Cognitive network neighborhoods quantify feelings expressed in suicide notes and Reddit mental health communities

Joseph, Simmi Marina; Citraro, Salvatore; Morini, Virginia; Rossetti, Giulio; Stella, Massimo

doi:10.1016/j.physa.2022.128336

Cited by 13 publications

(8 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While several studies used NLP to examine linguistic features in social media posts made by subjects ostensibly suffering from psychosis (Bae et al, 2021;Birnbaum et al, 2017;Dinev et al, 2022;Ernala et al, 2017;Joseph et al, 2023;Kim et al, 2020;Lejeune et al, 2022;Lyons et al, 2021;McManus et al, 2015;Mitchell et al, 2015;Zomick et al, 2019), this is the first study exploring the coherence of social media posts as an indicator of disorganization symptoms. A pronounced drawback of the present study is the lack of information on the users.…”

Section: Discussionmentioning

confidence: 99%

“…Posts from social media forums such as reddit can be analyzed efficiently by means of natural language processing (NLP) (Zhang et al, 2022) to understand processes related to substance abuse (Graves et al, 2022), anxiety and depression (Boettcher, 2021;Liu et al, 2023;Shen et al, 2017), suicidality (Alambo et al, 2019) or bipolar disorder (BD) (Sekulić et al, 2018). Social media data has also been utilized in the study of psychosis and schizophrenia (SZ) (Bae et al, 2021;Birnbaum et al, 2017;Dinev et al, 2022;Ernala et al, 2017;Joseph et al, 2023;Kim et al, 2020;Lejeune et al, 2022;Lyons et al, 2021;McManus et al, 2015;Mitchell et al, 2015;Zomick et al, 2019). Posts made by individuals suspected of suffering from psychosis were used to evaluate the nature and impact of COVID-19 on subjective well-being (Lyons et al, 2021) or to study their sleep behavior (Dinev et al, 2022).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

(PREPRINT) Plank & Zlomuzica, 2024, coherence in social media posts

Plank,

Zlomuzica

2024

Preprint

View full text Add to dashboard Cite

The extraction of linguistic markers from social media posts which are indicative of the onset and course of mental disorders offers great potential to mental healthcare. In the present study, we extracted over one million posts from the popular social media platform reddit to analyze speech coherence which reflects formal thought disorder and is a characteristic feature of schizophrenia and associated psychotic disorders. Natural language processing (NLP) models were used to perform an automated quantification of speech coherence. While controlling for the emotional tone and subjectivity, we demonstrated that posts made in forums geared towards disorders with a higher degree of psychotic and formal thought disorder symptoms tend to show a lower level of coherence. The lowest coherence scores were detected in posts made in forums on dissociative identity disorder, schizophrenia and bipolar disorder followed by a medium level of coherence in posts submitted to a forum on autism and post-traumatic stress disorder. In contrast, a relatively high level of coherence was detected in posts related to obsessive compulsive disorder, depression, anxiety and attention-deficit hyperactivity disorder. Our findings provide promising first evidence for a possible utility of NLP-based coherence analyses for the early detection and prevention of psychosis on the basis of posts gathered from publicly available social media data. This opens new avenues for large-scale prevention programs aimed at high-risk populations.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

(PREPRINT) Plank & Zlomuzica, 2024, coherence in social media posts

Plank,

Zlomuzica

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Similarly, the cyan community containing "er" that refers to Elliot Rodgers (a supposed Incel hero who committed mass violence before committing suicide (Broyd et al 2022)), "rope," "roping," "ascending," "neeting" and "ldaring" can also be categorized as nuanced Incel expressions based on our manual assessment. The presence of terms such as "er," "ering" (in cyan) and "erfuel" (red) in the figure also raises concerns about the connection of self-harm and mass violence among Incels as indicated by (Brunt et al 2021), (Joseph et al 2021) and as described in (EU 2021), the promotion to "go ER" means to commit murder-suicide as a way to inflict punishment on society for an individual's inceldom.…”

