Enhancing Crisis Management because of Deep Learning, Big Data and Parallel Computing Environment: Survey

The broad dissemination and scope of social media enables individuals to exchange information in realtime. This active involvement of societies plays a major role in reducing disaster risk and alleviating at-risk populations. While any operation needs accurate information in crisis management to enable a rapid response to decrease the potential loss of life. Timely retrieval of information from regions of an area affected by disaster is a demanding task. A catastrophe relief and response method's effectiveness depends largely on a prompt and accurate crisis assessment. This knowledge is primarily collected on site by first responders and can be updated later on. Several technics have been builts to automate this need through the extraction and analysis of appropriate content from social media. These approaches are not, however, well incorporated into the mechanism of relief. For more advancement, it would be important to reveal them. The era of big data and analytics is giving crisis management new possibilities (DM). Big data is fundamentally transforming humanitarian operations and crisis management due to its ability to simulate, evaluate, and forecast disasters. However, the related literature is complex and fragmented, necessitating a study in order to determine its progression. Big data and its uses for disaster mitigation have been the subject of many publications. This thesis explores big data in DM based on a systematic literature review to current key contributions, holes, threats, and future research agenda. The results are presented in terms of yearly dissemination, key articles, and-most cited papers in the review. The results also reveal a classifiable pattern.