Echo Insights: YouTube Sentiment Analysis App
In an era where social media is getting a wider attraction seeing an improvement in content creators than consumers, ECHO INSIGHTS is a tool for analyzing YouTube video sentiments. This mainly focusses on creating an analytical tool that inputs a YouTube video link from user and uses YouTube API V3 to fetch comments from the video and organize them into a dataframe using pandas. For global audacity, these extracted comments are translated into English using Google trans library in Python. Subsequently, each of the fetched results are assigned with sentiments using Vader, a pre trained sentiment analysis model with a better accuracy of 96%. Then the analyzed results are displayed in a intuitive dashboard designed using streamlit enabling users to interpret sentiment patterns in the content. The dashboard offers various visualization contents like, good and bad comments based on Vader sentiment providing a clear view on sentiments of comments. Additionally, a summarized sentiment score is also displayed, providing general tone of the video. One of the additional features of out webapp is Wordcloud, a python library that plots most comments used or relevant words in the fetched comments representing a quicker glance in video engagement and users’ opinions on video. ECHO INSIGHTS aims to empower content creators with better analysis and a healthier component and understanding of audience feedback and engagement with videos. Overall enabling users and creators with continuous improvement and optimization of their content, thus acting as a goof asset by translating raw comments till overall sentiment facilitating deeper connection with audience and crafting contents more efficiently.
Authors
- Kunamineni Vijay
- M Jaeyalakshmi
- Sharan Adhitya C D
- K Senthilnaathan
Institutions
- Rajalakshmi Engineering College
Publication Details
- Published
- 2024-09-13
- DOI
- https://doi.org/10.1109/icietdw61607.2024.10939985
- Primary Topic
- Sentiment Analysis and Opinion Mining
- Type
- article
- Field-Weighted Citation Impact
- 0.00