists of 12 pages. SRI/GF-0010 Approved For Release 0nMfffDTDCW1Pfjkt90788r001 300270001-9 CONTENTS LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . iii ,to I OBJECTIVE . . . . . . . . . . . . . . . . . . . . I . . . . . . 1 II INTRODUCTION . . . . . . . . . . . . . .. . . . . . . . . . . 2 00 111 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . 3 IV CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . 8 V RECOMMENDATIONS . . . . . . . . . . . . . . . . . . . . . . . 9 Aw ~qw Aw w CONFIDENTIAL Approved For Release 2000/08/07 : CIA-RDP96-0078MO1300270001-9 Approved For Releas t 26 -10 wmAfb:fftf fA9E-00788r001 300270001-9 TABLES 1 Resource Assessment Summary . . . . . . . . . . . . . . . . . 4 2 Training Target Data . . . . . . . . . . . . . * . . . . . . 6 iii CONFIDENTIAL Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 Approved For Release-2QD 00788r001300270001-9 I OBJECTIVE The objective of this program is to develop concepts and materials to determine if remote viewing data can be placed on a computerized data- base-management system for each access and manipulation. CONFIDENTIAL Approved For Release 2000/08/07 : CIA-RDP96-0078MO1300270001-9 Approved For Releas&02M"FgDEINqqFA4L-00788rOOl3OO270001.9 II INTRODUCTION The manipulation of textual data was among the earliest applications of digital computers. From these earlier uses, it was immediately apparent that two separate requirements should be imposed on data-base management (DBM) designers: 0 Develop task specific DBM programs. 6 Develop generalized DBM programs that allow any specific application to be designed at a later time by the end-user. The first requirement was easily met. An abundance of data-base management systems (DBMS) for specific tasks exist that handle everything from general accounting for business to map generation by graphic DBM. The second requirement, however, has not yet been adequately realized. Although many computer vendors offer generalized DBM design programs, they all appear to be of limited use when applied to a specific set of design requirements. The problem of creating a DBM design program is not trivial; to be useful to the widest possible audience, it must be very general. Yet, this very generality makes the final DBM cumbersome to use. This report describes test applications of a commercial DBMS to remote viewing (RV) data and recommends a threefold approach to the general problem of RV DBM. 40 2 CONFIDENTIAL Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 Approved For Releastnllpl-DE1471§NJ.00788rOO1 300270001-9 III DISCUSSION In the Grill Flame mid-year report, we demonstrated the use of a commercially available data-base management (DBM) program applied to two SG1J problems in RV data management. The first of these was the categorization and display of the results of a simulated RV analysis data base. Using the tentative assessment form, .which operating RV analysts had suggested, LUWO we analyzed 25 RVs each for 4 remote viewers. Table I shows a character- istic output from this data base, which is repeated here from The column headings are the seven categories taken from the assessment CO UJF- I-Pic form. Under each category the remote viewers are listed in ascending order of assessment averaged over all viewings to date for that category. For example, we note that remote viewer 007 was least successful at obtaining information about the geography of a site. However, when every- thing is considered, remote viewer 007 had the best overall utility. This and similar information might have been overlooked with manual inspection. 'OU ID The row below the dashed lines contains the averages for each column, which can be considered as the "facility" assessment for each category. For example, the represented facility does best on geography elements and second best when targeted against activity at the remote site. In a second example, we have used a DBMS to organize and manage a growing number of training targets from the National Geographic (375 as of this report). Our RV enhancement effort currently involves four remote viewers, six interviewers, four target preparers, and numerous target selection criteria. With this level of complexity and a growing SG1J "DoD GRILL FLAME Progress Report (U)," presented at Mid-Year GF Meeting at DIA on 30 April 1981 (April 1981, SECRET). 