The Emergence of Communities and Their Leaders on Twitter Following an Extreme Event

摘要: Twitter is presently utilized as a channel of communication and information dissemination. At present, government non-government emergency management organizations utilize to disseminate relevant information. However, these have limited ability evaluate the in order discover patterns, key players, messages that are being propagated through regarding event. More importantly there general lack knowledge who individuals or warning information, provide confirmations an event associated actions, urge others take action. This paper presents results analysis two events—2011 Japan Tsunami 2012 Hurricane Sandy. These insight into understanding human behavior, collectively part virtual communities on individually leaders members those communities. Specifically, their behavior evaluated terms obtaining propagating seeking additional confirmations, taking prescribed The will employ methodology shows how Natural Language Processing (NLP) Social Network Analysis (SNA) can be integrated results. allows extract actionable messages, construct network, find leaders, determine behaviors community leaders. Moreover, identifies specific roles Such include dispensing unique/new providing communities, urging them show agencies had participation during 2011 compared extensive users both events was consistent with issuance (i.e. warnings). findings suggest higher cohesion among than Sandy displayed agreement required protective action if some were other action). Additionally, differentiation leadership demonstrated stronger presence official sources roles.