A Method Extracting Task-related Information from Social Media based on Structured Domain Knowledge

Autor(es)

Link, Daniel ; Horita, Flavio E. A. ; de Albuquerque, Joao Porto ; Hellingrath, Bernd ; Ghasemivandhonaryar, Shabdiz ; Assoc Informat Syst

• Tipo de documento: Artigo Científico
• Fonte: AMCIS 2015 PROCEEDINGS; v. N/A, p. 16-pg., 2015-01-01.

Resumo

Social media platforms have come into the focus of research as sources of information about the unfolding situation in disaster contexts. Incorporating information from social media into decision-making is still difficult though. One reason may be that the prevalent approach to data analysis works bottom-up, which has several limitations. In this paper, we adopt a top-down approach by means of a novel keyword-based method for identifying potentially relevant information in social media data based on structured knowledge of activities undertaken in a domain. The application of the method to the context of humanitarian logistics using four social media datasets shows its capability to identify potentially relevant information via reference tasks and to match identified information with decision-makers' activities. In addition, we offer a first set of domain-specific keywords to identify information related to infrastructure and resources in humanitarian logistics. (AU)

• Processo FAPESP: 08/58161-1 - Assessment of impacts and vulnerability to climate change in Brazil and strategies for adaptation option
• Beneficiário: Jose Antonio Marengo Orsini
• Modalidade de apoio: Auxílio à Pesquisa - Programa de Pesquisa sobre Mudanças Climáticas Globais - Temático
• Processo FAPESP: 12/18675-1 - AGORA: uma arquitetura geoespacial aberta e colaborativa para construir resiliência contra desastres e impactos das mudanças climáticas em comunidades vulneráveis
• Beneficiário: João Porto de Albuquerque
• Modalidade de apoio: Bolsas no Exterior - Pesquisa