3 CONFIDENTIAL Approved For Release 2000/08/07 : CIA-RDP96-0078MO1300270001-9 Ik It. It % %~ 11, %~ %. 11 Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 Table 1 RESOURCE ASSESSMENT SUMMARY (23 Apr 1981) ri 0 z Geography Elements Elements Ambience Activity Personnel Utility LS SS ID Ave ID Ave ID Ave ID Ave ID Ave ID Ave ID Ave 007 1.39 712 1.15 007 - 712 0.95 712 1.14 007 0.53 712 0.55 0.71 126 1.44 126 1.33 712 0.73 126 1.50 126 1.62 531 1.21 531 1.38 531 1.59 531 1.57 531 1.31 531 1.75 531 1.75 712 1.56 126 1.55 712 2.63 007 2.09 126 1.56 007 1.81 007 1.86 126 2.00 007 2.63 1.76 1.53 1.08 1.50 1.59 1.32 1.53 Average of each column. Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 Approved For Release &WQ C"fAVMM01 300270001-9 qQ-WffFTD number of targets, a DBMS was mandatory to avoid duplication and to provide target statistics. As an example of the DBMS output, Table 2 shows a small portion of the existing data. The targets were selected solely on the basis of their use as calibrations for operational RV sessions. They are displayed chronologically as a demonstration of a sorting procedure. Both of the above DBM applications were done on a large-scale PRIME 750 computer using the PRIME Corporation's POWER+ packaged DBMS. In spite of the computing power of the PRIME 750 and generality of POWER+, four specific problems were identified: This DBMS: (1) Did not allow for the formation of specific data records that contained parts of separate records from the remaining data base. (2) Did not easily allow a change of data base. (3) Did not allow for tree structures. (4) Can be installed on the SRI International's PRIME 400 at a cost of $75,000 to $125,000. Table 1 is an example of the first problem described above. The simulated 0CF-S T141c, 113 - T"FAr 1A MMS IT VXY\ T U-109-r- data base was entered and stored by grouping together all the parameters from the analysis of a single remote viewer for a single operational *)OVA - session. Yet, to provide the summary information in Table 1, we must collect scores and IDs from across many sessions. In fact, Table 1 was _90~ manually constructed to demonstrate a concept and was beyond the capability of POWER+. The second problem applies to both DBM examples. In developing the 00 designs of the respective data bases, it was necessary to modify them in progress to assure an optimum configuration. POWER+ requires that all the data be reentered into the new design.. Although POWER+ has a file conver- W 441ckA ~s sion capability that makes possible reentry of data from a disc, it was K40V-L-: 40 qmzasonk cumbersome; therefore, we decided to reenter the data manually. 6 pec#z tum-nse 5 OF CONFIDENTIAL 10 Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 1k %_ %, k %. k %, k 4 Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 Table 2 TRAINING TARGET DATA (23 April 1981) Date Time LatitudeLongitudeTarget Location MonitorClassPreparer - 15 Feb 1021 42'341N08'441ECalvi, Corsica PuthoffC Unknown 1980 3 Mar 0910 11'26'S5~'041WCampo de Diauarum, Brazil PuthoffB Unknown 1980 3 Mar 0917 le OOIN121'001ELake Taal-Luzon, Phillipines PuthoffB Unknown 1980 3 Mar 1000 12'30'N70P001WAruba Island, Lesser AntillesPuthoffC Unknown 1980 1 Jul 0900 210381N1570041WOahu, Hawaii C Unknown 1980 I Jul 0952 31'301N35930'EDead Sea, Israel/Jordon SG1J C Unknown 1980 2 Oct 0820 18P29'N66" San Juan, Puerto Rico PuthoffC Puthoff 1980 08 1 W 2 Oct 0900 58012'N06P231WStornoway Lewis Is., ScotlandPuthoffC Unknown 1980 2 Apr 0905 3e 401S59'301WBuenos Aires, Argentina PuthoffC Harary 1981 2 Apr 0945 450301S165'301EDusky Sound, New Zealand PuthoffB Harary 1981 3 Apr 0807 22030'N88'20'ECalcutta, West Bengal-India C Harary 1981 I 3 Apr 0815 51'131N04925'EAntwerp, Belgium C Harary 1981 3 Apr 0840 41~02'N28P571EIstanbul, Turkey SG1J C Harary 1981 3 Apr 0920 39P441N4e23'E Mt. Ararat, Turkey B Harary 1981 3 Apr 1010 35'091N32'47'EVouni, Cyprus B Harary 1981 3 Apr 1015 41050'N71'281WProvidence, Rhode Island C Harary 1981 6 Apr 0817 OUP141S7BP30'WQuito, Ecuador PuthoffC Harary 1981 6 Apr 0824 390181N7&38'W Baltimore, Maryland PuthoffB Harary 1981 6 Apr 0827 36P36'N830401WCumberland Gap, Kentucky-Tennessee-VirginiaPuthoffB Harary 1981 6 Apr 0835 0OP19IN3235'E Kampala, Uganda PuthoffB Harary 1981 6 Apr 0845 370371N719P Natural Bridge, Virginia PuthoffB Harary 1981 33 1 W 6 Apr 0930 130321S71o57'WCuzco, Peru