testBug

10:30:00

1512988200

October 25, 2017

Semantic Wide and Deep Learning for Detecting Crisis-Information Categories on Social Media

Gregoire Burel, Hassan Saif and Harith Alani

Research

clock_eventOctober 25, 2017, 10:30.
house Lehár 1-3
download Download paper (preprint)

Abstract

When crises hit, many flog to social media to share or consume information related to the event. Social media posts during crises tend to provide valuable reports on affected people, donation offers, help requests, advice provision, etc. Automatically identifying the category of information (e.g., reports on affected individuals, donations and volunteers) contained in these posts is vital for their efficient handling and consumption by effected communities and concerned organisations. In this paper, we introduce Sem-CNN; a wide and deep Convolutional Neural Network (CNN) model designed for identifying the category of information contained in crisis-related social media content. Unlike previous models, which mainly rely on the lexical representations of words in the text, the proposed model integrates an additional layer of semantics that represents the named entities in the text, into a wide and deep CNN network. Results show that the Sem-CNN model consistently outperforms the baselines which consist of statistical and non-semantic deep learning models.