MS, Kent State University, 2018, College of Arts and Sciences / Department of Computer Science

Posting on the internet, including weblogs or social media, is one of the ways individuals seek for an outlet to express themselves or mental health concerns. For many mental health issues such as psychosis, the timing of detection and treatment is critical; short and long-term outcomes are better when individuals begin treatment close to the onset of psychosis. While the internet offers a positive medium for short term therapy, it is not a face to face therapy session, wherein a trained professional is better able to deduce the root of the problem. Many clinicians are adopting electronic communication to strengthen their therapeutic alliance with their patients. The drawback of psychiatry is that it lacks objectified tests for mental illnesses that would otherwise be present in medicine. Current neuroscience has yet not found genetic markers that can characterize individual mental illnesses. A thought disorder (ThD) which is a widely found symptom in people suffering from schizophrenia, is diagnosed from the level of coherence when the flow of ideas is muddled without word associations. A system that can explore the use of speech analysis for aiding in psychiatric diagnosis is highly desirable and would help early detection and effective treatment results. This thesis introduces a framework – Prediction Onset Prediction System (POPS) - to predict the onset of psychosis based on written language habits. A scrape of a multitude of individual comments is analyzed using a trained psychosis prediction module that is able to predict if an individual is psychotic (based on the semantics) using natural language processing, machine learning techniques and a customized corpus with terms consist with psychotic language tendencies created using speech analysis techniques. The effectiveness of the corpus and its implication in psychosis detection is explored.

Committee: Kambiz Ghazinour (Advisor) Subjects: Computer Science; Health; Mental Health; Psychology; Sociology; Teaching; Technology

Master of Arts, The Ohio State University, 2018, East Asian Languages and Literatures

The current study surveys social media data to identify regional lexical variants in modern written Chinese and to characterize their geographical distributional patterns. A large amount of geo-tagged linguistic data was obtained from a corpus containing 5.1 million messages posted on a Chinese micro-blogging website (Weibo). A list of lexical items obtained from the book "Lexicon of Chinese Dialects" was searched in the corpus to generate word counts by location. It was found that a portion of regional lexical variants from this book appeared in the written corpus. Closer examination of these variants revealed different patterns in their geographical distributions. This study also investigated the regional specificity of these lexical variants by calculating their cumulative frequencies across space, which led to different conclusions about their usage when compared with survey results found in previous literature. In order to find out if there are regional sub-varieties of written Chinese characterized by lexical variation, a machine learning algorithm (k-means) was trained on the word frequency data gathered from the corpus to cluster the locations based on their uses of lexical items most clearly signaling regional differences. The cluster analysis suggests the existence of three clusters, reflecting the north-south contrast in modern written Chinese that is associated with the linguistic history of China, as well as the strong influence of Cantonese in areas around the Guangdong province. Through the above-mentioned analyses, this study provideds some insights into the lexical norms of written Chinese in contemporary China. It also contributes to the development of methods for processing Chinese texts on computer.

Committee: Marjorie Chan (Advisor); Marie-Catherine de Marneffe (Committee Member) Subjects: Asian Studies; Language; Linguistics

MA, University of Cincinnati, 2017, Arts and Sciences: Psychology

Participants provided decisions in three-word recognition experiments designed to keep stimulus and response constant while simultaneously manipulating task demands: go-no-go naming, lexical decision, and semantic categorization. The aggregate response time distributions of these tasks were analyzed with a lognormal inverse power-law mixture distribution, the cocktail model (Holden & Rajaraman, 2012), to probe the influence of longer timescale demands on cognitive dynamics. Changes in task demands influenced the shape of the response time distributions, despite identical stimuli across the tasks. The outcomes are discussed in terms of the influence of temporally nested networks of performance. Constraints are aspects of physical, perceptual or cognitive systems that shape behavior in particular ways. For instance, linguistic development unfolds a long timescale of development. However, the influence of these constraints are expressed on much faster timescale activities, such as acts of individual word recognition and articulation. Fast timescale cognitive dynamics are influenced by a web of longer contextual and historical constraints (Van Orden, Hollis, & Wallot, 2012). The narrative explains how the characteristic shapes of response time distributions reveal these influences.

Committee: John Holden Ph.D. (Committee Chair); Anthony Chemero Ph.D. (Committee Member); Quintino Mano Ph.D. (Committee Member) Subjects: Cognitive Therapy