Suicidal Ideation Detection Using Machine Learning

Naga Prabhakar Ejaru, Ramya Sree Gangana, Karuna Kuruba,

Pallavi Uppara and Pavan Sai Chinduluru

Department of CSE, Srinivasa Ramanujan Institute of Technology, Anantapuramu -515701, Andhra Pradesh, India

Keywords: Self‑Destructive Contemplations, Early Recognizable Proof, Social Checking, Mental Assessment,

Intercession Techniques, Psychological Wellness, Risk Appraisal, Medical Services Experts.

Abstract: Recognizing self-destructive ideation is a basic area of examination that expects to distinguish people who

might be in danger of self-damage or self-destruction. Early ID is fundamental for further developing

intercession and treatment results, eventually diminishing the probability of deadly outcomes. This

exploration centers around making a successful structure for perceiving self-destructive considerations

utilizing different strategies, like social and mental examination. By following key markers like changes in

correspondence, profound prosperity, and social separation, this system plans to anticipate self-destructive

ideation and empower convenient mediations. The review inspects the utilization of clinical assessments,

mental polls, and example acknowledgment strategies to recognize people in danger. Moreover, it investigates

how medical services experts can coordinate these devices into their clinical practices to upgrade the precision

of appraisals and intercessions. The general objective is to create a harmless, dependable, and open location

framework that can be utilized in different conditions, including clinics, emotional wellness places, and local

area-based drives. This preventive methodology plans to lessen the disgrace related with looking for help for

self-destructive considerations and to raise worldwide mindfulness about emotional wellness challenges.

1 INTRODUCTION

Self-destructive ideation, or the thought of taking

one's life, has arisen as a developing worldwide

emotional wellness issue. Early distinguishing proof

and mediation in self-destructive contemplations can

fundamentally further develop results, lessening the

gamble of extreme outcomes like demise. This study

centers around fostering a structure pointed toward

distinguishing self-destructive ideation by joining

techniques like conduct examination, mental

evaluations, and information driven forecast models.

Recognizing self-destructive contemplations early is

vital, as it considers opportune intercession before the

condition grows into hurtful activities. Nonetheless,

perceiving self-destructive ideation has customarily

been provoking for medical services suppliers

because of its intricate nature, the shame

encompassing it, and its generally expected covered

up or implicit appearance. People encountering self-

destructive considerations may not transparently

express their expectations, making it harder for

clinicians to distinguish advance notice signs. In this

way, an effective recognition framework should

depend on numerous information sources, including

both verbal and non-verbal prompts, personal conduct

standards, and profound disturbances. The essential

objective of this examination is to make a dependable,

open, and harmless device that helps medical care

experts in recognizing people in danger for self-

destruction. By following key conduct markers like

changes in correspondence, profound prosperity,

social confinement, and emotional wellness decline,

this framework can foresee self-destructive ideation.

A critical part of this structure includes mental

assessments that survey an individual's psychological

state, close by social information that features

indications of self-hurt propensities. Through

normalized risk evaluations and mental surveys,

medical care suppliers can acquire a more profound

comprehension of a person's emotional wellness,

empowering them to answer as quickly as possible to

forestall self-destruction. Also, this exploration

investigates the capability of man-made brainpower

and AI to distinguish designs and foresee self-

destructive ideation. By dissecting assorted

information, including web-based entertainment

posts, instant messages, and clinical appraisals, AI

Ejaru, N. P., Gangana, R. S., Kuruba, K., Uppara, P. and Chinduluru, P. S.

Suicidal Ideation Detection Using Machine Learning.

DOI: 10.5220/0013892300004919

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies (ICRDICCT‘25 2025) - Volume 3, pages

