Discovering Shifts to Suicidal Ideation from Mental Health Content in Social Media

Choudhury, Munmun De; Kıcıman, Emre; Dredze, Mark; Coppersmith, Glen; Kumar, Mrinal

doi:10.1145/2858036.2858207

Cited by 577 publications

(447 citation statements)

References 63 publications

(72 reference statements)

Supporting

Mentioning

436

Contrasting

Unclassified

Order By: Relevance

“…Additionally, online linguistic patterns match previous findings regarding language use of depressed individuals [16]. More recently, De Choudhury et al [10] have shown that linguistic features are important predictors in identifying individuals transitioning from mental discourse on social media to suicidal ideation. The authors showed a number of markers characterizing these shifts including social engagement, manifestation of hopelessness, anxiety and impulsiveness based on a small subset of Reddit posts.…”

Section: Literature Reviewsupporting

confidence: 80%

See 1 more Smart Citation

Detection of suicide-related posts in Twitter data streams

Vioulès¹,

Moulahi²,

Azé³

et al. 2018

IBM J. Res. & Dev.

120

View full text Add to dashboard Cite

Suicidal ideation detection in online social networks is an emerging research area with major challenges. Recent research has shown that the publicly available information spread across social media platforms holds valuable indicators to effectively detecting individuals with suicidal intentions. The key challenge of suicide prevention is understanding and detecting the complex risk factors and warning signs that may precipitate the event. In this paper, we present a new approach that uses the social media platform Twitter to quantify suicide-warning signs for individuals and to detect posts containing suicide-related content. The main originality of this approach is the automatic identification of sudden changes in a user's online behavior. To detect such changes, we combine natural language processing techniques to aggregate behavioral and textual features and pass these features through a martingale framework, which is widely used for change detection in data streams. Experiments show that our text-scoring approach effectively captures warning signs in text compared to traditional machine learning classifiers. Additionally, the application of the martingale framework highlights changes in online behavior and shows promise for detecting behavioral changes in at-risk individuals.

show abstract

Section: Literature Reviewsupporting

confidence: 80%

“…Social computing research has focused on this topic in recent years [6,9,10]. However, few initiatives have been concerned with the real time detection of suicidal ideation on Twitter.…”

Section: Introductionmentioning

confidence: 99%

Detection of suicide-related posts in Twitter data streams

Vioulès¹,

Moulahi²,

Azé³

et al. 2018

IBM J. Res. & Dev.

120

View full text Add to dashboard Cite

show abstract

“…Monitoring such venues can thus offer a window into behaviors which are otherwise difficult to study. Specifically, in the case of suicides, social media has been used to detect suicidality [20], and search engine logs were utilized to analyze suicides in general [21], [22] and the Werther Effect ("copycat suicides") in particular [21].…”

Section: Internet Data As a Source For Health Informationmentioning

confidence: 99%

How Search Engine Data Enhance the Understanding of Determinants of Suicide in India and Inform Prevention: Observational Study (Preprint)

Adler¹,

Cattuto²,

Kalimeri³

et al. 2018

Preprint

View full text Add to dashboard Cite

BackgroundIndia is home to 20% of the world's suicide deaths. In India, and around the world, young people are especially at risk of suicide. While statistics regarding suicide in India are distressingly high, data and cultural issues likely contribute to a widespread underreporting of the problem. Social stigma and only recent de-criminalization of suicide are but two factors hampering official agencies' collection and reporting of suicide rates. ObjectiveAs the product of a data collaborative -the cross-sector exchange of data to create new public value -this paper leverages private-sector search engine data toward gaining a fuller, more accurate picture of the suicide issue among young people in India. By combining official statistics on suicide with data generated through search queries, this paper seeks to: 1) add an additional layer of information to more accurately represent the magnitude of the problem; 2) determine whether search query data can serve as an effective proxy for factors contributing to suicide that are not represented in traditional datasets; and 3) consider how data collaboratives built on search query data could inform future suicide prevention efforts in India and beyond. MethodsWe combined official statistics on demographic information with data generated through search queries from Bing to predict suicide rates per state in India as reported by the National Crimes Record Bureau of India.We have extracted English language queries on five topics ("suicide", "depression", "hanging", "pesticide", "poison"). For each query, we recorded the time and date of the query, the state in India from which the user made the query, and the text of the query.We have then collected data on demographic information at state level in India, including: Urbanization, Growth Rate, Sex Ratio, Internet Penetration, Population. We have modeled the suicide rate per state as a function of the queries on each of the 5 topics considered as linear independent variables. We also built a second model by integrating the demographic information on Urbanization, Growth Rate, Sex Ratio, Internet Penetration and Population, all considered as additional linear independent variables in the model. ResultsResults of the first model fit (R 2 ) when predicting the suicide rates from the fraction of queries in each of the 5 topics, as well as the fraction of all suicide methods, show a correlation of about 0.5. The correlation increases significantly with the removal of even 3 outliers, and improves slightly when 5 outliers are removed. In all cases, statistically significant correlation is reached, but the best correlation is obtained for suicide methods (hanging, pesticide, and poison), and only to a lesser extent for depression.Results for the second model fit using both query data and demographic data show that for all categories, if no outliers are removed, demographic data predict suicide rates better than query data. However, when 3 outliers are removed, query data about pesticides or poisons improves the model ov...

show abstract

“…Considering the frequent use of emotion language in status updates that relate to current experiences [13,14], for a large proportion of the population social media can provide unobtrusive access to time-sensitive and ecologically valid samples of expressed emotion [2,[41][42][43]. While observations of emotion variability and instability are yet to be applied to social media as a means of automatically screening for individuals at risk of depression, depression change has been demonstrated to be visible through changes in status update posting activity and content [2,4,10,44,45].…”

Section: Social Media and Emotion Dynamicsmentioning

confidence: 99%