Search results

This paper aims to examine tweet posts regarding Typhoon Washi to contend the usefulness of social media and big data as an aid of post-disaster management. Through topic modelling and content analysis, this study examines the priorities of the victims expressed in Twitter and how the priorities changed over a year.

Social media, particularly Twitter, was where the data gathered. Using big data technology, the gathered data were processed and analysed according to the objectives of the study. Topic modelling was used in clustering words from different topics. Clustered words were then used for content analysis in determining the needs of the victims. Word frequency count was also used in determining what words were repeatedly used during the course period. To validate the gathered data online, government documents were requested and concerned government agencies were also interviewed.

Findings of this study argue that housing and relief goods have been the top priorities of the victims. Victims are seeking relief goods, especially when they are in evacuation centres. Also, the lack of legal basis hinders government officials from integrating social media information unto policymaking.

This study only reports Twitter posts containing keywords either, Sendong, SendongPH, Washi or TyphoonWashi. The keywords were determined based on the words that trended after Typhoon Washi struck.

For social media and big data to be adoptable and efficacious, supporting and facilitating conditions are necessary. Structural, technical and financial support, as well as legal framework, should be in place. Maintaining and sustaining positive attitude towards it should be taken care of.

Although many studies have been conducted on the usefulness of social media in times of disaster, many of these focused on the use of social media as medium that can efficiently spread information, and little has been done on how the government can use both social media and big data in collecting and analysing the needs of the victims. This study fills those gaps in social big data literature.