93-99

ISBN: 978-989-758-777-1

calculations can distinguish cautioning signs and

banner people needing further assessment. A

fundamental piece of the review is exploring how

medical services experts can incorporate these

apparatuses into their clinical practices. Considering

the delicate idea of tending to self- destructive

ideation, it is basic for clinicians to approach exact,

modern devices to help their appraisals. Routine

psychological well-being screenings could be

incorporated into standard clinical assessments to

assist with recognizing people in danger. Besides,

making a cooperative model including medical

services suppliers, emotional wellness experts, and

local area associations could cultivate viable and

facilitated intercession endeavors. By giving clinical

experts these devices, they can offer directing,

emotional well-being treatment, and references to

trained professionals, guaranteeing that those in

danger get suitable consideration. Also, a significant

part of this approach is pointed toward lessening the

disgrace connected with psychological well-being

and self-destructive ideation. Numerous people try

not to look for help because of dread of judgment or

misinterpretations about dysfunctional behavior,

which frequently defers intercession. A framework

that joins conduct and mental investigation could

assist with normalizing emotional wellness care and

urge people to look for help before their condition

declines. This examination likewise advocates for the

production of a versatile and versatile self-destructive

ideation recognition framework that can be applied

across different conditions, including medical clinics,

centres, and local area outreach programs. Such a

framework could give nonstop checking to high-

gamble with people, cautioning guardians and

medical services suppliers when intercession is

required. The drawn-out objective is to foster an all-

inclusive instrument for distinguishing self-

destructive contemplations early, no matter what the

setting. This drive points not exclusively to save lives

yet additionally to work on worldwide consciousness

of emotional well-being difficulties and lessen the

shame related with self-destruction. The objective is

to develop a climate were looking for psychological

wellness support is viewed as ordinary, fundamental,

and energized, enabling people to connect for help

unafraid of judgment. All in all, early identification

of self-destructive ideation is basic for forestalling

death toll and improving psychological well-being

results. The joining of social checking, mental

assessments, and AI models can shape an extensive

structure for recognizing those in danger. By

outfitting medical care suppliers with these devices,

this examination looks to empower proactive

mediation before pointless ways of behaving grab

hold. Eventually, the objective is to make a humane,

compelling, and open self-destruction counteraction

strategy that helps people in recapturing command

over their lives while advancing a more extensive

comprehension of psychological well-being

difficulties.

2 RELATED WORKS

Self-destructive ideation recognition has turned into a

urgent exploration region, expecting to lessen self-

mischief and self-destruction rates by distinguishing

people in danger early. The test lies in perceiving

unobtrusive signs of self-destructive considerations

and separating them from other psychological

wellness conditions. Research has investigated the

conduct, mental, and social variables adding to self-

destructive ideation. This frequently includes a mix

of clinical assessments, psychometric evaluations,

and social perceptions to identify advance notice

signs, for example, state of mind changes,

correspondence moves, and modified social ways of

behaving. One of the earliest ways to deal with

recognizing self-destruction risk implied clinical

appraisals and organized interviews led via prepared

experts. These appraisals offer a top to bottom

assessment of an individual's mental state, past

psychological well-being issues, and any indications

of despondency, uneasiness, or injury. A critical part

of these assessments incorporates perceiving direct

articulations of self-destructive contemplations and

connecting them to realized risk factors, for example,

substance misuse, earlier self-destruction endeavors,

and family ancestry. Mental scales like the Beck

Gloom Stock (BDI) and the Self destruction Ideation

Survey (SIQ) have been much of the time utilized in

exploration to gauge the seriousness of self-

destructive considerations. In any case, these self-

announced apparatuses might be restricted by

underreporting or social allure predispositions.

Notwithstanding conventional strategies, there has

been a rising spotlight on friendly observing to

recognize people in danger. Web-based entertainment

stages, for instance, stand out as many individuals

share their own encounters, battles, and feelings on

the web. Studies have investigated the capability of

utilizing AI calculations to break down virtual

entertainment posts and distinguish examples, for

example, changes in language use, negative opinion,

or sadness. Research recommends that people

encountering self-destructive considerations might

show a change in tone, jargon, and composing style,

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

frequently reflecting pessimistic feelings or

detachment. Albeit these methodologies can be

harmless and contact a more extensive crowd, they

raise worries about security, assent, and moral issues.

Notwithstanding these worries, web-based

entertainment-based recognition frameworks are

being considered as a strengthening instrument for

early mediation. Mental overviews and surveys are

additionally generally used to recognize those in

danger. These instruments plan to evaluate a person's

close to home and mental prosperity, investigating

regions like pressure, self-esteem, and strategies for

dealing with stress. Studies show that sensations of

sadness, defenselessness, and tireless pity are firmly

connected to self-destructive considerations.

