Context-Aware Information Retrieval

Customer U.S. Air Force and NASA
Users Analysts and general Web users
Need In an era of rapid change, smart organizations worldwide have recognized the need to establish technical intelligence capabilities of various sorts. The central objective of these efforts is to identify emerging technologies that can fulfill recognized shortcomings in manufacturing processes or a company’s product line, or otherwise reshape the competitive landscape. The Air Force’s Information Warfare Solution Analysis Integrated Product Team (SAIPT) is a particularly robust example of such a capability wherein requirements identified in the course of Mission Needs Analysis (MNA) are used to propel the search for valuable technologies and process improvements. But, given a dynamic set of hundreds of requirements and a large number of potential solutions hidden within vast numbers of research proposals, whitepapers, product descriptions, and PowerPoint presentations, the task of matching requirements to solutions can be a daunting task.As has been discovered by analysts of all sorts, the application of general purpose information retrieval and filtering tools to such complex tasks is complicated by a number of factors including: query imprecision, terminology mismatch, and information overload.

Generally, users of search engines must engage in a lengthy process of query writing, result review, and query refinement – attempting various combinations of search terms in the hopes of finding the information or solution they seek. While techniques such as automated query expansion can help mitigate the term-mismatch problem, trial and error still play a central role. Aware addresses this problem through the use of a novel ‘multi-search’ strategy in which it generates a number of qualitatively different queries by drawing search terms from a broader context.In its most recent incarnation we have incorporated a ‘glassbox’ approach to semantic search in which users may build up task vocabularies and define (and redefine) semantic concepts in an ad hoc fashion. By providing the user with control over the definition of concepts (instead of a centralized authority), users are better able to target their searches.


Our system also increases practical search precision by improving the ability of users to identify documents of high utility even when those results do not appear at the top of the rankings. We accomplish this through a unique faceted browsing approach and real-time subject/genre labeling.

Aware was initially launched in 2005 as a desktop application.

Status The InfoTracker prototype has been developed for Windows XP and heavily tested. InfoTracker’s built-in viewer includes document converters that cover PDF, HTML, and Microsoft Office. Users may also utilize InfoTracker through a Microsoft Word toolbar. We continue to extend the capabilities of InfoTracker and are actively seeking Beta users. Please see the InfoTracker Product Page for additional information.
Related Applications Applications of the Aware technology span enterprise search, intelligence analysis, and general Web research.