Mental Health Detection using Online Social Media Mining

ABSTRACT:

Online Social Media (OSM) platforms have become integral to modern communication, allowing users to express their thoughts, emotions, and opinions in real-time. With the widespread use of social media, there is an increasing concern regarding its impact on mental health. Identifying mental health conditions through social media activity can provide early intervention opportunities for individuals experiencing distress. Traditional mental health assessments rely on self-reporting and clinical evaluations, which may not always be timely or accessible. Therefore, leveraging social media data for mental health detection presents a promising approach to monitoring emotional well-being and identifying individuals who may require support.

This project, “Mental Health Detection using Online Social Media Mining” is developed using Java as the programming language, JSP, CSS, and JavaScript for the frontend, and MySQL as the database. A distinguishing feature of this system is its ability to analyze user-generated content for mental health assessment.

When users post tweets, the content is analyzed for the presence of these keywords. If a user’s tweets contain a high occurrence of such words, the system categorizes the user as being in an “abnormal” mental health state. This information is updated in the admin panel, allowing the administrator to monitor mental health trends within the platform.

This project demonstrates the potential of using social media mining techniques for mental health detection. By analyzing textual data and identifying patterns in user behavior, the system offers a novel approach to early mental health assessment. The integration of OSN functionalities with mental health monitoring provides a scalable and efficient solution for detecting emotional distress in social media users.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

The existing systems for mental health detection primarily relied on traditional psychological assessments, surveys, and clinical evaluations conducted by mental health professionals. These assessments were based on self-reported data collected through structured questionnaires, interviews, or direct interactions with individuals. Healthcare providers and psychologists analyzed these responses to diagnose mental health conditions and provide necessary interventions.
With the advancement of technology, mental health detection systems evolved to incorporate digital tools such as mobile applications and online self-assessment platforms. These platforms allowed users to answer predefined sets of questions related to their emotional state, daily activities, and behavioral patterns. The responses were then processed using predefined rules or scoring mechanisms to provide insights into the user’s mental health condition. Additionally, some systems utilized Natural Language Processing (NLP) techniques to analyze text-based responses from users to detect early signs of mental health concerns.
In parallel, social media platforms have been extensively used for sentiment analysis and behavioral studies. Researchers and organizations leveraged social media data to study user sentiments, engagement levels, and content trends. Some early systems applied basic keyword-based analysis to detect mood variations by scanning user posts for specific emotional expressions. These approaches provided valuable insights into social media behavior and its correlation with mental health but were not directly integrated into an interactive social networking platform.
Overall, the existing systems before the development of this project mainly focused on structured assessments, self-reported data collection, and early sentiment analysis techniques. They played a significant role in understanding mental health trends and provided a foundation for integrating social media-based mental health detection in more advanced systems.

DISADVANTAGES OF EXISTING SYSTEM:

Despite the advancements in mental health detection, the existing systems had several limitations that affected their accuracy, efficiency, and accessibility. One major drawback was the reliance on self-reported data, which often led to biased or inaccurate assessments. Individuals might not always provide truthful responses due to stigma, lack of awareness about their condition, or difficulty in expressing their emotions. This resulted in incomplete or misleading evaluations, reducing the effectiveness of the mental health assessment process.
Another limitation was the time-consuming nature of traditional psychological assessments and clinical evaluations. These assessments required individuals to undergo structured interviews or fill out lengthy questionnaires, making them less convenient for real-time or large-scale mental health monitoring. Additionally, the dependence on trained mental health professionals for analysis and diagnosis further limited accessibility, especially in regions with a shortage of mental health experts.
Early digital mental health detection systems, such as mobile applications and online self-assessment tools, had restricted capabilities in detecting real-time emotional fluctuations. Since they primarily relied on predefined question sets or manual inputs, they lacked continuous monitoring and failed to capture sudden mood changes. Moreover, many of these systems did not consider social and contextual factors influencing an individual’s mental health, leading to a lack of personalized insights.
In sentiment analysis-based approaches using social media, previous systems often relied on basic keyword detection without considering context, sarcasm, or evolving linguistic expressions. As a result, misinterpretations were common, leading to false positives or negatives in detecting mental health concerns. Additionally, these systems lacked direct integration with interactive social networking platforms, making it difficult to track behavioral trends over time or provide real-time analysis based on user activity.
Overall, the limitations of existing systems highlighted the need for a more advanced, real-time, and context-aware approach to mental health detection, integrating social media mining techniques to enhance accuracy and efficiency.

