2 June 2020 in Suicide and Life-Threatening Behavior
Objective: With early identification and intervention, many suicidal deaths are preventable. Tools that include machine learning methods have been able to identify suicidal language. This paper examines the persistence of this suicidal language up to 30 days after discharge from care.Method: In a multi‐center study, 253 subjects were enrolled into either suicidal or control cohorts. Their responses to standardized instruments and interviews were analyzed using machine learning algorithms. Subjects were re‐interviewed approximately 30 days later, and their language was compared to the original language to determine the presence of suicidal ideation.Results: The results show that language characteristics used to classify suicidality at the initial encounter are still present in the speech 30 days later (AUC = 89% (95% CI: 85–95%), p < .0001) and that algorithms trained on the second interviews could also identify the subjects that produced the first interviews (AUC = 85% (95% CI: 81–90%), p < .0001).Conclusions: This approach explores the stability of suicidal language. When using advanced computational methods, the results show that a patient’s language is similar 30 days after first captured, while responses to standard measures change. This can be useful when developing methods that identify the data‐based phenotype of a subject.
7 August 2017 in BMC Bioinformatics
Probabilistic assessments of clinical care are essential for quality care. Yet, machine learning, which supports this care process has been limited to categorical results. To maximize its usefulness, it is important to find novel approaches that calibrate the ML output with a likelihood scale. Current state-of-the-art calibration methods are generally accurate and applicable to many ML models, but improved granularity and accuracy of such methods would increase the information available for clinical decision making. This novel non-parametric Bayesian approach is demonstrated on a variety of data sets, including simulated classifier outputs, biomedical data sets from the University of California, Irvine (UCI) Machine Learning Repository, and a clinical data set built to determine suicide risk from the language of emergency department patients. The method is first demonstrated on support-vector machine (SVM) models, which generally produce well-behaved, well understood scores. The method produces calibrations that are comparable to the state-of-the-art Bayesian Binning in Quantiles (BBQ) method when the SVM models are able to effectively separate cases and controls. However, as the SVM models’ ability to discriminate classes decreases, our approach yields more granular and dynamic calibrated probabilities comparing to the BBQ method. Improvements in granularity and range are even more dramatic when the discrimination between the classes is artificially degraded by replacing the SVM model with an ad hoc k-means classifier. The method allows both clinicians and patients to have a more nuanced view of the output of an ML model, allowing better decision making. The method is demonstrated on simulated data, various biomedical data sets and a clinical data set, to which diverse ML methods are applied. Trivially extending the method to (non-ML) clinical scores is also discussed.
30 April 2022 in Bilişim Teknolojileri Dergisi
Social media data can provide a general idea of people’s response towards the COVID-19 outbreak and its reflections, but it cannot be as objective as the news articles as a source of information. They are valuable sources of data for natural language processing research as they can reveal various paradigms about different phenomena related to the pandemic. This study uses a news collection spanning nine months from 2019 to 2020, containing COVID-19 related articles from various organizations around the world. The investigation conducted on the collection aims at revealing the repercussions of the pandemic at multiple levels. The first investigation discloses the most mentioned problems covered during the pandemic using statistics. Meanwhile, the second investigation utilizes machine learning to determine the most prevalent topics present within the articles to provide a better picture of the pandemic-induced issues. The results show that the economy was among the most prevalent problems. The third investigation constructs lexical networks from the articles, and reveals how every problem is related through nodes and weighted connections. The findings exhibit the need for more research using machine learning and natural language processing techniques on similar data collections to unveil the full repercussions of the pandemic.