|
Capabilities for a Future National Intelligence Enterprise
Gary W. Fuller
Booz Allen Hamilton
he National Intelligence Enterprise (NIE) will be a unified, coordinated, and effective entity that continuously performs analysis of data from multiple sources, extracts, integrates, and correlates information, and produces and distributes, at the appropriate security classification, the most timely, most relevant, and most accurate intelligence content for its clients, customers, and consumers around the world. The NIE, though comprised of 15 different U.S. Government agencies and departments (or portions thereof) will provide a variety of data, information and knowledge products and services to satisfy a broad range of needs by the U.S. Policy, Defense, Homeland Security and Law Enforcement communities, as well as allies and coalition partners.
Ten Critical Capabilities for the NIE
For the NIE to be both effective and efficient in the future it will have ten critical capabilities across the Enterprise. These capabilities are fundamental to being able to function in a unified and coordinated manner while recognizing that 15 different organizations constitute the Enterprise.
Clients, Customers, and Consumers
The NIE will have capabilities to provide responsive products and services to clients, customers and consumers. Clients are persons with whom the Enterprise has a one-on-one relationship like the President, National Security Adviser, Secretary of State, or Secretary of Defense. Customers are persons that have a continuing relationship with the Enterprise either online, by telephone, or through collaborative communications. Customers have profiles and typically have subscriptions to various types of content. Consumers are persons who regularly use intelligence content, but tend to do so in such a way that their preferences are generally unknown to the enterprise. Most Internet website users could be viewed as consumers of their information content, but in a somewhat anonymous manner. Therefore, the NIE must be knowledgeable of their Clients and Customers by name and must have a profile of their intelligence needs so that they may adequately serve them. Internet websites frequently make use of Customer Relationship Management (CRM) tools to do this in addition to personal contact. The most important consideration here is that Clients and Customers are people—not organizations.
Responsive Products and Services
The NIE will have the capability to produce intelligence products and services that are designed to satisfy the needs of each customer segment, i.e., U.S. Policy, Defense, Homeland Security, and Law Enforcement. These product and service offerings will, in most cases, be both customizable and personalizable by an end-user. Customization allows an end-user to select from a number of pre-defined options, while personalization allows the end-user to configure the product (or service) exactly the way they want based on profiles and preferences. Intelligence products and services will be continuously improved—and sometimes retired—based on customer needs. But, to be perceived as very responsive to customers, not every product or service should be based on a formal customer-initiated requirements process, but rather some of them will be conceptualized and developed internally within the Enterprise based on customersÕ preferences and an analysis of CRM data. Some will be created based on broad customer segment understanding, applying the "if-we-build-it-they-will-come" model. In this way, the Enterprise will reach out to the customer base well beyond their formally written and validated requirements.
An example of a new service offering for all customers will be an Intelligence News Network or ITN.
Intelligence News Network
The NIE will have a capability to produce and distribute recorded and real-time intelligence news on a variety of regions, countries, transnational issues, and topics for users within the Enterprise as well as clients, customers, and consumers around the world. This activity will not be centralized, but rather each of the member organizations of the NIE will operate one or more intelligence news centers depending on their missions and specialties. The content will be streamed over IP networks as well as be available in high definition through such satellite systems as those provided by Global Broadcast Services. (GBS) and others. The goal of ITN will be to provide continuous coverage of the world and its activities and events from an intelligence perspective. Content will include intelligence highlights as well as in-depth analysis. Although some of the content will originate from studios, it seems appropriate that analysts will also be capable of broadcasting directly from their workstation. These capabilities will contribute to the overall requirements placed on the infrastructure of the Enterprise.
Online Data, Information & Knowledge Service—National Intelligence Online
There is a difference between the Internet and all of its World Wide Web sites and America Online (AOL). Companies, government agencies, educational institutions and individuals put websites online for their own "information sharing" reasons—whether to market a product, to sell a product, or to merely publish useful, interesting, or entertaining content. America Online on the other hand is a managed set of content that a user views as "an information service." AOL also provides connectivity and access to the Internet as well as e-mail and other Internet Service Provider (ISP) functions, but these are ancillary to being an aggregator and provider of information to millions of subscribers. To deliver their content, AOL first has to gain access to it from the original provider, as well as categorize it so that it can be listed in a directory. For some content it also controls the user interface and presentation of the information.