Subsequently, customary evaluating for these

elements in clinical settings and local area wellbeing

programs is being proposed as a method for working

on early location and mediation. Late mechanical

headways have opened additional opportunities for

distinguishing self-destructive ideation. Wearable

gadgets and portable applications currently consider

persistent checking of physiological signals, for

example, pulse inconstancy, rest designs, and actual

work. These markers can assist with identifying close

to home pain, as people with self-destructive

considerations might encounter disturbed rest or

diminished action levels. A few examinations have

effectively utilized portable wellbeing (mHealth)

devices to screen emotional well-being progressively,

giving significant information that can make medical

care suppliers aware of expected gambles. These

devices are especially valuable for individuals in

distant regions or those reluctant to look for

customary expert assistance. Research has

additionally featured the significance of preparing

medical care suppliers to perceive indications of self-

destructive ideation and answer suitably. Particular

preparation programs for clinicians have been found

to work on their capacity to lead careful gamble

appraisals and mediate early. The utilization of

normalized risk evaluation apparatuses, similar to the

Columbia-Self destruction Seriousness Rating Scale

(C-SSRS), has been displayed to work on the

exactness of self-destruction risk ID. These projects

plan to decrease self-destruction rates by

guaranteeing that in danger people get ideal and

viable help. Past individual mediations, general

wellbeing endeavors have zeroed in on more

extensive cultural anticipation measures. Emotional

wellness mindfulness crusades and instructive

projects have picked up speed, meaning to diminish

the shame encompassing self-destruction and

advance open conversations about psychological

well-being. These drives urge people to look for help

unafraid of judgment. The mix of emotional well-

being instruction into schools, work environments,

and local area settings is assisting with bringing

issues to light of the significance of mental prosperity

and the requirement for proactive counteraction of

self-hurt. Remedial mediations for those

distinguished as in danger have likewise been a

critical examination region. Mental social treatment

(CBT), persuasive conduct treatment (DBT), and

other remedial strategies have demonstrated

compelling in lessening self-destructive

contemplations. These treatments center around

changing pessimistic idea designs, upgrading survival

methods, and working on profound guideline. Joining

psychotherapy with pharmacological medicines, for

example, antidepressants or mind-set stabilizers, has

been investigated as a method for addressing

fundamental psychological wellness conditions

adding to self-destructive ideation. All in all, research

on distinguishing self-destructive ideation has

progressed essentially, with enhancements in clinical

appraisals, social checking, mechanical apparatuses,

and preparing for medical services suppliers. In spite

of the fact that difficulties remain, especially in

regards to protection and moral worries, the

advancement of exhaustive, diverse methodologies

offers expect diminishing worldwide self-destruction

rates. A definitive point is to make a viable, open

framework for early recognition that can save lives by

forestalling reckless ways of behaving and advancing

psychological wellness care.

3 METHODOLOGY

3.1 Logistic Regression

Definition: Logistic Regression is a statistical

method primarily used for binary classification tasks.

It predicts the likelihood that a given input belongs to

a specific class, such as identifying whether someone

may have suicidal ideation or not.

How it Works: The model applies a logistic or

sigmoid function to predict probabilities within the

range of 0 to 1. It assumes a linear relationship

between the input features and the output. To find the

best parameters (weights), the algorithm minimizes

the log-loss (binary cross-entropy), which measures

the difference between predicted probabilities and

actual outcomes. The training process typically uses

optimization techniques like Gradient Descent.

Suicidal Ideation Detection Using Machine Learning

Application: This technique can be used to predict

the likelihood of suicidal ideation based on features

such as behavioral patterns or survey responses.

3.2 Decision Tress Classifier

Definition: A Decision Tree Classifier is a tree-

structured model used for classification tasks. It

divides the dataset into subsets based on feature

values, creating a tree-like structure where each node

represents a decision, and each leaf node represents a

class label.

How it Works: The model recursively splits the

dataset by selecting the feature that best separates the

data using metrics such as Gini Impurity or

Information Gain (Entropy). This splitting continues

until a stopping condition is met, such as when all

data points in a node belong to the same class or a

predefined tree depth is reached. While Decision

Trees are interpretable, they may overfit if not

properly pruned.

Application: Decision Trees can be used to identify

patterns in communication, emotional states, or

behavior that are strongly associate with suicidal

ideation.

3.3 Random Forest Classifier

Definition: Random Forest is an ensemble learning

method that builds multiple Decision Trees and

merges their outputs to enhance accuracy and reduce

overfitting.

How it Works: The model uses a technique called

bagging (bootstrap aggregating), where each tree is

trained on a random subset of the data. At each node,

a random subset of features is selected for splitting,

promoting diversity among the trees. The final output

is determined by taking the majority vote from all the

trees (for classification) or averaging the results (for

regression).

Application: Random Forest can improve

classification accuracy by averaging the predictions

from multiple trees, making it useful for detecting

suicidal ideation with higher reliability.

3.4 AdaBoost Classifier

Definition: AdaBoost, or Adaptive Boosting, is an

ensemble method that combines weak classifiers

(often Decision Trees) into a stronger model. It

focuses more on cases where previous models have

failed.

