Smart web miner - extending web browser with mining framework based on user behavior & web-of-thing patterns for web personalization

As Web is proliferating so is the user behavior and the browsing pattern changing rapidly making it difficult & challenging task to make prediction´s based on the occurrence patterns from the analysis of the device Web logs, in a way ensuring that the patterns are not obsolete. Through this paper we address this problem by proposing a novel way of enabling Web engines with device intelligent FW which can extract the contextual user´s behavior patterns in real time through device soft sensors, WoT Sensors and browsing/Web-apps usage logs. These data are mined on device to generate prediction rules & also identify Interesting Web zones & frequent reoccurring patterns, discovering very long sequential patterns .This approach is further enriched with a mechanism to intelligently inform/alert users about a regular co-occurrence event(s), contextually controlling the WoT devices and application recommendations. We also propose a novel way to sync the different context data by different users (friends/family) making more collaborative contextual applications possible. This approach not only helps in diluting the impact of obsolescence of the frequent occurring patterns but also provide an enriched set of rules by which many innovative use cases for Web engine platform can be envisioned.