DETECTION OF DEPRESSION STATUS FROM TEXT USING NLP APPROACH: COMPARATIVE PERFORMANCE OF DIFFERENT ENSEMBLE ALGORITHMS

Zülfikar Aslan

Makale detayları

DETECTION OF DEPRESSION STATUS FROM TEXT USING NLP APPROACH: COMPARATIVE PERFORMANCE OF DIFFERENT ENSEMBLE ALGORITHMS

Yazarlar:

Zülfikar Aslan

Yayıncı:

Gaziantep University

Ana Başlıklar:

Information Systems Development Methodologies and Practice

Dergi:

The International Journal of Energy and Engineering Sciences Vol.9 Issue 2, pp.79-92

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

İngilizce

Document ID (Org No.):

2001319139

pISSN:

2602-294X

Anahtar kelimeler:

Özet:

 This study aims to compare the performance of different ensemble learning algorithms using Natural Language Processing NLP approach for the detection of depression status from text. Depression is a common mental health problem worldwide and early diagnosis and intervention are of critical importance. A dataset consisting of 124,017 tweets collected between 2019-2020 was used in the study. These tweets were classified according to their depressive and non-depressive content. Five different ensemble learning algorithms, namely Random Forest, AdaBoost, Gradient Boosting, XGBoost and Voting Classifier, were applied in the study. The performance of the models was evaluated using various metrics such as accuracy, precision, recall and F1-score. In addition, ROC curves and learning curves were analyzed. The results revealed that all ensemble learning algorithms showed high performance, but Random Forest gave the best results in all metrics. Voting Classifier performed the second best, while Gradient Boosting performed relatively poorly. This study shows that ensemble learning algorithms are effective in detecting depression from text, and Random Forest in particular can be used as a potential screening tool in this area. The findings may contribute to the development of technology-supported approaches for early diagnosis and intervention in the field of mental health.

 This study aims to compare the performance of different ensemble learning algorithms using Natural Language Processing NLP approach for the detection of depression status from text. Depression is a common mental health problem worldwide and early diagnosis and intervention are of critical importance. A dataset consisting of 124,017 tweets collected between 2019-2020 was used in the study. These tweets were classified according to their depressive and non-depressive content. Five different ensemble learning algorithms, namely Random Forest, AdaBoost, Gradient Boosting, XGBoost and Voting Classifier, were applied in the study. The performance of the models was evaluated using various metrics such as accuracy, precision, recall and F1-score. In addition, ROC curves and learning curves were analyzed. The results revealed that all ensemble learning algorithms showed high performance, but Random Forest gave the best results in all metrics. Voting Classifier performed the second best, while Gradient Boosting performed relatively poorly. This study shows that ensemble learning algorithms are effective in detecting depression from text, and Random Forest in particular can be used as a potential screening tool in this area. The findings may contribute to the development of technology-supported approaches for early diagnosis and intervention in the field of mental health.

 This study aims to compare the performance of different ensemble learning algorithms using Natural Language Processing NLP approach for the detection of depression status from text. Depression is a common mental health problem worldwide and early diagnosis and intervention are of critical importance. A dataset consisting of 124,017 tweets collected between 2019-2020 was used in the study. These tweets were classified according to their depressive and non-depressive content. Five different ensemble learning algorithms, namely Random Forest, AdaBoost, Gradient Boosting, XGBoost and Voting Classifier, were applied in the study. The performance of the models was evaluated using various metrics such as accuracy, precision, recall and F1-score. In addition, ROC curves and learning curves were analyzed. The results revealed that all ensemble learning algorithms showed high performance, but Random Forest gave the best results in all metrics. Voting Classifier performed the second best, while Gradient Boosting performed relatively poorly. This study shows that ensemble learning algorithms are effective in detecting depression from text, and Random Forest in particular can be used as a potential screening tool in this area. The findings may contribute to the development of technology-supported approaches for early diagnosis and intervention in the field of mental health.

 This study aims to compare the performance of different ensemble learning algorithms using Natural Language Processing NLP approach for the detection of depression status from text. Depression is a common mental health problem worldwide and early diagnosis and intervention are of critical importance. A dataset consisting of 124,017 tweets collected between 2019-2020 was used in the study. These tweets were classified according to their depressive and non-depressive content. Five different ensemble learning algorithms, namely Random Forest, AdaBoost, Gradient Boosting, XGBoost and Voting Classifier, were applied in the study. The performance of the models was evaluated using various metrics such as accuracy, precision, recall and F1-score. In addition, ROC curves and learning curves were analyzed. The results revealed that all ensemble learning algorithms showed high performance, but Random Forest gave the best results in all metrics. Voting Classifier performed the second best, while Gradient Boosting performed relatively poorly. This study shows that ensemble learning algorithms are effective in detecting depression from text, and Random Forest in particular can be used as a potential screening tool in this area. The findings may contribute to the development of technology-supported approaches for early diagnosis and intervention in the field of mental health.

Makale detayları

DETECTION OF DEPRESSION STATUS FROM TEXT USING NLP APPROACH: COMPARATIVE PERFORMANCE OF DIFFERENT ENSEMBLE ALGORITHMS

Yazarlar:

Zülfikar Aslan

Yayıncı:

Gaziantep University

Ana Başlıklar:

Information Systems Development Methodologies and Practice

Dergi:

The International Journal of Energy and Engineering Sciences Vol.9 Issue 2, pp.79-92

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

İngilizce

Document ID (Org No.):

2001319139

pISSN:

2602-294X

Depression Detection

Natural Language Processing (NLP)

Ensemble Learning Algorithms

Social Media Analysis

Sayı

Makale detayları

DETECTION OF DEPRESSION STATUS FROM TEXT USING NLP APPROACH: COMPARATIVE PERFORMANCE OF DIFFERENT ENSEMBLE ALGORITHMS

Yazarlar: Zülfikar Aslan

Yayıncı: Gaziantep University

Ana Başlıklar: Information Systems Development Methodologies and Practice

Dergi: The International Journal of Energy and Engineering Sciences Vol.9 Issue 2, pp.79-92

Belge Tipi: Makale

Yayın Tarihi: 12/2024

Belge Dili: İngilizce

Document ID (Org No.): 2001319139

pISSN: 2602-294X

Depression Detection

Natural Language Processing (NLP)

Ensemble Learning Algorithms

Social Media Analysis

Sayı

Yazarlar:

Zülfikar Aslan

Yayıncı:

Gaziantep University

Ana Başlıklar:

Information Systems Development Methodologies and Practice

Dergi:

The International Journal of Energy and Engineering Sciences Vol.9 Issue 2, pp.79-92

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

İngilizce

Document ID (Org No.):

2001319139

pISSN:

2602-294X