PuthoffB Swann 1981 6 Apr 0936 28591N 139401WMontana del Fuego, Lanzarote PuthoffB Swann 1981 Canary Island 6 Apr 0945 51 029'NOOP38'WWindsor Castle, Windsor EnglandPuthoffB Swann 1981 7 Apr 0940 29025'N98030'WSan Antonio, Texas PuthoffB Harary 1981 8 Apr 0836 20P19'N1030101WChapala Lake, Mexico PuthoffB Swann 1981 8 Apr 0956 33391S 780581WRobinson Crusoe Is., Juan PuthoffB Unknown 1981 Fernandez Island 8 Apr 1116 51051'NOlo2l'WBlenheim Palace, Oxon EnglandPuthoffB Swann 1981 8 Apr 1119 60POO'N15~'OO'WCook Inlet, Alaska PuthoffB Swann 1981 9 Apr 0845 030021S37020'EMt. Kilimanjaro, Tanzania PuthoffC Humphrey 1981 ' I 9 Apr 0923 3822_'N11OP211WCanyonlands Park, Utah PuthoffB 1981 1 I I 1 0 Z Wn M Z SG1J Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 1W Approved For Release(2 (Eff"ML00788r001 300270001-9 00 The third problem arises in a training evaluation environment. The target management data shown in Table 2 displays only a small portion of the data that is available for each of the targets. Each target is classified as to the principal type of target with regard to large-scale geographical features (e.g.; a body of water, an island, or a mountain). For purposes of evaluation, it is necessary to know details of each of these major features. For example, one may want to determine how many targets had "muddy streams" as the particular form of "water." POWER+ was incapable of providing this capability. Thus, only a coarse target classification was possible. A "tree-structured" data base is one of the easiest solutions for this problem. A properly designed tree structure allows for easy modification and simple implementation of complex searches such as the one described above. It is not that tree-structured data-base systems do not exist. Rather, they tend to be large (require large main-frame memories) and require fast processors. Typically, these data bases include a large number of features that are not required in our application and unneces- sarily burden the speed/memory requirement of the system. so The deficiencies described here are not unique to POWER+. Each wq ime onE Tvwx beLT commercially available DBMS will have similar problems with regard to W trs Tile JJ our application. However, there are portions of our application that may have reduced requirements and can be solved with simple data-base manage- ment systems. ZA-no P 7 CONFIDENTIAL Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 Approved For Release(ZQ&WrD(EfffP 00788r001300270001-9 .* L IV CONCLUSIONS Depending upon the exact data-base design criteria, commercially available DBMS may possibly be applied to RV data and target management on a limited basis. Such systems are available for a combined hardware/ software cost of approximately $5,000. These systems have the added advantage of being "desk-top" size and are single user so that the entire system may be maintained within a secure environment. If we consider the security requirement of the general GRILL FLAME 15MOVE program and budget constraints,,we conclude that: ~ Full data-base management of RV data is possible, but must be developed in house. ~ A complete DBMS must be self-contained, flexible, and operate as a single-user on a micro-mini-computer. CONFIDENTIAL A P ?1E:TZ/1AR_ 1-0 Do TIMS Approved For Release 2000/08/07 : CIA-RDP96-00788r001300270001-9 Approved For ReleaECO"f8flD EWM*-00788rOO 1300270001-9 d V RECOmMFNDATIONS As a result of the demonstration applications, discussed in the body RIO of this report, and of a survey of available DBM systems, we recommend a threefold approach to implement computerized DBM of RV data, RV training, and psychoenergetic intelligence information. We recommend: (1) That a complete DBMS be purchased to facilitate training. This system would be exclusively used for training target management and can be completely maintained within the training program. (2) That a micro-mini-computer-based DBM system be developed that is tailored to the RV requirements to provide optimal evaluation and management of operational RV sessions. (3) That the requirements of such a system first be defined to initiate development of a psycho- energetic intelligence data-base system. These requirements would then aid in identifying the particular DBMS that is needed. C'Mr 40 40 DOOT TV\N "ME ZeTrm -nAwris ML kA\H 1"D DO 9 CONFIDENTIAL %4 _