Detecting Stress Based on Social Interactions in Social Networks

PROJECT ABSTRACT:

Psychological stress has become a significant concern, impacting individuals’ overall well-being. Early detection is crucial for effective intervention and mental health support. With the widespread use of social media, people frequently share their daily experiences and engage with others online, providing an opportunity to analyze their digital interactions for stress detection. This study explores the connection between a user’s stress levels and their social network, revealing that an individual’s emotional state is closely linked to that of their online friends.

This project, “Detecting Stress Based on Social Interactions in Social Networks,” aims to develop a web-based application that identifies stress levels in users based on their social media activity. The system is built using Java for backend processing, JSP, CSS, and JavaScript for frontend development and MySQL for database management.

The web application mimics an online social networking system, where users can register, log in, find friends, follow friends, post tweets, retweet, and like tweets. An admin panel is integrated, allowing the administrator to monitor user activities and determine stress levels based on the content of their tweets. The system leverages a predefined database of stress-related keywords to analyze user posts and classify them as either normal or stressed. A graphical representation is generated in the admin dashboard, comparing the number of normal and stressed tweets, thereby providing valuable insights into user mental health trends.

This system can be beneficial in early stress detection and intervention, helping organizations, psychologists, or researchers analyze stress patterns and take proactive measures to promote mental well-being.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

In the existing system, several studies have explored emotion analysis on social media by examining user-generated content at the tweet level. These systems primarily rely on text-based linguistic features and classic classification approaches to analyze emotions in social media interactions. One such system, MoodLens, was developed for the Chinese microblogging platform Weibo, classifying emotions into four categories: angry, disgusting, joyful, and sad. This classification framework helped in understanding the sentiment behind users’ posts and how emotions are expressed in online social interactions.
In the existing system, research has been conducted on emotion propagation in social networks, focusing on how different emotions spread among users. Studies have found that negative emotions, such as anger, propagate more rapidly and extensively than positive emotions like joy. This finding highlights the importance of analyzing the influence of social connections in stress detection. Since stress is often categorized as a negative emotional state, incorporating insights from social interactions can enhance the accuracy of stress detection models.
The existing system primarily uses text-based sentiment analysis and classification techniques to detect emotional states from user posts. While these approaches provide valuable insights into social media behavior, they primarily focus on individual tweet-level analysis rather than holistic, user-level stress assessment. The correlation between a user’s social connections and their stress levels has not been extensively explored, presenting an opportunity for improvement in stress detection methodologies.

DISADVANTAGES OF EXISTING SYSTEM:

Stress Detection Methods Are Reactive: The existing psychological stress detection primarily relies on face-to-face interviews, self-report questionnaires, or wearable sensors. While these methods provide direct insights, they are often time-consuming, labor-intensive, and delayed, making them ineffective for real-time stress monitoring.
Limited to Text-Based Analysis: The existing system mainly focuses on analyzing textual content from social media posts to detect emotions. However, social media data consists of multiple modalities, including images, videos, and interactions, which are not utilized. This limited scope reduces the accuracy of stress detection, as users often express emotions through more than just text.
Lack of Social Relationship Consideration: Though some studies have explored user-level emotion detection, the impact of social relationships on stress levels has not been effectively analyzed. The influence of a user’s friends, followers, and network interactions plays a significant role in stress propagation, but the existing system does not incorporate this factor into its detection models.
Inability to Capture Emotion Propagation Patterns Effectively: Research indicates that negative emotions, such as anger and stress, spread more rapidly and broadly across social networks compared to positive emotions. However, the existing system does not efficiently utilize this insight to enhance stress detection, limiting its ability to predict stress trends based on social influence.
Lack of Real-Time Monitoring and Analysis: The existing systems are not designed for real-time stress detection, as they focus on analyzing past tweets and interactions. This delayed approach fails to provide timely interventions, making it less effective for proactive mental health monitoring.
Limited Scalability and Adaptability: Existing solutions are often tailored to specific languages or platforms (e.g., Weibo), making them difficult to scale for multi-platform or multilingual analysis. This lack of adaptability limits their applicability in a diverse social media environment.
These limitations highlight the need for an advanced system that integrates real-time analysis, multi-modal data processing, and social network influence factors to improve stress detection accuracy and efficiency.

PROPOSED SYSTEM:

The proposed system introduces an advanced web-based application designed to detect psychological stress among users by analyzing their social interactions on a social networking platform. Developed using Java for backend processing, JSP, CSS, and JavaScript for the frontend, and MySQL for database management, this system offers an automated approach to stress identification based on user-generated content.
In this system, users can register, log in, find and follow friends, post tweets, like, and retweet posts, mimicking a real-world social media environment. The admin panel provides monitoring capabilities, allowing the administrator to track user activity and analyze stress levels based on the content of their tweets. A predefined database of stress-related words is integrated into the system, enabling automated detection of stress indicators within user posts.
Additionally, the system employs data visualization techniques, presenting stress analysis results through graphical reports that compare normal and stress-related tweets. This visualization helps in understanding behavioral patterns and social influences on stress levels. The system aims to offer a structured approach to stress detection, leveraging text processing techniques and social network analysis to assess mental well-being effectively.

ADVANTAGES OF PROPOSED SYSTEM:

Automated Stress Detection: The system uses a predefined database of stress-related words to automatically identify and classify stress levels in users based on their social media interactions, reducing the need for manual intervention.
Real-Time Monitoring: The web application continuously monitors user posts, likes, retweets, and interactions, providing real-time stress analysis for timely detection.
Enhanced Social Network Analysis: By analyzing interactions between users and their social circles, the system provides insights into how stress propagates through a network, offering a more holistic understanding of user mental health.
User-Friendly Web Interface: The system is designed with an intuitive user interface, making it easy for users to engage in social networking activities while the system works in the background to assess stress levels.
Graphical Representation of Stress Trends: The admin panel features a visual representation of stress trends, displaying comparative graphs of normal tweets and stress-indicating tweets, allowing better interpretation of mental health patterns.
Efficient Database Management: Using MySQL, the system effectively stores and manages user data, including stress-related posts, ensuring structured retrieval and analysis of information.
Scalability and Extensibility: The system can be expanded with additional functionalities such as machine learning-based sentiment analysis to improve stress detection accuracy and adapt to new trends in digital behavior.
Proactive Mental Health Awareness: By identifying stress indicators early, the system can be used for research, intervention, and support strategies, contributing to better mental health awareness in online communities.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

System : Pentium i3 Processor.
Hard Disk : 500 GB.
Monitor : 15’’ LED.
Input Devices : Keyboard, Mouse.
Ram : 4 GB.

SOFTWARE REQUIREMENTS:

Operating system : Windows 10/11.
Coding Language : JAVA.
Frontend : JSP, CSS, JavaScript.
JDK Version : JDK 23.0.1.
IDE Tool : Apache Netbeans IDE 24.
Tomcat Server Version : Apache Tomcat 9.0.84
Database : MYSQL.

Java Projects

Detecting Stress Based on Social Interactions in Social Networks