The NIE will develop and operate a National Intelligence Online (NIO) information service for its clients, customers and consumers. It will provided integrated access to all intelligence content, regardless of producer—and it will do so on an on-demand basis. Thus, NIO will be the gateway to all intelligence data, information, and knowledge—as well as the wisdom of human analysts. No single undertaking will revolutionize the intelligence business more than having such integrated access to source data, in-work analyses, and finished intelligence then this. NIO will be managed similar to AOL, therefore it will provide a business framework for sharing beyond connectivity and content by adding the value proposition of aggregating and presenting the most timely, most relevant, and most accurate intelligence content to each customer segment at an appropriate security classification. The management of NIO will have to secure adequate connectivity and satellite links to ensure that the content is readily accessible around the world. NIO will also ensure that the content is accessible through a variety of devices from Top Secret/SCI inside-the-Beltway high-end workstations to police car screens on the road. This may very well be superior to 15 different organizations doing this independently.
Global Intelligence Knowledge Base
The NIE will have a knowledge base that is continuously being updated by the intelligence analysis and production process. A Global Intelligence Knowledge Base (GIKB) will capture, in a structured manner, all current intelligence about regions, countries, transnational issues, topics, and targets in such a way that the analysts across the NIE can continuously update it depending on their specialty and authority.
One way to look at such a concept is a hybrid of the content of the World Factbook and all National Intelligence Estimates. One way or the other, the NIE will have such an online resource for use by its analysts. An important factor to note here is that the GIKB will be dynamically changing as a result of ongoing multi-source analysis. Furthermore, it is equally as important to note that the GIKB will not a directory of finished intelligence reports or briefings, but rather it will have content that has been extracted from these items that provides the latest and most accurate assessment of capabilities and intentions.
The GIKB content will be drawn from existing fileservers and databases (owned by NIE member organizations) so that duplication of content is all but eliminated and the accuracy and currency of the intelligence extracted and entered into the base is the best. Multiple views of the knowledge base will be possible so that an analyst is capable of retrieving a current assessment from any number of perspectives. A GIS or geospatial view is critical so that the existence of knowledge, as a function of geography, can be portrayed. In other words, not only are natural and man-made features displayed (as in any conventional GIS) but geo-referenced icons will also be portrayed to represent the type, accuracy and currency of knowledge that the NIE has for a given facility or installation.
The GIKB will form the foundation of knowledge that the NIE has on essentially any subject, and all intelligence analysts across the Enterprise will be charged with maintaining and updating its content. Rather than beginning an analysis task with a series of searches of finished intelligence reports, an analyst will enter a query into the GIKB to retrieve the multi-dimensional assessment of the situation. He/she will then make whatever updates necessary to ensure that the entries are current and accurate.
In addition to the examples above, it is worth comparing the online Encyclopedia known as Wikipedia to the GIKB. In its English version, there are nearly 500,000 articles that are continuously being updated by people around the world. Every day hundreds of thousands of visitors from around the world make tens of thousands of edits and they create thousands of new articles. Each article is heavily laced with hyperlinks to other articles, making it a highly interactive research experience. Its currency on many topics is usually intra-day.
Here is the first paragraph of the article on Iraq:
The Republic of Iraq is a Middle Eastern country in southwestern Asia encompassing the ancient region of Mesopotamia. It shares borders with Kuwait and Saudi-Arabia to the south, Jordan to the west, Syria to the northwest, Turkey to the north, and Iran to the east. Its current leadership was put in place on June 28, 2004, following a March 2003 invasion led by British and American forces which drove Saddam Hussein and the Ba'ath Party from power.
As you can see, hyperlinks are provided to allow you to move effortlessly to other related articles, terminology and dates. The NIE could readily initiate development of a GIKB using Wikipedia technology and the expertise of its thousands of analysts from across its 15 member organizations.
Discovery Services
NIE Clients, Customers and Consumers will have comprehensive discovery services that will enable them to easily browse through a hierarchical list of topics, search for files, and query the content of shared databases across the Enterprise. These discovery services will be implemented at the Enterprise level rather than by means of unique, agency-specific browse, search and query tools or websites. Discovery services will include capabilities to find documents and files, to understand the meaning of text (in multiple languages) and to query databases.