How it Works: Initially, all data points are assigned

equal weights. A weak classifier is trained on the

weighted data, and after each iteration, the weight of

incorrectly classified points is increased. The final

prediction is made by combining the results of all

classifiers, with more weight given to those with

higher accuracy.

Application: AdaBoost can help detect subtle signs

of suicidal ideation by focusing on instances that were

previously misclassified, thereby improving the

model’s sensitivity.

3.5 Gradient Boosting Classifier

Definition: Gradient Boosting is an ensemble

technique that builds models sequentially, where each

new model corrects the errors made by the previous

one. It employs gradient descent to minimize the loss

function.

How it Works: Initially, a weak model is trained, and

subsequent models focus on correcting the errors

(residuals) of the previous models. Gradient descent

is used to fine-tune the model parameters step by step,

ensuring that each new model targets the mistakes of

the ensemble. The final prediction combines the

outputs of all models using a weighted sum.

Application: Gradient Boosting can enhance

predictive accuracy for detecting suicidal ideation,

particularly in complex datasets with intricate

relationships.

3.6 Gaussian Naive Bayes

(GaussianNB)

Definition: Gaussian Naive Bayes is a probabilistic

classifier based on Bayes' theorem, which assumes

that features follow a Gaussian (normal) distribution.

It’s often used for classification tasks where the

features are considered independent.

How it Works: The algorithm computes the

probability of each class (e.g., suicidal ideation vs. no

suicidal ideation) using Bayes' theorem. It assumes

that the features are independent (a naive assumption)

and follow a normal distribution. The class with the

highest probability is chosen as the prediction.

Application: GaussianNB can be applied to classify

individuals based on statistical patterns in their

behaviors or survey responses, particularly if these

features are normally distributed. Table 1: show the

Comparison table for all the algorithms.

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

Table 1: Comparison Table for All the Algorithms.

Model Accuracy

Logistic Regression 0.7376

Decision Tree Classifier 0.8156

Random Forest

Classifie

0.8014

AdaBoost Classifier 0.8085

Gradient Boosting

Classifie

0.8227

Gaussian Naive Bayes

(GaussianNB)

0.7376

4 DISCUSSION AND RESULT

This study underscores the essential job of early

identification and mediation in distinguishing people

in danger of self-destructive contemplations. The

created framework, which coordinates social, mental,

and conduct examination strategies, demonstrated

exceptionally successful in distinguishing

unpretentious changes in people's way of behaving

and figuring designs that could show self-destructive

ideation. Social checking, which tracks varieties in

correspondence designs, social segregation, and

everyday propensities, arose as a solid strategy for

early recognition. Besides, psychological well-being

assessments, including surveys pointed toward

evaluating profound misery and distinguishing risk

factors, fundamentally added to the ID interaction. By

using AI calculations to perceive designs, the

framework showed a serious level of precision in

distinguishing people in danger, limiting bogus

negatives and guaranteeing opportune mediation.

Joint effort with medical services experts improved

the framework's general viability. Clinical experts

could consolidate bits of knowledge from

psychological wellness evaluations and social

conduct investigations into their clinical assessments,

offering a more intensive comprehension of a

patient's condition. This incorporated methodology

considered focusing on mediations in light of hazard

seriousness and fitting treatment intends to address

individual issues. A critical component of this study

was the improvement of a non-meddlesome and

effectively open identification framework. Intended

to be applied across different settings, for example,

clinics and local area emotional well-being programs,

the framework guaranteed availability for people in

both metropolitan and country regions. Furthermore,

by limiting the disgrace around emotional well-being

evaluations, the review advances a more

comprehensive, proactive way to deal with tending to

psychological well-being issues. The outcomes show

that when executed accurately, such structures further

develop discovery and mediation times as well as

establish a steady climate, empowering people to look

for help unafraid of judgment. All in all, the

discoveries of this study underline the benefit of

joining different strategies clinical appraisals,

mental assessments, and social conduct observing to

upgrade the exactness of recognizing self-destructive

considerations. The proposed system is supposed to

further develop early intercession methodologies,

prompting improved results for those in danger of

self-mischief or self-destruction significantly.

