From Raw Data to Intelligence Information
No matter what the setting or type of collection, gathered information must be pack-aged meaningfully before it can be used in the production of intelligence. Processing methods will vary depending on the form of the collected information and its intended use, but they include everything done to make the results of collection efforts usable by intelligence producers. Typically, "processing" applies to the techniques used by government intelligence services to transform raw data from special-source technical collection into intelligence information.37
While collectors collect "raw" information, certain [collection] disciplines involve a sort of pre-analysis in order to make the information "readable" to the average all-source analyst. For instance: imagery analysts "read-out" the basic information on the image; foreign language broadcasts must be literally translated by linguists and analyzed for linguistic "context"; electronic signals require sorting out to be intelligible to the uninitiated in that arcane art; agent reports also need literal translations and perhaps comments as to access, context, assumed or proven past veracity.38In the private sector, some processing activities are analogous to those of the government. Interpreting and annotating open-source information for a business intelligence ser-vice may include: marking locations of interest on a map or photograph, "translating" press releases or technical reports, transcribing the words of a speaker from video or audiotape into text, or drafting a detailed commentary from a personal interview.
Another term for processing, collation, encompasses many of the different operations that must be performed on collected information or data before further analysis and intelligence production can occur. More than merely physically manipulating information, collation organizes the information into a usable form, adding meaning where it was not evident in the original. Collation includes gathering, arranging, and annotating related information; drawing tentative conclusions about the relationship of "facts" to each other and their signiﬁcance; evaluating the accuracy and reliability of each item; grouping items into logical categories; critically examining the information source; and assessing the meaning and usefulness of the content for further analysis.
Collation reveals information gaps, guides further collection and analysis, and provides a framework for selecting and organizing additional information.39
Examples of collation include ﬁling documents, condensing information by categories or relationships, and employing electronic database programs to store, sort, and arrange large quantities of information or data in preconceived or self-generating patterns. Regardless of its form or setting, an effective collation method will have the following attributes:
- Be impersonal. It should not depend on the memory of one analyst; another person knowledgeable in the subject should be able to carry out the operation.
- Not become the "master" of the analyst or an end in itself.
- Be free of bias in integrating the information.
- Be receptive to new data without extensive alteration of the collating criterion.40
To prepare collected information for further use, one must evaluate its relevance and value to the speciﬁc problem at hand. An examination of the informations source and applicability to the intelligence issue can determine whether that information will be further employed in the intelligence production process. Three aspects to consider in evaluating the relevance of information sources are reliability, proximity, and appropriateness.
Reliability of a source is determined through an evaluation of its past performance; if the source proved accurate in the past, then a reasonable estimate of its likely accuracy in a given case can be made. A human sources own testimony of reliability may also be taken into account; qualiﬁers such as "certain," "believe," and "guess" indicate how sure the source is of the information being conveyed. However, if the source is completely untested, then evaluation of the information must be done solely on its own merits, independent of its origin.41
Proximity refers to the sources closeness to the information. The direct observer or participant in an event may gather and present evidence directly, but in the absence of such ﬁrsthand information, the analyst must rely on sources with varying degrees of proximity to the situation. A primary source passes direct knowledge of an event on to the analyst. A secondary source provides information twice removed from the original event; one observer informs another, who then relays the account to the analyst. Such regression of source proximity may continue indeﬁnitely, and naturally, the more numerous the steps between the information and the source, the greater the opportunity for error or distortion.42
Appropriateness of the source rests upon whether the source speaks from a position of authority on the speciﬁc issue in question. As no one person or institution is an expert on all matters, the sources individual capabilities and shortcomings affect the level of validity or reliability assigned to the information it provides regarding a given topic.43
The following examples illustrate the use of reliability, proximity, and appropriateness to evaluate a source.
The mail clerk at 3rd Army Headquarters told me that, according to the 1st Armored Division Supply Ofﬁcer, the Division is being deployed to Site Y in three days.The reliability of the mail clerk as a source (questionable), his proximity to the information (secondary), and the appropriateness of the Supply Ofﬁcer as a source on the fact of deployment (uncertain), make this information of little value to the intelligence production process.
A major national newspaper published an interview with the CEO of Big Company, quoting the CEOs announcement of a merger the company had just secretly concluded with Large Company.The reliability of a major national newspaper as a source (good), its proximity to the information (secondary), and the appropriateness of the CEO as the source of the merger announcement (high) make this information of high value to intelligence production.
Three aspects of the information itself have a bearing on its applicability to intelligence issues: plausibility, expectability, and support. Plausibility refers to whether the information is true under any circumstances or only under certain conditions, either known or possible. Expectability is assessed in the context of the analysts prior knowledge of the subject. Support for information exists when another piece of evidence corroborates it - either the same information from a different source, or different information that points to the same conclusion.44
For example, a source contends that the President of Country X recently died, but the death is being kept secret from all but a few members of his regime. Although unusual, this information is plausible, and even has precedent in history. The scenario may meet the expectability criterion, if the country or this particular regime is known to be extremely secretive and paranoid about being vulnerable to hostile internal or external takeover movements. Support for this information may come from the same source pro-viding details on the Presidents secret burial ceremony, or a different source, such as an actor who was hired to play the part of the President in a false Presidential address televised to the nation.
All these factors of source and content contribute to an initial assessment of the value of a particular piece of information to the intelligence production process. Those pieces that are judged to be useful may then undergo further scrutiny in light of customer needs, while items of questionable value may be rejected or set aside for further processing and comparison with other information. This initial selection of intelligence information sets the stage for intelligence analysis and production, as explained in the following Parts of the primer.
37 The Department of Defense deﬁnes intelligence information as "unprocessed data of every description which may be used in the production of intelligence." (Department of Defense, Joint Chiefs of Staff, Joint Pub 1-02, Dictionary of Military and Associated Terms (Washington, DC: 23 March 1994), 184.)
38 Dearth, "National Intelligence," 19.
39 R.H. Mathams, "The Intelligence Analysts Notebook," in Strategic Intelligence: Theory and Application, eds. Douglas H. Dearth and R. Thomas Goodden, 2d ed. (Washington, DC: Joint Military Intelligence Training Center, 1995), 85-86.
40 Mathams, 86.
41 adapted from Gary Harris, "Evaluating Intelligence Evidence," in A Handbook of Intelligence Analysis, ed. Ronald D. Garst, 2d ed. (Washington, DC: Defense Intelligence College, January 1989), 34-35. For an in-depth treatment of evidence evaluation techniques and factors, see David Schum, Evidence and Inference for the Intelligence Analyst, Vols. I and II (Lanham, MD: University Press of America, 1987).
42 Harris, 35.
43 adapted from Harris, 36.
44 adapted from Harris, 36-38