PROPOSED SYSTEM:

The proposed system, “Mental Health Detection using Online Social Media Mining,” integrates an Online Social Networking (OSN) platform with real-time mental health analysis. Developed using Java for backend processing, JSP, CSS, and JavaScript for the frontend, and MySQL as the database, the system enables user interactions while monitoring their mental well-being through social media activity.
The system consists of two key entities: Admin and Users. New users must register on the platform before accessing its features. Upon successful authentication, users can explore the network, follow other users, view their tweets, and post their own tweets, including text and images. The OSN module replicates standard social networking functionalities, allowing seamless interaction and engagement among users.
A significant aspect of this system is its ability to analyze user-generated content to detect potential mental health conditions. A predefined database contains a list of words associated with mental health concerns. When users post tweets, the system scans the content and compares it with the predefined word set. If a tweet contains words linked to abnormal mental health states, the system flags the user’s mental state accordingly.
The admin panel provides comprehensive monitoring features, including access to user details, posted tweets, individual mental health analysis, and overall mental health trends. The admin can view all registered users along with their profile information, including User ID, profile photo, name, email, date of birth, gender, state, and country. Additionally, the admin can track all tweets posted on the platform, including tweet IDs, user IDs, content, and associated images.
For mental health monitoring, the system offers detailed insights into each user’s emotional well-being. The admin can analyze individual users’ mental health by evaluating the frequency of flagged words in their tweets. A graphical representation displays the proportion of normal and abnormal tweets, helping in visualizing trends. The system also provides an overall mental health analysis, aggregating data from all users and generating a graphical representation of mental health trends across the platform.
The entire mental health detection mechanism is implemented using a flexible rule-based system that ensures systematic evaluation based on predefined linguistic patterns. This approach enables continuous monitoring of user activity, facilitating real-time detection and tracking of mental health indicators through social media interactions.

ADVANTAGES OF PROPOSED SYSTEM:

The proposed system, “Mental Health Detection using Online Social Media Mining” offers several advantages by integrating an Online Social Networking (OSN) platform with real-time mental health analysis. This innovative approach enhances mental health monitoring through social media interactions while providing seamless user engagement.
One of the key advantages of the system is real-time mental health detection. Unlike traditional methods that rely on self-reported data or scheduled assessments, this system continuously analyzes user-generated content. By monitoring tweets in real-time, the system can detect potential mental health concerns early, allowing timely interventions.
The system also ensures automated mental health assessment using a predefined database of words associated with mental health conditions. This rule-based detection mechanism eliminates the need for manual screening, improving efficiency and scalability. The automated process enables the system to assess multiple users simultaneously, making it suitable for large-scale deployment.
Another significant advantage is the integration of social networking features. The system is designed as a fully functional OSN platform, where users can register, follow others, post tweets, and share images. This approach ensures a natural and engaging user experience while enabling mental health detection without disrupting typical social media interactions.
The graphical representation of mental health trends provides clear insights into individual and overall platform-wide mental health status. By visualizing mental health trends using graphs, the system enhances the interpretation of data, allowing administrators to make informed decisions. Individual mental health status is tracked based on tweet patterns, while an aggregated view of all users’ mental health is also available for broader analysis.
Additionally, the admin panel offers comprehensive monitoring capabilities, allowing administrators to track user activity, review posted tweets, and analyze mental health trends. The structured organization of user details, tweet records, and mental health insights ensures efficient management and supervision of the platform.
Lastly, the system ensures flexibility and scalability by employing a rule-based detection mechanism that can be updated with new keywords or expanded to include advanced analytics. Future improvements, such as machine learning integration, can further enhance the accuracy and adaptability of the detection process.
Overall, the proposed system provides an efficient, automated, and interactive solution for mental health monitoring using online social media mining, making it a valuable tool for early detection and awareness.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

System : Pentium i3 Processor.
Hard Disk : 20 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

Mental Health Detection using Online Social Media Mining