Structuring Open Source Information to Support Intelligence Analysis

Open source information on the internet can contribute significantly to intelligence assessments. Unfortunately, this information is mostly unstructured text, and varies widely in accuracy, focus, and level of impartiality. Though such unstructured information may be organized for convenient review in information portals, it is difficult to visualize, link, consolidate, summarize, compare, fuse, or otherwise analyze unstructured free text information. Once information is structured, it may be examined and processed using a broad spectrum of powerful commercial tools. This paper reviews the benefits from structuring open source information, describes the criteria that structured information should meet, and outlines a structuring methodology that EBR is not employing for a major U.S. government organization. This methodology employs an advanced service-based architecture for integrating collection and analysis tools, commercial text extraction software, formal ontologies, and tools for operator review and refinement of the machine-structured text. 1. An Example of Structuring Structuring creates well-defined data records from unstructured material. The following example, based on material that EBR is processing for one of it’s clients, illustrates the end result of structuring, and serves to illustrate the important properties of structured information. In this case, the client is interested in understanding commercial relationships among companies within the telecommunications industry. Among their interests are various technical and marketing associations. In this example, there are two sources, PWID 1 (Published Works #1) and PWID 2. Each source contains multiple references to the nature of the association, with each reference contributing some information. PWID 1 Indonesia's Telkom awards Ericsson with broadband contract