Natural Language Processing on SNS Data to Extract Features of Psychiatric Disorders

Natural Language Processing on SNS Data to Extract Features of Psychiatric Disorders - Natural Language Processing on SNS Data to Extract Features of Psychiatric Disorders

Status

Active, not recruiting

Phases

Unknown

Study type

Observational

Source

JPRN

Registry ID

JPRN-UMIN000034489

Enrollment

300

Registered

2018-10-15

Start date

2018-10-15

Completion date

Unknown

Last updated

2026-06-29

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Patients with major depressive disorder, bipolar I/II disorder, schizophrenia, and anxiety disorders (including obsessive compulsive disorder) by DSM-5 or ICD-10

Interventions

None listed

Eligibility

Sex/Gender

All

Inclusion criteria

Inclusion criteria: As patients (1) Out/in-patients regularly visiting or hospitalized at the study sites and/or other related facilities, and diagnosed with major depressive disorder, bipolar I/II disorder, schizophrenia, and anxiety disorders (including obsessive compulsive disorder) according to DSM-5 or ICD-10. (2) Those who post on social network services, such as Twitter or Facebook. (3) 20 years old or older. (4) Decisionally unimpaired as judged by treating physician. If judged as decisionally impaired, patients' guardians should give consent. As healthy volunteers (1) Healthy volunteers who offered to participate in the study through study website. (2) Those who post on social network services, such as Twitter or Facebook. (3) 20 years old or older.

Exclusion criteria

Exclusion criteria: As patients (1) Patients whose illness can exacerbate by interview of the study. (2) Patients who have comorbidities that can interfere with posting to social network service; such as patients with hand paralysis or visual impairment. (3) Those who are considered to be ineligible by the PI or investigators. As healthy volunteers (1) Those who have comorbidities that can interfere with posting to social network service; such as patients with hand paralysis or visual impairment. (2) Those who are considered to be ineligible by the PI or investigators

Design outcomes

Primary

Measure	Time frame
Linguistic features of each psychiatric diagnosis identified through natural language processing and machine learning	—

Countries

Japan

Contacts

Public ContactMomoko Kitazawa

Keio University School of Medicine Department of Neuropsychiatry

m-kitazawa@keio.jp03-5363-3492

Outcome results