Depression is the leading cause of disability, often undiagnosed, and one of the most treatable mood disorders. As such, unobtrusively diagnosing depression is important. Many studies are starting to utilize machine learning for depression sensing from social media and Smartphone data to replace the survey instruments currently employed to screen for depression. In this study, we compare the ability of a privately versus a publicly available modality to screen for depression. Specifically, we leverage between two weeks and a year of text messages and tweets to predict scores from the Patient Health Questionnaire-9, a prevalent depression screening instrument. This is the first study to leverage the retrospectively-harvested crowd-sourced texts and tweets within the combined Moodable and EMU datasets. Our approach involves comprehensive feature engineering, feature selection, and machine learning. Our 245 features encompass word category frequencies, part of speech tag frequencies, sentiment, and volume. The best model is Logistic Regression built on the top ten features from two weeks of text data. This model achieves an average F1 score of 0.806, AUC of 0.832, and recall of 0.925. We discuss the implications of the selected features, temporal quantity of data, and modality.

Foodborne outbreaks are a serious but preventable threat to public health that often lead to illness, loss of life, significant economic loss, and the erosion of consumer confidence. Understanding how consumers respond when interacting with foods, as well as extracting information from posts on social media may provide new means of reducing the risks and curtailing the outbreaks. In recent years, Twitter has been employed as a new tool for identifying unreported foodborne illnesses. However, there is a huge gap between the identification of sporadic illnesses and the early detection of a potential outbreak. In this work, the dual-task BERTweet model was developed to identify unreported foodborne illnesses and extract foodborne-illness-related entities from Twitter. Unlike previous methods, our model leveraged the mutually beneficial relationships between the two tasks. The results showed that the F1-score of relevance prediction was 0.87, and the F1-score of entity extraction was 0.61. Key elements such as time, location, and food detected from sentences indicating foodborne illnesses were used to analyze potential foodborne outbreaks in massive historical tweets. A case study on tweets indicating foodborne illnesses showed that the discovered trend is consistent with the true outbreaks that occurred during the same period.

Foodborne diseases and outbreaks are significant threats to public health, resulting in millions of illnesses and deaths worldwide each year. Traditional foodborne disease surveillance systems rely on data from healthcare facilities, laboratories, and government agencies to monitor and control outbreaks. Recently, there is a growing recognition of the potential value of incorporating social media data into surveillance systems. This paper explores the use of social media data as an alternative surveillance tool for foodborne diseases by collecting large-scale Twitter data, building food safety data storage models, and developing a novel frontend foodborne illness surveillance system. Descriptive and predictive analyses of the collected data were conducted in comparison with ground truth data reported by the U.S. Centers for Disease Control and Prevention (CDC). The results indicate that the most implicated food categories and the distributions from both Twitter and the CDC were similar. The system developed with Twitter data could complement traditional foodborne disease surveillance systems by providing near-real-time information on foodborne illnesses, implicated foods, symptoms, locations, and other information critical for detecting a potential foodborne outbreak.

Given that depression is one of the most prevalent mental illnesses, developing effective and unobtrusive diagnosis tools is of great importance. Recent work that screens for depression with text messages leverage models relying on lexical category features. Given the colloquial nature of text messages, the performance of these models may be limited by formal lexicons. We thus propose a strategy to automatically construct alternative lexicons that contain more relevant and colloquial terms. Specifically, we generate 36 lexicons from fiction, forum, and news corpuses. These lexicons are then used to extract lexical category features from the text messages. We utilize machine learning models to compare the depression screening capabilities of these lexical category features. Out of our 36 constructed lexicons, 14 achieved statistically significantly higher average F1 scores over the pre-existing formal lexicon and basic bag-of-words approach. In comparison to the pre-existing lexicon, our best performing lexicon increased the average F1 scores by 10%. We thus confirm our hypothesis that less formal lexicons can improve the performance of classification models that screen for depression with text messages. By providing our automatically constructed lexicons, we aid future machine learning research that leverages less formal text.

Depression, a prevalent mental health disorder with severe health and economic consequences, can be costly and difficult to detect. To alleviate this burden, recent research has been exploring the depression screening capabilities of deep learning (DL) models trained on videos of clinical interviews conducted by a virtual agent. Such DL models need to consider the challenges of modality representation, alignment, and fusion as well as small sample sizes. To address them, we propose WavFace, a multimodal deep learning model that inputs audio and temporal facial features. WavFace adds an encoder-transformer layer over pre-trained models to improve the unimodal representation. It also applies an explicit alignment method for both modalities and then uses sequential and spatial self-attention over the alignment. Finally, WavFace fuses the sequential and spatial self-attentions among the two modality embeddings, inspired by how mental health professionals simultaneously observe visual and vocal presentation during clinical interviews. By leveraging sequential and spatial self-attention, WavFace outperforms pre-trained unimodal and multimodal models from the literature. With a single interview question, WaveFace screened for depression with a balanced accuracy of 0.81. This presents a valuable modeling approach for audio-visual mental health screening.