Finding Documents and Files
Although the most obvious content is that of finished intelligence, capabilities will also be provided to discover ongoing analysis, as well as to determine the availability of raw data from all relevant sources.
Once the Enterprise posts all—or at least most—content on servers for discovery and access by customers, the issue will then be how to qualify sources and screen incredibly long lists of search results. To do this effectively, adequate metadata will be applied to content so as to support "advanced search" capabilities. But this is sometimes a tedious process, and is subject to the idiosyncrasies of constructing a good Boolean search specification. Therefore, a set of "rules" that can be applied to how search results are presented will be available. Capabilities will be provided to group results by provider, source type, date of publication, most recent update, and other parameters that can make the process of sorting through results more efficient and effective.
In addition to linear lists, albeit selected and sorted as suggested above, graphical representations that show the interrelationship of a number of variables will also be provided. The critical factor here is providing a capability to a user to ascertain what sources he/she should use without having to go to the individual site and open a webpage or document, first.
Understanding Text
When the body (or corpus) of documents is large, say in the thousands to tens of thousands and beyond, it becomes unwieldy to deal with search results that are based on indexed documents. Keyword search, though powerful, can produce way too many results, and only produces results based on the "pattern" of words searched, i.e., individual words, groups of words, or exact phrases. But, although useful, this technology as practiced by Google, Yahoo and many other search engines, it cannot understand the meaning of the words, and therefore cannot automatically extract information from data, nor can it understand relationships, networks, and transactions.
If the objective is to get answers (or possible answers), search must not result in the raw records from the database, but rather with extracted information such as relationships between people, countries, actions, and events. This powerful kind of textual understanding is possible, but it requires complete dictionaries of each language to be searched as well as a complex set of semantic rules that allow the software to truly understand what has been written. Of course the power of such software is that it can "read" millions of pages to identify potential relationships and present them to a human analyst for validation. This kind of screening is nearly impossible, even when a large number of human analysts are applied to the task.
Text understanding services therefore will be provided by the NIE to both internal (analyst) users as well as external customers so that the most timely, most relevant, and most accurate intelligence can be either produced or discovered.
An example here is the content produced and distributed by the Foreign Broadcast Information Service (FBIS). FBIS provides translations of foreign news sources—whether print or broadcast—and posts it for intelligence users on Intelink. The translations, especially of newscasts on radio and TV, are frequently very current. This content is also made available to the public through subscription services on the Internet such as the World News Connection by Dialog. But this service provides only traditional Boolean search capabilities using simple, advanced, and regions and topic search. The result of a search is a lengthy search results list.
If text-understanding technology were applied, using a product like Insightful CorporationÕs InFact, it would be possible to automatically generate alerts and warnings from the meaning of the translated newscasts and newspaper articles. An example of this technology can be viewed online by visiting the InFact 9/11 Commission Report Index. This index was automatically generated to extract instances of people, places and organizations that occurred within the body of the 9/11 Report. Each indexed item has one or more hyperlinks associated with it that take you to the body of the report to the exact sentence in which the term occurred. The InFact product, through its understanding of the English language and its syntax, could automatically identify those sentences that dealt with people, places, and organizations.
Querying Databases
As almost everyone knows, the real "jewels" of information are frequently found deep in databases—not in documents, files, or web pages. It is for this reason that the NIE will create a directory of Enterprise databases and characterize their content. It will be possible, based on knowing the kind of information needed, to target one or several databases and search them without having specific access to or training on the data system that created the database content. It will also be possible to perform a "federated" search that spans dozens—if not hundreds—of databases to identify information necessary to support intelligence analysis and production efforts within the Enterprise, or to ascertain the situation for any location around the world.
The Enterprise will have a single gateway or portal to database queries in addition to those provided by the NIE member organizations that "own" the data systems that hold the information. By implementing the data, information and knowledge service model, it will be possible for the Enterprise to offer a family of database query services that will meet the needs of internal Enterprise users as well as clients, customers, and consumers.
