This project designs and implements tools for frictionless information systems that work behind the applications that people use in their work and everyday lives, automatically finding and presenting contextually relevant, useful, and interesting information. To be relevant to a user's information needs in a given context, information must add something new: it must be similar in certain respects to the context, but dissimilar in certain other respects. The technology utilized in this project explicitly and specifically searches or filters for interestingly different information, dissimilar in systematic ways that reflect information attributes relevant to a user's goals.
The project develops useful information attributes and dimensions of dissimilarity as well as technologies that can utilize these attributes and dimensions to find useful and relevant information based on the user?s current context. The particular dimensions that matter in a given context will depend upon the user, task, and domain at hand. To address this variability, the project is aimed at a general architecture and a set of tools and libraries that will support the rapid development of frictionless information systems for a wide range of application settings.
The results are expected to lead to development of information systems that can provide people with information that is contextually relevant, genuinely interesting, and diverse, resulting in broader and deeper understanding. The results will be disseminated in academic venues as well as through public release of prototype systems via the project Web site (http://infolab.northwestern.edu/). The project is well integrated with educational activities. Relevance Engine platform will be utilized in class projects for advanced undergraduates and graduate students, and some of these systems may be aimed at educational settings. Students will be involved in development of corpora for testing prototypes, informal assessment of prototypes using web resources, and gain valuable experience in research, design and implementation, and validation.
