YouTube Sentiment-Analysis

With the prevalence of social media, people are more familiar with publicly sharing their opinions these days. Some people even become content creators on major social media platforms, such as Instagram and YouTube. Therefore, these websites are ideal resources for sentiment analysis. By deploying the technique, the efficiency of social monitoring and brand marketing are enhanced. Creators and brands can perceive the viewers' emotion provoked by the content they create, and further decide whether there are adjustments to make.

Methodology:

1. Description
2. Data collection
3. Objective Statement
4. Prerequisite
5. Data Preprocessing
6. Analyze the sentiments
7. Conclusion

Description:

This project works by scraping YouTube comments and identify the sentiment of comments and shows the WordCloud of most frequent words/emojis and positive/negative sentiment.

Basic sentiment analysis of comments on a youtube video using a builtin python package "TextBlob Sentiment Analyser".

Objective:

• Performing Sentiment analysis
• WordCloud of your positive & negative sentences
• Performing Emoji's analysis
• Collecting the Entire data of Youtube
• Analysing the most liked category
• Finding out whether audience is engage or not
• Analysing the Trending videos

Prerequisite :

pandas package :

$ sudo pip install pandas

seaborn package :

$ sudo pip install seaborn

matplotlib package :

$ sudo pip install matplotlib

plotly package :_

$ sudo pip install plotly

wordcloud package :

$ sudo pip install wordcloud

textblob package :_

$ sudo pip install TextBlob

emoji package :_

$ sudo pip install emoji

Analyze the sentiments with stop words using WordCloud and Plotly.

Full report of this project: Sentiment Analysis

* positive comments

* Top 20 Most-Frequent Emojis

* Like-rate according to the categories

* Trending-Videos

Conclusion:

The discussion demonstrates that sentiment analysis is still subject to the context. Without the context, emotions cannot be precisely expressed. Stopword is proven by the study that it is one of the aspects that we can still work on.

From this study, we learn lots of sentiment analysis tools that are handy to use. Although it did not provide 100% accuracy when interpreting the emotions, it saved our time, and showed its potential in the combination of natural language processing and linguistics.