Global Geospatial Information Network
Although GIS technology and geospatial information have come to the forefront in recent years (due to emphasis by CIA and NGA) it is still not as pervasive as it should be. Every NIE analyst will be able to access an incredible amount of geospatial data and information from their workstation. They will be capable of accessing an intelligence version of a Global Geospatial Information Network (GGIN). This network will link all geospatial content—regardless of provider—so that all types of intelligence can be co-registered and displayed over maps and imagery. This is not to say that there are not initiatives in this arena, but the scale of such activities has not been global. There are hundreds (if not thousands) of organizations that deal with maps, imagery, feature data, and the significance of installations, emitters, and other facilities but they are not linked. The NIE will be proactive in establishing an intelligence-oriented GGIN so that if something can be displayed in the context of a set of earth coordinates—it is! GGIN, then is a Internet technology-based global extent GIS where the themes and layers are from multiple providers around the globe. NGA is clearly an obvious supplier of foundation data and intensified data to support particular missions, but there are so many other providers both within the Enterprise, and as mission and production partners that they will all become part of a common, standards-based network. Geospatial presentation of content will become ubiquitous across the NIE and its customer base. To accomplish this, organizations like the OpenGIS Consortium will continue to encourage industry to build interoperable GIS databases that exploit Internet technology so that a Global Geospatial Information Network can become a reality—and be accessible from every analystÕs desktop.
One factor that will contribute to the success of GGIN is an initiative to create and maintain a Seamless Imagery Database.
Seamless Imagery Database
GIS databases typically contain both raster and vector data, however vector data is what gives a GIS its unique "signature." Vector data can be used to represent both natural and man-made features and is therefore quite appropriate for various theme "layers" such as boundaries, topographic elevation contours and the transportation network, e.g., roads. Imagery, on the other hand, is raster data. It has the benefit of accurately portraying all of the earthÕs features in shades of gray or color so that every nuance can be seen. Features are not "symbolized" as they are on a map or in a vector layer of a GIS. The features however are not labeled. Raster imagery, when combined with vector data, provide a powerful combination.
Unfortunately, due to the way imagery collectors operate, image frames are still the primary mode of collection, processing, storage, and distribution. After collecting satellite imagery of the earth for more than 45 years, the U.S. Government does not have a seamless, high-resolution imagery database of the earthÕs landmasses. Mosaics of several images are routinely created, but the business process continues to be oriented toward delivering discrete images.
The NIE therefore will have a Seamless Imagery Database (SID) in lieu of imagery (frame) fileservers as its primary resource. All collectors—satellite, aircraft, and UAV, will continuously update the database. Rather than exploiting images as they arrive, imagery will be accessed from the database as research on various intelligence problems is conducted. Image understanding technologies such as change detection and automatic target recognition will be applied to images as they are "deposited" into the database to identify changes that should immediately be screened by human analysts. The database will be an integral part of the GGIN and will have multiple sub-layers that represent the earthÕs surface and features in different bands as well as at different qualities.
The SID will not only be accessible as an element of the GGIN, but will also be a logical part of the Global Intelligence Knowledge Base.
System-to-System Data and Information
Products and Services produced by the NIE will extend beyond consumption by human customers, clients, and consumers. The world has become a system-of-systems that is enabled by Internet (IP) technology and standards. Though weapon systems are a good example and current user of data produced by NGA, there are and will be many more system-to-system data and information transfers. The motivation for this kind of connectivity and distribution is not to remove the human analyst from the loop, but rather to ensure that the timeliness of delivery is as necessary to support a customerÕs mission. Electronic alerts and notifications as well as the transfer of results from a family of intelligence information services will be moved between computers as fast as possible.
System-to-system transfers of product and service content will be tailored to the customerÕs needs from the outset. Thus when supporting Defense, Homeland Security, and Law Enforcement, it will be critical to not only have the content be created and distributed at the correct security classification—especially when it is unclassified—but also, it will be capable of being received on the appropriate device—whether a mainframe computer, a police car display, or a hand-held computer.
Thus, this type of product or service offering will be conceptualized, from the beginning, as to the environment of the user, the type of connectivity, security considerations, the amount of data, the type of receiving device and the quality of the display. When customers are operating with short decision cycles, that are minutes (and sometimes seconds), it is expected that the NIE will be capable of supporting the necessary timelines. To do this, autonomous system-to-system data and information transfers (and broadcasts) will be mandatory.
