ChatGPT, a machine learning model, has been found to write college-level essays and hold human-like conversations. A study led by the University of Buffalo trained ChatGPT to recognize locations in social media posts from disaster victims. The goal is to help first responders find those in distress during natural disasters by processing social media data for emergency services. The study showed that ChatGPT’s language skills could help first responders reach victims more quickly and save lives. The researchers hope to use AI systems to automatically process social media data for emergency services.
The study showed that ChatGPT could recognize location descriptions more accurately than conventional tools and was able to quickly learn and adapt to the problem. However, providing good prompts is crucial for accurate results. The researchers plan to further research to use ChatGPT’s extracted location descriptions to geolocate victims and filter out irrelevant or false posts.
The researchers hope that their efforts will simplify the use of AI technologies for emergency managers, allowing them to focus on saving lives. The study was supported by the National Science Foundation.
>>Join our Facebook Group be part of community. <<