# Deteksi Depresi dan Kecemasan Pengguna Twitter Menggunakan Bidirectional LSTM

**Authors:** Kuncahyo Setyo Nugroho, Ismail Akbar, Affi Nizar Suksmawati,  Istiadi  
**arXiv:** [2301.04521](https://arxiv.org/abs/2301.04521)  
**Date:** 2023-01-11  
**Summary page:** https://www.summarizepaper.com/en/paper/2301.04521v1/

## Key points

- Study presented at CIASTECH 2021 focused on detecting depression and anxiety in Twitter users using Bidirectional LSTM
- Aimed to address social stigma surrounding mental health issues, particularly depression and anxiety
- Proposed BiLSTM model as effective approach for identifying potential users experiencing depression and anxiety based on online posts
- Researchers developed BiLSTM model that outperformed traditional machine learning models and standard LSTMs in detecting mental health issues in Twitter users with an accuracy rate of 94.12%
- Emphasized importance of deep learning approaches like LSTM networks in various fields such as sentiment analysis, traffic forecasting, and price prediction models

## Summary

The study presented at The 4th Conference on Innovation and Application of Science and Technology (CIASTECH 2021) focused on the detection of depression and anxiety in Twitter users using Bidirectional LSTM. The research aimed to address the social stigma surrounding mental health issues, particularly depression and anxiety, which often leads individuals to seek support on social media platforms like Twitter. Traditional methods of detecting mental health disorders through textual data are challenging due to users not explicitly discussing their mental state. Therefore, the study proposed a BiLSTM model as an effective approach for identifying potential users experiencing depression and anxiety based on their online posts. is a crucial aspect of overall well-being that affects individuals worldwide. However,  often prevents people from seeking help for conditions like  and . To overcome this barrier, researchers have turned to platforms like , where individuals may feel more comfortable sharing their thoughts and feelings. The team of researchers, including Kuncahyo Setyo Nugroho from Universitas Brawijaya, Ismail Akbar from UIN Maulana Malik Ibrahim, Affi Nizar Suksmawati, and Istiadi from Universitas Widyagama Malang, developed a BiLSTM model that outperformed traditional machine learning models and standard LSTMs in detecting mental health issues in Twitter users. The model achieved an impressive accuracy rate of 94.12%, showcasing its effectiveness in early detection and intervention for individuals experiencing depression and anxiety. Furthermore, the study referenced previous research on neural network techniques for text data analysis, emphasizing the importance of deep learning approaches like LSTM networks in various fields such as sentiment analysis, traffic forecasting, and price prediction models. By leveraging advanced technologies like , researchers can enhance their ability to analyze textual data effectively and contribute to improving mental health awareness and support systems for individuals struggling with depression and anxiety.