ALENTAR-J-CM PROJECT
Research for youth well-being and mental health
ALENTAR-J-CM PROJECT
Research for youth well-being and mental health
Our mission
ALENTAR-J-CM comes from the project’s acronym Large Language Models Application for large-scale healthcare prevention of mental health problems and suicide risk among young people.
This research project will develop and implement an advanced system based on Large Language Models (LLMs) for the early detection and assessment of emotional disorders and suicide risk among young people (ages 12 to 25).
AI to detect and prevent emotional disorders in young people
Our main research objective is to develop an advanced technological prototype, based on Large Language Models (LLMs), to support clinical health professionals in their daily work—particularly in the detection and evaluation of emotional disorder risks (anxiety, depression, etc.) and suicide risk in adolescents and young people.
We aim to be pioneers in offering fast, effective and reliable tool to prevent the emergence and development of these mental health problem, while contributing with our knowledge to research and clinical intervention.
ALENTAR-J-CM is constantly evolving, adapting to mental health challenges and strengthening its commitment through research and the implementation of affective language models.
OUR APPROACH
Objective and action plan
Detecting linguistic markers
We analyze discourse markers and paralinguistic features to support the detection and prevention of emotional problems in young people.
Developing algorithms
We design systems based on LLMs for the early detection and prevention of mental health issues and suicidal behavior.
Detection of linguistic markers
Identification
We will identify and define paralinguistic and discourse parameters.
Analysis
We will collect and analyze written and vocal expression data from diverse populations.
Validation
We will validate the obtained parameters in real-world contexts.
Development
We will develop evidence-based prevention resources and services.
Identification
We will identify and define paralinguistic and discourse parameters.
Analysis
We will collect and analyze written and vocal expression data from diverse populations.
Validation
We will validate the obtained parameters in real-world contexts.
Development
We will develop evidence-based prevention resources and services.
Algorithm development
Identification
We will develop LLM-based algorithms to identify linguistic markers.
Validation
We will design and validate LLM-based algorithms using new field data obtained from our clinical studies.
Implementation
We will create software prototypes for social and healthcare centers and hospitals.
Identification
We will develop LLM-based algorithms to identify linguistic markers.
Validation
We will design and validate LLM-based algorithms using new field data obtained from our clinical studies.
Implementation
We will create software prototypes for social and healthcare centers and hospitals.
IMPLEMENTATION
Project phases
Model training with existing data
Identification and definition of paralinguistic and discourse parameters.
Automatic reading of medical records.
Model training with data from psychological support hotlines and chat services
Field studies
Written and vocal expression data collection and analysis.
Validation of parameters in real-world contexts.
Development and validation of LLM-based algorithms using new data.
Advanced Large Language Models
Development and validation of LLM-based algorithms using new field data derived from clinical studies.
Software prototype
Development of a support system prototype for community and hospital-based services.
Implementation of a software prototype for social and healthcare centers and hospitals.
implementation
Project phases
Model training with existing data
Identification and definition of paralinguistic and discourse parameters.
Automatic reading of medical records.
Model training with data from psychological support hotlines and chat services
Field studies
Written and vocal expression data collection and analysis.
Validation of parameters in real-world contexts.
Development and validation of LLM-based algorithms using new data.
Advanced Large Language Models
Development and validation of LLM-based algorithms using new field data derived from clinical studies.
Software prototype
Development of a support system prototype for community and hospital-based services.
Implementation of a software prototype for social and healthcare centers and hospitals.
Every step matters.
Let’s join efforts to support emotional well-being in our society.
This website is supported by the Dirección General de Investigación e Innovación Tecnológica de la Comunidad de Madrid (Orden 3177/2024) through the I+D Technological activities program (TEC-2024/COM-224) and the European Regional Development Fund (ERDF)