Expert Consulting
In addition to data, information and knowledge content and electronically transferred system-to-system content, there is a real opportunity for the NIE to operate a "consulting business" staffed by human intelligence analysts. This business will provide "advisory services" to NIE customers to either help them interpret and use other intelligence products or to address problems that are not strictly tied to Enterprise information content products.
This is not to say that this kind of work does not already go on, for example in the Homeland Security community, but rather that the consulting services business will be institutionalized and expanded—especially as databases and automation increase.
Of course to be efficient and effective at providing consulting services, NIE analysts will not only have access to databases, but customers will be capable of understanding all of the specialties and areas of expertise for which analyst-consultants can provide support. Some of the same databases and tools that are required for collaboration services among analysts can be applied to the interface between analysts and customers.
Collecting and Obtaining New Data
The analysis and production "division" of the Enterprise needs to be able to deal with the various collection domains, i.e., suppliers, in new and innovative ways that can change the business model from one that is collection driven to one that is focused on customers and satisfying their needs and mission requirements with innovative intelligence products and services. When new data is needed, the process for requesting it should be simple, reliable, and responsive. Data content from suppliers should always be available in databases accessible on an on-demand basis
Providing Current Status Data
The NIE will provide integrated current status data to its internal users and its external customers. Status data will encompass the entire intelligence business and its processes such that authorized individuals will be capable of determining collection and processing related status as well as intelligence analysis and production status. In the former case, a user will be able to know what collection (by INT) is planned, scheduled, and completed either on a browse or search basis, or by specific requests that had been submitted by the user. Because all collectors and collection organizations will be considered suppliers to the Enterprise, an integrated report (that is current to the time of the query) will be prepared and presented to the user. In the latter case, users will be able to determine the status of ongoing intelligence analysis and intelligence production across the Enterprise. Both kinds of status data will be available in real time on an interactive basis. Supply chain and workflow management applications will provide the data from each member organization in a federated manner such that a very organized presentation of status is available on demand. It will also be possible to know the status of content using such constructs as region, country, transnational issue, or topic. Additionally it will be possible to perform arbitrary geographic searches of any location to determine the most recent intelligence available, including the ability to delimit the source, the type of intelligence, the confidence level, etc.
Subscriptions, Profiles and Preferences
The NIE will provide users with the capability to establish subscriptions to content it produces as well as related information. Categories of news, including planned and scheduled activity notifications will also be available by subscription. As in the case of other services, a user will have a single entry point to establish a subscription for anything produced by the Enterprise, irrespective of the producing organization. Once a subscription for content, or a notification or alert is entered, the subscriber will receive a confirmation e-mail or webpage. When a relationship exists between the NIE and a commercial or international production partner, a user will be able to enter a subscription for that content, too.
Profiles and preferences will complement subscription services so that when content is delivered or when interactions with the NIE are conducted, the correct customization and personalization is automatically applied.
Traditional, Transformational, or Revolutionary Approaches
The NIE will be a unified, coordinated, and effective entity that continuously performs analysis of data from multiple sources, extracts, integrates, and correlates information, and produces and distributes, at the appropriate security classification, the most timely, most relevant, and most accurate intelligence content for its clients, customers, and consumers around the world. It will not be a traditional enterprise in terms of the approaches it takes to satisfying customersÕ needs, because to do that would ignore the world situation, best commercial practices, and the capabilities provided by current and emerging technologies. Though the world is talking about transformations, there are even aspects to this process that may be too formal, too structured, and too common—maybe even too traditional. For the NIE to be truly successful, a revolution is necessary. This revolution should apply to an analysis of where weÕve been, where we are now, and where we want the Enterprise to go. It is human nature to be conservative, to be cautious, to be risk averse. People tend to stay on the straight and narrow, in other words, most people are traditional thinkers. However, if revolutionary thinking is applied to formulating the NIE, then producers and users of intelligence can be certain that we will have at least achieved a transformation in the business of intelligence—even after it has been "softened" by the traditionalists.
Gary W. Fuller
Booz Allen Hamilton
703 902 7022
7 March 2005
Version 0.2
|
|||