Predicting Emotions of User in Online Social Network

ABSTRACT:

Predicting emotions of users in an online social network is a crucial aspect of understanding user interactions, sentiments, and overall engagement within a digital community. With the increasing usage of online social networking platforms, users frequently share their thoughts, opinions, and experiences through posts and messages. These interactions can reflect the emotional state of individuals, influencing social trends, public sentiment, and even mental health assessments. Analyzing and predicting emotions from such interactions can provide valuable insights for user engagement, content moderation, and community well-being.

The need for emotion prediction in online social networks arises from the growing importance of sentiment analysis and user behavior understanding. Identifying user emotions can help platform administrators in filtering harmful content, providing personalized recommendations, and fostering positive interactions. Many existing approaches leverage complex machine learning and deep learning models for sentiment analysis, making them computationally expensive and difficult to implement in real-time applications. To overcome this limitation, a lightweight yet effective approach is required, ensuring efficiency and accuracy while maintaining the scalability of the system.

The developed system is designed as an online social networking (OSN) platform, similar to Twitter, and incorporates fundamental OSN features along with emotion prediction capabilities. The system is built using Java for backend development, JSP, CSS, and JavaScript for the frontend, and MySQL as the database.

By integrating fundamental social networking features with an innovative emotion prediction mechanism, this system enhances user experience and administrative control. The lightweight approach to sentiment classification ensures efficient processing, making it a practical solution for online social networking platforms seeking to understand user emotions without relying on complex computational models. This project demonstrates a practical and scalable approach to emotion prediction in online interactions, paving the way for improved user engagement, content moderation, and social network analytics.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

The existing system of online social networking platforms primarily revolves around facilitating user interactions through various means such as posting updates, sharing media, messaging, and engaging with other users’ content. Platforms like Facebook, Twitter, and Instagram allow users to create profiles, connect with friends, and share thoughts, images, and videos in real time. These systems provide features such as news feeds, messaging services, and reaction mechanisms (likes, comments, and shares) to enhance social engagement.
In existing system platforms, users can follow or add friends, enabling them to view updates and interact with posts. Social networking sites use various engagement techniques, including recommendation systems, trending topics, and personalized content feeds. Additionally, administrative roles in these systems allow for user management, content moderation, and analytics for better understanding user behavior and trends.
The primary focus of existing systems has been on enabling seamless social interaction, content sharing, and networking. While they incorporate basic sentiment-driven interactions such as likes and emojis, they typically do not analyze emotions explicitly based on textual content or categorize posts according to emotional states.

DISADVANTAGES OF EXISTING SYSTEM:

The existing online social networking (OSN) platforms primarily focus on enabling communication, content sharing, and engagement among users. However, they have several limitations when it comes to sentiment analysis, emotion detection, and user behavior insights. Below are some of the key disadvantages of the existing system:
Lack of Emotion-Based Classification: The existing system social networking platforms do not have built-in mechanisms to classify user posts based on emotions. While users can express their feelings through tweets, likes, or comments, there is no direct categorization of emotions such as happiness, sadness, anger, or surprise based on textual content.
Limited Sentiment Analysis: The existing social networking sites focus on displaying content chronologically or based on engagement metrics. They do not analyze user emotions or provide insights into the overall sentiment of user posts. Sentiment analysis in these platforms often requires third-party tools or external machine learning models for interpretation.
Manual Interpretation of User Emotions: In the absence of an automated emotion detection system, understanding user sentiment requires manual effort. Users, moderators, or analysts must read and interpret posts to gauge emotions, making the process time-consuming and less efficient.
No Emotional Trends Analysis: Existing OSN platforms do not provide a systematic way to analyze emotional trends over time. While platforms may track engagement metrics (such as the number of likes, shares, and comments), they do not categorize content based on emotional states. This limitation restricts the ability to assess mood trends and behavioral patterns among users.
Limited Administrator Insights: Admins in traditional social media platforms primarily focus on content moderation, user management, and security. However, they lack the ability to analyze user emotions and trends. Without an integrated emotion analysis feature, administrators cannot effectively monitor the emotional impact of posts or detect patterns in user sentiment.
Difficulty in Identifying Distress or Negative Emotions: Since existing systems do not categorize emotions explicitly, identifying users who frequently post about distress, sadness, or negative emotions becomes challenging. Platforms rely on self-reporting or user reports rather than proactive detection, which limits their ability to provide necessary support or intervention when needed.
No Emotion-Based Graphical Analysis: Social networking platforms provide various data visualizations, such as post engagement metrics and activity graphs. However, they do not generate graphs or reports that categorize posts based on emotional sentiment, making it difficult to visualize and track emotional trends over time.
The existing online social networking systems serve their primary purpose of connecting users, sharing information, and facilitating interactions. However, they lack advanced sentiment analysis and emotion classification features. The absence of automated emotion detection and trend analysis restricts user experience and administrative insights, leaving a gap in understanding user emotions in real-time.

