Akademik Veri Yönetim Sistemi
North American Chapter of the Association for Computational Linguistics (NAACL) Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology , Washington, Amerika Birleşik Devletleri, 10 - 15 Temmuz 2022, ss.219-225
In this shared task, we focus on detecting mental health signals in Reddit users’ posts through
two main challenges: A) capturing mood
changes (anomalies) from the longitudinal set
of posts (called timelines), and B) assessing
the users’ suicide risk-levels. Our approaches
leverage emotion recognition on linguistic
content by computing emotion/sentiment
scores using pre-trained BERTs on users’ posts
and feeding them to machine learning models,
including XGBoost, Bi-LSTM, and logistic
regression. For Task-A, we detect longitudinal
anomalies using a sequence-to-sequence
(seq2seq) autoencoder and capture regions
of mood deviations. For Task-B, our two
models utilize the BERT emotion/sentiment
scores. The first computes emotion bandwidths
and merges them with n-gram features, and
employs logistic regression to detect users’
suicide risk levels. The second model predicts
suicide risk on the timeline level using a BiLSTM on Task-A results and sentiment scores.
Our results outperformed most participating
teams and ranked in the top three in Task-A.
In Task-B, our methods surpass all others and
return the best macro and micro F1 scores.