5 CONCLUSIONS

Recognizing self-destructive ideation is a crucial

examination region zeroed in on distinguishing

people in danger of self-mischief or self-destruction,

offering a chance for opportune mediation. By

utilizing a scope of procedures, like social

examination, mental assessments, and example

acknowledgment, this study highlights the meaning

of early identification of self-destructive

contemplations. These methods, which remember

following changes for correspondence, close to home

wellbeing, and social way of behaving, empower

specialists to anticipate and recognize expected

gambles. Integrating clinical appraisals and

emotional wellness overviews into regular practices

can upgrade the accuracy of hazard assessments and

mediations, making a more all-encompassing way to

deal with mental medical services. Also, this

exploration features the significance of medical

services experts utilizing these instruments inside

clinical conditions to make self-destruction

counteraction methodologies more open and viable.

The production of a harmless, easy to understand, and

dependable identification framework is basic to

cultivating a more proactive psychological well-

being system. Such a framework wouldn't just guide

in recognizing in danger people in clinical settings yet

in addition assist with lessening the disgrace related

with looking for help for self-destructive

considerations. The far and wide execution of this

framework could have huge positive effects,

particularly in clinics, emotional well-being focuses,

and local area programs, where early and precise

distinguishing proof can significantly bring down the

gamble of lethal results. This preventive model

advances a worldwide comprehension of emotional

wellness issues, pushing for improved public

mindfulness and a shift toward a steadier and less

Suicidal Ideation Detection Using Machine Learning

critical environment for those confronting mental

difficulties. Eventually, by putting resources into

recognition frameworks and further developing

intercession draws near, social orders can gain

significant headway toward bringing down self-

destruction rates and advancing mental prosperity

across different populaces.

6 FUTURE ENHANCEMENT

Future upgrades in self-destructive ideation

identification frameworks can zero in on a few

significant regions to improve both accuracy and

openness. One potential improvement is

consolidating man-made brainpower (artificial

intelligence) and AI (ML) calculations to upgrade

identification exactness. These advancements can

deal with a lot of information from different sources

like text, discourse, and virtual entertainment action,

recognizing unobtrusive, complex examples that

conventional strategies could miss. By consistently

gaining from new data, these frameworks can remain

refreshed on developing social patterns, giving

constant gamble assessments and working with

quicker intercessions. One more improvement could

be the advancement of portable applications and

wearable gadgets that screen people continuously,

following social and physiological pointers, for

example, rest designs, action levels, and voice tone.

These gadgets could offer nonstop, aloof perception,

alarming medical care suppliers or encouraging

groups of people assuming it are distinguished to

concern signs. This proactive methodology would

empower faster intercessions and guarantee more

prominent availability to those in danger, particularly

for people who may not effectively look for help.

Furthermore, improving the social responsiveness of

recognition frameworks is a vital region for

improvement. Self-destructive ideation can introduce

contrastingly across different societies, and fitting

evaluation devices to reflect social and etymological

contrasts would work on the framework's capacity to

recognize takes a chance in a worldwide setting. This

could include preparing AI models with different

datasets that address a great many social foundations

Another improvement includes incorporating

emotional well-being experts into the input circle.

While artificial intelligence and computerized

frameworks can offer important experiences, human

mastery is fundamental while deciding the gamble

level and proper game-plan. By joining clinical

master input with artificial intelligence driven

expectations, independent direction can be refined,

prompting more customized care. Also, bringing

issues to light about psychological well-being and

decreasing the shame encompassing looking for help

is essential. Future frameworks ought to incorporate

instructive parts to educate clients about advance

notice signs regarding self-destructive ideation and

empower taking care of oneself methodologies, while

additionally offering prompt admittance to advising

administrations. In outline, the fate of self-destructive

ideation identification will probably fixate on joining

trend setting innovations, social mindfulness,

continuous observing, and human joint effort to make

a more viable and open framework pointed toward

forestalling self-mischief and saving lives.

REFERENCES

Chadha, A., & Kaushik, B. (2021). A Survey on

Prediction of Suicidal Ideation Using Machine and

Ensemble Learning. The Computer Journal, 64(11),

1617–1632. https://doi.org/10.1 093/COMJNL/BXZ12

Dhelim, S., Chen, L., Ning, H., & Nugent, C. (2023).

Artificial intelligence for suicide assessment using

Audiovisual Cues: a review. Artificial Intelligence

Review, 56(6), 5591–5618. https://doi.org/10.1007/S

10462-022-10290-6/METRICS

Dr. P. Golda Jeyasheeli, C. K. K. S. A. (2022). Deep

Learning Methods for Suicide Prediction using Audio

Classification. Journal of Positive School Psychology,

10479-10485–10479–10485. https://journalppw.com

/index.php/jpsp/article/view/6384

Galatzer-Levy, I., Abbas, A., Ries, A., Homan, S., Sels, L.,

Koesmahargyo, V., Yadav, V., Colla, M., Scheerer, H.,

Vetter, S., Seifritz, E., Scholz, U., & Kleim, B. (2021).

