Potential use of text classification tools as signatures of suicidal behavior: A proof-of-concept study using Virginia Woolf’s personal writings

Abstract

BackgroundThe present study analyzes the feasibility of text classification to predict individual suicidal behavior. Entries from Virginia Woolf’s diaries and letters were used to assess whether a text classification algorithm could identify written patterns associated with suicide.MethodsThis is a text classification study. We compared 46 text entries from the two months before Virginia Woolf’s suicide with 54 texts randomly selected from Virginia Woolf’s work during other periods of her life. Letters and diaries were included, while books, novels, short stories, and article fragments were excluded. The data was analyzed using a Naïve-Bayes machine-learning algorithm.ResultsThe model showed a balanced accuracy of 80.45%, sensitivity of 69%, and specificity of 91%. The Kappa statistic was 0.6, which means a good agreement, and the p-value of the model was 0.003. The area under the ROC curve (AUC) was 0.80. In other words, the model exhibited good performance when used for classifying Virginia Woolf’s diaries and letters.DiscussionThe present study showed the feasibility of a machine-learning model coupled with text to identify individual written patterns associated with suicidal behavior. Our text signature was able to identify the period of two months preceding suicide with a high accuracy. This technique may be applied to subjects with psychiatric disorders by means of data captured from social media, e-mail, among others. The algorithm may then predict a specific outcome and enable early intervention by clinicians.

Machine learning and big data analytics in bipolar disorder: A position paper from the International Society for Bipolar Disorders Big Data Task Force

“…It is also worth mentioning that a recent text classification study used letters and diaries of Virginia Woolf to identify written patterns associated with suicide. 65 Authors found an AUC of 0.80 and a balanced accuracy of 80.45% by using Naïve-Bayes machine-learning algorithm.…”

“…Other studies also predicted suicidality by using machine learning coupled with a combined genomic and clinical risk assessment approach and built models with an AUC of 0.98 and 0.82 in patients with BD. It is also worth mentioning that a recent text classification study used letters and diaries of Virginia Woolf to identify written patterns associated with suicide . Authors found an AUC of 0.80 and a balanced accuracy of 80.45% by using Naïve‐Bayes machine‐learning algorithm.…”

The utility of artificial intelligence in suicide risk prediction and the management of suicidal behaviors

“…Beyond content analysis, word valence further indicates risk, with positively valenced words associated with a 30% reduced likelihood of suicide (McCoy et al, 2016). Moreover, the application of ML algorithms to patient writing patterns has been shown to predict a suicide attempt up to 2 months in advance, with an 80% accuracy rate (de Avila Berni et al, 2018); this may enable early identification of suicidality if applied to data sources that can be captured in real time such as social media and other personal text-based communications. NLP has also been applied to acoustic analysis of audio conversations to discriminate suicidal and non-suicidal individuals based on spoken language differences (Pestian et al, 2016).…”

Linguistic Features and Psychological States: The Case of Virginia Woolf

“…Meanwhile, Eichstaedt et al (2018) built a model with some linguistic predictors to identify depressed Facebook users, with a higher prediction accuracy (AUC = 0.72). In addition, de Ávila Berni et al (2018) developed a model to identify texts proposed by suicidal individuals based on the Naïve-Bayes machine-learning algorithm. The model achieved a higher performance, with an accuracy of 80%.…”