Using HyperLogLog to Prevent Data Retention in Social Media Streaming Data Analytics

Social media data are widely used to gain insights about social incidents, whether on a local or global scale. Within the process of analyzing and evaluating data, it is common practice download store locally. Considerations privacy protection users often neglected thereby. However, protecting when dealing with personal demanded by laws ethics. In this paper, we introduce method using cardinality estimator HyperLogLog. Based an exemplary disaster management scenario, show that can be analyzed counting occurrences posts, without becoming in possession actual raw data. For analyses like these, based occurrences, estimation suffices task. Thus, risk abuse, loss, public exposure mitigated preserved. The ability do unions intersections multiple datasets further encourages use technology. We provide proof-of-concept implementation for our introduced method, provided Twitter API.