Validation of visual and auditory digital markers of

suicidality in acutely suicidal psychiatric inpatients:

Proof-of-concept study. Journal of Medical Internet

Research, 23(6), e25199. https://doi.org/10.2196/2519

Heckler, W. F., de Carvalho, J. V., & Barbosa, J. L. V.

(2022). Machine learning for suicidal ideation

identification: A systematic literature review.

Computers in Human Behavior, 128, 107095.

https://doi.org/10.1016/J.CHB.2021.107095

Ji, S., Pan, S., Li, X., Cambria, E., Long, G., & Huang, Z.

(2021). Suicidal Ideation Detection: A Review of

Machine Learning Methods and Applications. IEEE

Transactions on Computational Social Systems, 8(1),

214–226. https://doi.org/10.1109/TCSS.2020.3021467.

Kancharapu, R., & Ayyagari, S. N. (2024). Suicidal

ideation prediction based on social media posts using a

GAN-infused deep learning framework with genetic

optimization and word embedding fusion. International

Journal of Information Technology (Singapore), 16(4),

2577–2593. https://doi.org/10.1007/S41870-023-

01725-6/METRICS

ICRDICCT‘25 2025 - INTERNATIONAL CONFERENCE ON RESEARCH AND DEVELOPMENT IN INFORMATION,

COMMUNICATION, AND COMPUTING TECHNOLOGIES

Li, T. M. H., Chen, J., Law, F. O. C., Li, C.-T., Chan, N.

Y., Chan, J. W. Y., Chau, S. W. H., Liu, Y., Li, S. X.,

Zhang, J., Leung, K.-S., & Wing, Y.-K. (2023).

Detection of Suicidal Ideation in Clinical Interviews for

Depression Using Natural Language Processing and

Machine Learning: Cross-Sectional Study. JMIR

Medical Informatics, 11(1), e50221. https://doi.org/10

.2196/50221

Parsapoor, M., Koudys, J. W., & Ruocco, A. C. (2023).

Suicide risk detection using artificial intelligence: the

promise of creating a benchmark dataset for research on

the detection of suicide risk. Frontiers in Psychiatry, 14,

1186569. https://doi.org/10.3389/FPSYT.2023.11865

69/BIBT

Pillai, A., Nepal, S. K., Wang, W., Nemesure, M., Heinz,

M., Price, G., Lekkas, D., Collins, A. C., Griffin, T.,

Buck, B., Preum, S. M., Cohen, T., Jacobson, N. C.,

Ben-Zeev, D., & Campbell, A. (2024). Investigating

Generalizability of Speech-based Suicidal Ideation

Detection Using Mobile Phones. Proceedings of the

ACM on Interactive, Mobile, Wearable and Ubiquitous

Technologies, 7(4). https://doi.org/10.1145/3631452

Sawhney, R., Manchanda, P., Mathur, P., Shah, R., &

Singh, R. (2018). Exploring and Learning Suicidal

Ideation Connotations on Social Media with Deep

Learning. 167–175. https://doi.org/10.18653/V1/W18-

6223

Tlachac, M. L., Dixon-Gordon, K., & Rundensteiner, E.

(2021). Screening for suicidal ideation with text

messages. BHI 2021 - 2021 IEEE EMBS International

Conference on Biomedical and Health Informatics,

Proceedings. https://doi.org/10.1109/BHI50953.2021.

9508486

Tlachac, M. L., Flores, R., Reisch, M., Kayastha, R.,

Taurich, N., Melican, V., Bruneau, C., Caouette, H.,

Lovering, J., Toto, E., & Rundensteiner, E. A. (2022).

StudentSADD: Rapid mobile depression and suicidal

ideation screening of college students during the

coronavirus pandemic. Proceedings of the ACM on

Interactive, Mobile, Wearable and Ubiquitous

Technologies 6(2). https://doi.org/10.1145/3534604/S

UPPL_FILE/TLACHAC-1.ZIP

Yao, H., Rashidian, S., Dong, X., Duanmu, H., Rosenthal,

R. N., & Wang, F. (2020). Detection of suicidality

among opioid users on reddit: Machine learning-based

approach. Journal of Medical Internet Research, 22(11),

e15293. https://doi.org/10.2196/15293

Suicidal Ideation Detection Using Machine Learning