Project Proposal: Emotion Classification for Social Media Posts BACKGROUND: Social media platforms like Instagram, YouTube, and Twitter play a major role in shaping public opinion. However, on some platforms, features such as dislike counts are either hidden, unavailable, or absent altogether. This can make it difficult for content creators, brands, and everyday users to understand the emotional reactions to their posts. Despite the absence of explicit dislike indicators, analysing the text interactions (comments, replies) can provide valuable insight into how a post is perceived emotionally by the audience. This project seeks to address this gap by using automated sentiment analysis to classify emotions surrounding social media content.

AIMS AND OBJECTIVES: The primary aim of this project is to develop an application that allows users to detect and classify the emotional sentiment surrounding social media posts. By analysing user-generated text data such as comments or replies, the system will provide an indication of whether a post is generally perceived e.g. happy, sad, angry. The specific objectives include:

Building a system that extracts text data from social media platforms based on user-provided links. Developing a model to classify emotions e.g. (happy, sad, angry) from the extracted text data. Providing an accessible user interface that allows individuals to input a social media post link and receive emotion classification results. Offering insight into public sentiment for platforms where explicit dislike counts are absent or hidden. METHOD: The proposed method for solving this problem involves the following steps:

Data Collection: The system will retrieve text data (comments, replies) from social media posts. This will involve extracting relevant textual interactions from user-provided links. Text Processing: The collected text data will be processed and cleaned to prepare it for emotion classification. This includes tasks such as tokenization, stop word removal, and other natural language processing techniques. Model Development: A machine learning/deep learning model will be developed and trained to classify the emotional sentiment of the processed text data. The model will predict whether the overall sentiment is positive, negative, or neutral based on patterns in the text. User Interface: An interface will be created to allow users to input social media post links and receive the sentiment analysis results in a clear and accessible format. Testing and Evaluation: The system will be tested using real-world social media posts to ensure that the emotion classification model performs accurately and provides valuable insights into public sentiment.