PROPOSED SYSTEM:

The proposed system aims to enhance the traditional Online Social Networking (OSN) platform by integrating an emotion classification mechanism that analyzes and categorizes user posts based on their emotional content. Developed using Java, with JSP, CSS, and JavaScript for the frontend and MySQL for the database, the system retains the core functionalities of an OSN while incorporating an additional feature to classify tweets into predefined emotional categories.
The system consists of two main entities: Users and Admin. Users can perform standard OSN activities such as creating a profile, posting tweets, viewing timelines, following other users, sending private messages, and managing their followers. The key enhancement is the ability to classify tweets into one of six emotional categories: Happy, Sad, Surprise, Anger, Disgust, or Neutral. This classification is based on a predefined dataset stored in the database, without the use of complex machine learning algorithms.
For administrators, the system provides additional functionalities to oversee user activity and analyze emotions in posted content. Admins can view all user details, access posted tweets, and perform tweet analysis, which categorizes each post based on its detected emotion. Additionally, a graphical representation of tweet distributions across different emotional categories is provided, offering insights into the emotional trends within the network.
By integrating emotion detection, the proposed system moves beyond basic user engagement, allowing for an improved understanding of user sentiment in social interactions. It enables a structured and lightweight approach to categorizing user emotions while maintaining the core features of a traditional OSN platform.

ADVANTAGES OF PROPOSED SYSTEM:

The proposed system enhances the traditional Online Social Networking (OSN) experience by integrating an emotion classification mechanism that categorizes user posts based on their emotional content. This feature brings several advantages, improving user engagement, administrative oversight, and sentiment analysis capabilities.
Automated Emotion Classification: Unlike traditional social networking platforms, the system automatically categorizes user tweets into predefined emotional categories (Happy, Sad, Surprise, Anger, Disgust, and Neutral). This eliminates the need for manual interpretation and enhances the understanding of user sentiment.
Enhanced User Experience: By classifying posts based on emotions, the system provides users with a more interactive and insightful experience. Users can gain better awareness of their emotional expressions in their posts, leading to more meaningful engagements.
Efficient Sentiment Analysis for Administrators: The system provides the admin with the ability to analyze tweets based on emotions, helping in monitoring user interactions and trends. This feature enables better moderation and oversight of platform activities.
Graphical Representation of Emotional Trends: The system includes a dynamic graphical representation of tweets categorized by emotions. This feature helps in visualizing sentiment trends over time, offering useful insights into user behavior and emotional patterns.
Lightweight and Easy to Implement: The system does not rely on complex machine learning algorithms, making it lightweight and easy to deploy. This approach ensures that sentiment analysis can be performed efficiently without requiring extensive computational resources.
Better User Interaction and Engagement: By analyzing emotions, users can relate better to posts and interactions. This emotional classification encourages users to engage with content that resonates with their mood, fostering a more personalized social networking experience.
Improved Moderation and Safety: The emotion classification system allows administrators to identify negative sentiments such as anger or sadness, helping them take appropriate action if necessary. This feature contributes to a safer and more positive online environment.
Real-Time Emotion Analysis: As tweets are posted, they are analyzed and classified in real time. This immediate categorization enables quick insights into the overall emotional state of users on the platform.
The proposed system enhances the conventional OSN model by introducing emotion detection, providing users with a more engaging experience while giving administrators valuable insights into sentiment trends. With its lightweight implementation and graphical analysis, the system improves user interaction, content moderation, and overall platform intelligence.

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

Predicting Emotions of User in Online Social Network