Identifying emergency stages in Facebook posts of police departments with convolutional and recurrent neural networks and support vector machines
This work addresses a domain-specific problem for emergency responders and the public by improving classification accuracy for social media posts, though it is incremental as it applies existing methods to a new dataset.
The study tackled the problem of classifying Facebook posts from US police departments into emergency stages (preparedness, response, recovery, and general engagement) to aid in automated processing and timely emergency response, achieving an F1 score of 0.839 with an RNN using custom-trained word2vec features.
Classification of social media posts in emergency response is an important practical problem: accurate classification can help automate processing of such messages and help other responders and the public react to emergencies in a timely fashion. This research focused on classifying Facebook messages of US police departments. Randomly selected 5,000 messages were used to train classifiers that distinguished between four categories of messages: emergency preparedness, response and recovery, as well as general engagement messages. Features were represented with bag-of-words and word2vec, and models were constructed using support vector machines (SVMs) and convolutional (CNNs) and recurrent neural networks (RNNs). The best performing classifier was an RNN with a custom-trained word2vec model to represent features, which achieved the F1 measure of 0.839.