Hot Set Identification for Social Network Applications

摘要: Several operations of Web-based applications are optimized with respect to the set resources that will receive majority requests in near future, namely hot set. Unfortunately, existing algorithms for identification do not work well emerging social network applications, characterized by quite novel features traditional Web: highly interactive user accesses, upload and download operations, short lifespan resources, interactions among members online communities. We propose evaluate innovative combinations predictive models social-aware solutions Experimental results demonstrate some considered improve accuracy up 30% if compared models, they guarantee stable robust even context high variability.