Abstract

This paper describes the approachwe take to social media analysis, combining the exploration of the opinion oftext and centered on the recognition of entities and events. We examine aparticular use case, which is to help archivists select materials for inclusionin a social media archive to preserve community memories, moving towardsstructured preservation around semantic categories. The textual approach weadopt is rule-based and relies on a number of sub-components, taking intoaccount issues inherent in social media such as noisy non-grammatical text, useof insulting words, short language popularly called as SLANG, and so on. Inorder to resolve the ambiguity and provide additional contextual information.We propose two major innovations in this work: first, the novel combination oftools for extracting texts and multimedia opinions; And second, the adaptationof NLP tools for the analysis of opinion specific to the problems of socialmedia.

Keywords:- Security,k-NN classifier, cloud databases, encryption