Social media today is an important source of information about various events happening around the world. Among various social networking platforms, microtext based ones such as Twitter are of special interest as they are also a rich source of real-time events. In this thesis, our goal is to study the effectiveness of using Twitter as a social sensor for obtaining real-time information on road traffic conditions. Specifically, we focus on: i) identifying tweets that contain traffic event related information, ii) classify such tweets into six main groups of accident, fire, road construction, police activities, weather and others, iii) extract fine-grained location information about the traffic incident by analyzing tweet text. Our experimental results show that using Twitter as a social sensor for obtaining rich information about traffic events is indeed a promising approach. We show that we can correctly detect traffic related tweets with an accuracy of 81%. Moreover, the accuracy of correctly classifying traffic related tweets into one of the six categories is 97%. Lastly, our experimental results show that using only geo-tags of tweets is not sufficient for fine-grained localization of traffic incidents due to two reasons: i) a vast majority of traffic related tweets do not contain geo-tags, and ii) the location mentioned in the tweet text and the geo-tag of a tweet do not always agree. Such observations prove that fine-grained localization of traffic incidents from tweet must also include analysis of the tweet text using Natural Language Processing techniques.

Master of Science / Department of Computer Science / Doina Caragea / Huge amounts of data that are generated on social media during emergency situations are regarded as troves of critical information. The use of supervised machine learning techniques in the early stages of a disaster is challenged by the lack of labeled data for that particular disaster. Furthermore, supervised models trained on labeled data from a prior disaster may not produce accurate results. To address these challenges, domain adaptation approaches, which learn models for predicting the target, by using unlabeled data from the target disaster in addition to labeled data from prior source disasters, can be used. However, the resulting models can still be affected by the variance between the target domain and the source domain. In this context, we propose to use a hybrid feature-instance adaptation approach based on matrix factorization and the k-nearest neighbors algorithm, respectively. The proposed hybrid adaptation approach is used to select a subset of the source disaster data that is representative of the target disaster. The selected subset is subsequently used to learn accurate supervised or domain adaptation Naïve Bayes classifiers for the target disaster. In other words, this study focuses on transforming the existing source data to bring it closer to the target data, thus overcoming the domain variance which may prevent effective transfer of information from source to target. A combination of selective and transformative methods are used on instances and features, respectively. We show experimentally that the proposed approaches are effective in transferring information from source to target. Furthermore, we provide insights with respect to what types and combinations of selections/transformations result in more accurate models for the target.

How do people discuss mental health on social media? Can we train a computer program to recognize differences between discussions of depression and other topics? Can an algorithm predict that someone is depressed from their tweets alone? In this project, we collect tweets referencing “depression” and “depressed” over a seven year period, and train word embeddings to characterize linguistic structures within the corpus. We find that neural word embeddings capture the contextual differences between “depressed” and “healthy” language. We also looked at how context around words may have changed over time to get deeper understanding of contextual shifts in the word usage. Finally, we trained a deep learning network on a much smaller collection of tweets authored by individuals formally diagnosed with depression. The best performing model for the prediction task is Convolutional LSTM (CNN-LSTM) model with a F-score of 69% on test data. The results suggest social media could serve as a valuable screening tool for mental health.