Section: Demystifying Expressions Of Self-harm (Rq1)mentioning

confidence: 99%

From Isolation to Desolation: Investigating Self-Harm Discussions in Incel Communities

Ali,

Zannettou

2024

ICWSM

View full text Add to dashboard Cite

Incel communities have recently attracted the public's interest mainly due to their high degree of extreme views and involvement in real-world violence. A common theme in Incel communities is self-harm discussions. Despite this, beyond small-scale qualitative analyses of self-harm discussions in Incel communities, we lack a large-scale quantitative understanding of how Incels discuss self-harm and how it differs from mainstream communities. In this work, we aim to demystify self-harm discussions in Incel communities using a data-driven approach and understand how Incels differentiate from mainstream communities. We use a dataset of 6.4M posts from 18 Incel subreddits and 2.4M posts from an Incel forum, as well as 5.8M posts from two mainstream subreddits discussing mental health. Using word embedding approaches, temporal analyses, topic modeling, and qualitative analysis, we shed light on self-harm discussions in Incel and mainstream communities and their evolution over time. We find substantial differences in the language related to self-harm deployed among the communities; we find that Incels use niche terms related to self-harm, which is not the case in mainstream communities. We observe that over time, language related to self-harm evolves considerably more among Incels than in mainstream communities. Also, we observe that negative perception of their physical appearance is the most recurrent theme in self-harm conversations for Incels, which does not feature in mainstream communities. Finally, by analyzing social factors, we find that Substance abuse is the most closely associated social factor to self-harm in Incel and mainstream communities and that Physical Appearance, over time, is becoming increasingly closely related to self-harm discussions in Incel communities.

show abstract

“…Conversely, when extracting keywords from long texts, each text is modeled as a word cooccurrence network [8,19,38].…”

Section: Network Creationmentioning

confidence: 99%

Using citation networks to evaluate the impact of text length on the identification of relevant concepts

Tohalino¹,

Silva²,

Amancio³

2023

Preprint

View full text Add to dashboard Cite

The identification of the most significant concepts in unstructured data is of critical importance in various practical applications. Despite the large number of methods that have been put forth to extract the main topics of texts, a limited number of studies have probed the impact of the text size on the performance of keyword extraction (KE) methods. In this study, we adopted a networkbased approach to evaluate whether keywords extracted from paper abstracts are compatible with keywords extracted from full papers. We employed a community detection method to identify groups of related papers in citation networks. These paper clusters were then employed to extract keywords from abstracts. Our results indicate that while the various community detection methods employed in our KE approach yielded similar levels of accuracy, a correlation analysis revealed that these methods produced distinct keyword lists for each abstract. We also observed that all considered approaches, however, reach low values of accuracy. Surprisingly, text clustering approaches outperformed all citation-based methods. The findings suggest that using different sources of information to extract keywords can lead to significant differences in performance, and this effect can play an important role in applications relying upon the identification of relevant concepts.

show abstract

Cognitive network neighborhoods quantify feelings expressed in suicide notes and Reddit mental health communities

Cited by 13 publications

References 40 publications

(PREPRINT) Plank & Zlomuzica, 2024, coherence in social media posts

(PREPRINT) Plank & Zlomuzica, 2024, coherence in social media posts

From Isolation to Desolation: Investigating Self-Harm Discussions in Incel Communities

Using citation networks to evaluate the impact of text length on the identification of relevant concepts

Contact Info

Product

Resources

About

Cognitive network neighborhoods quantify feelings expressed in suicide notes and Reddit mental health communities

Cited by 13 publications

References 40 publications

(PREPRINT) Plank &amp; Zlomuzica, 2024, coherence in social media posts

(PREPRINT) Plank &amp; Zlomuzica, 2024, coherence in social media posts

From Isolation to Desolation: Investigating Self-Harm Discussions in Incel Communities

Using citation networks to evaluate the impact of text length on the identification of relevant concepts

Contact Info

Product

Resources

About

(PREPRINT) Plank & Zlomuzica, 2024, coherence in social media posts

(PREPRINT) Plank & Zlomuzica, 2024, coherence in social media posts