Using Deep Learning Techniques to Detect Traffic Information in Social Media Texts

Preview this chapter:

Using Deep Learning Techniques to Detect Traffic Information in Social Media Texts, Page 1 of 1

< Previous page | Next page > /docserver/preview/fulltext/9789815305395/chapter-40-1.gif

Given the real-time and pervasive nature of social media, data mining for traffic-related insights is a newly emerging area of study. In this work, we discuss the challenge of mining social media for traffic-generating microblogs on Sina Weibo, the Chinese equivalent of Twitter. It is recast as a classification issue in short texts for machine learning. In the first step, we use a dataset of three billion microblogs and the continuous bag-of-word model to learn word embedding representations. Word embedding, as opposed to the standard one-hot vector representation of words, has been shown to be useful in natural language processing tasks because of its ability to capture semantic similarity between words. Then, we suggest feeding the learned word embeddings into convolutional neural networks (CNNs), long short-term memory (LSTM) models, and their combination LSTM-CNN to extract traffic-related microblogs. We evaluate the suggested techniques against state-of-the-art methods such as the support vector machine (SVM) model using a bag of n-gram features, the SVM model using word vector features, and the multi-layer perceptron model using word vector features. The proposed deep learning methods are shown to be useful in experiments.