2022-10-26
Team Members
- Snehal Bailmare
- Shiwani Yadav
- Satwinder Kaur
- Jahnawi Chanamolu
Data
- We are using Youtube API V3 to collect data.
- Resulted csv files of Music,Sports and News will be used as input.
- Input Variables :Video id,text display,text original,user name,video url,channel url, channel name,no of likes and ratings
- Output Variables :Positive,Negative and Neutral.
- Data wrangling techniques will be used to clean and transform the data.
Problem Description
- To use sentiment analysis/text analysis technique to determine whether the users comments are a positive, negative or neutral and whether comments are parallel with the content they are watching.
- In online video platform youtube is the primary choice for everyone and having a wide audience thousands of videos and comments are uploaded every minute.
- We will understand the sentiments expressed in comments to see how they will effect the brand.
Analytics Plan
- To execute Natural Language Processing for classification of comments as positive, negative or neutral.
- Visualization techniques like ggplots will be use to analyse the data.
Evaluation
- Based on predicted models accuracy for different data sets we will predict the best model.
- Our evaluation will also include whether our data is biased towards positive, negative or neutral sentiments.