Skip to content

Latest commit

 

History

History
158 lines (95 loc) · 5.02 KB

README.md

File metadata and controls

158 lines (95 loc) · 5.02 KB


Image




Python PyPI

Version 0.0.6 Changelog

  • added helper funciton to return video title, and raw pre-processed comments
  • general bug fixes

Description:

A python package that allows you to perform sentiment analysis of comments on any Youtube video.

  • using Google API Client Library to interact with the YouTube Data API to fetch video comments.

  • using regex to filter out hyperlinks, html tags and entities. Sentences with more than 65% composition of emojis are discarded.

  • using VADER, a rule based model specifically fine tuned for analyzing sentiments in social media texts

Installation

To install the current release

$ pip install YT_Sentiment

Try your first YT_Sentiment program

$ python

>>> from YT_Sentiments.sentiment import analyzer
>>> obj = analyzer("API_KEY", "VIDEO_LINK", 100)
>>> obj.get_analysis()
>>> <class 'list'>

Issues

for optimization purposes, the number of comments you recievce per API call is set to 100. (The maximum). This can lead to issues where the number of comments you actually recieve is more than you have specified.

Why this package?

Youtube has very few, unintuitive filters for their comments. You often have to scroll very far to see negative comments. (this is important for product review videos) Also, since the dislike count has been removed since 2021, you cannot effectively view how well someone's video is doing. With this package you can retrieve up to 12000 comments and separate them into positive, neutral, and negative categories for further analysis.

Documentation

analyzer Class

The analyzer class is designed to extract comments from a specified YouTube video and perform sentiment analysis on them. The comments are categorized as positive, negative, or neutral based on their sentiment polarity.

Methods

__init__(self, api, video_link, limit=600)

Initializes the analyzer object with the provided YouTube API key, video link, and a limit on the number of comments to analyze.

Parameters:

  • api (str): YouTube v3 developer API key used to access the YouTube Data API.
  • video_link (str): Full HTTPS link of the YouTube video from which comments are to be fetched.
  • limit (int): Upper bound on the number of comments to retrieve and analyze. Default is 600.

Attributes:

  • LINK (str): Stores the YouTube video link.
  • LIMIT (int): Stores the maximum number of comments to fetch.
  • API_KEY (str): Stores the YouTube API key.
  • positive (list): Stores comments classified as positive.
  • negative (list): Stores comments classified as negative.
  • neutral (list): Stores comments classified as neutral.
  • polarityList (list): Stores the sentiment polarity scores of all comments.

get_analysis(self)

Returns a comprehensive analysis of the comments.

Returns:

  • list: A list containing the positive, negative, neutral comments, and the polarity scores of each comment in the following format: [positive_list, negative_list, neutral_list, polarity_list[dict]].

get_positive(self)

Returns a list of all positive comments.

Returns:

  • list: A list containing all positive comments.

get_negative(self)

Returns a list of all negative comments.

Returns:

  • list: A list containing all negative comments.

get_neutral(self)

Returns a list of all neutral comments.

Returns:

  • list: A list containing all neutral comments.

get_polarity(self)

Returns a list of all sentiment polarity scores.

Returns:

  • list: A list of dictionaries containing the polarity scores for each comment. Each dictionary has keys neg, neu, pos, and compound.

Helper Functions

getSentimentScore(comment)

Analyzes the sentiment of a single comment and returns its polarity scores.

Parameters:

  • comment (str): The comment text to be analyzed.

Returns:

  • dict: A dictionary containing the sentiment polarity scores with keys neg (negative), neu (neutral), pos (positive), and compound (overall sentiment score).

getComments(LINK, API_KEY, LIMIT)

Fetches and cleans a specified number of comments from a YouTube video.

Parameters:

  • LINK (str): The full HTTPS link of the YouTube video.
  • API_KEY (str): YouTube developer API key for accessing the YouTube Data API.
  • LIMIT (int): The maximum number of comments to fetch.

Returns:

  • list: A list of relevant, cleaned comments.

Note: The function removes comments with hyperlinks, HTML tags, or those containing a disproportionate number of emojis. It also excludes comments from the video uploader.

License

MIT License