Incremental Mining of Information Interest for Adaptive Scanning of Business Environments through the Internet


Rey-Long Liu



Businesses often hierarchically organize their internal and environmental information of interest (IOI) into folders (or categories). Such personalized hierarchical folders may not only facilitate the management of IOI, but also reflect the interest of each individual business. A folder corresponds to an interest type. The interest is relatively long-term when compared with one-shot queries. For such interest, environmental scanning through the Internet (ESI) should be a continuous job directed by the specifications of the interest. The specifications should be both precise and comprehensible in order to make ESI more cost-effective and controllable. However, expressing such specifications are quite difficult for the business, since each interest type is implicitly and collectively defined by the content (i.e. documents) of the corresponding folder, which may also evolve over time. In this project, based on our previous experiences in information need identification, text mining, and ESI, we plan to develop an incremental text mining technique to identify the business’s current interest by mining the business's information folders, making ESI more adaptive to the business’s evolving interest. The specification mined for each interest type specifies the context of the interest type in suitable form (e.g. conjunction-of-disjunctions form), which is easy for business users to comprehend and refine. It helps the scanner to comprehensively start from proper seed sites and focus on those sites that are more likely to provide the information really of the business’s interest. The business may thus maintain her folders to constantly get IOI without paying much attention to the difficult tasks of interest specification and seed identification.

Keywords: context of interest, precise interest specifications, comprehensible interest specifications, incremental text mining, adaptive environmental scanning


     Back to Research Project