This document contains a description of all included data files.

=====================================================

FILE NAME: Comment-Sentiment.csv.zip

A collection of all comments with accompanying sentiment score, parent, lag, and cardinality data.

COLUMN NAMES:

video_nid
A unique ID for the parent video.

comment_nid
A unique ID for the comment.

sentiment_pos
The positive sentiment value of the comment as determined by SentiStrength.

sentiment_neg
The negative sentiment value of the comment as determined by SentiStrength.

parent_nid
The unique ID of the parent comment if the row is a reply.

cardinality
The order that the reply comes in the thread.

like_count
The number of times the comment/reply was liked.

parent_sentiment_pos
The positive sentiment value of the reply's parent comment as determined by SentiStrength.

parent_sentiment_neg
The negative sentiment value of the reply's parent comment as determined by SentiStrength.

replies_siblings
The number of sibling replies in the thread.

sentiment_pos_lag_X
The positive sentiment of the previous reply in the thread with a lag of X (up to 4).

sentiment_neg_lag_X
The negative sentiment of the previous reply in the thread with a lag of X (up to 4).

video_format
Whether the parent video's format was a typical TED Talk or an Animation.

presenter_gender
The gender of the parent video's presenter (not available for Animations).

row_type
Comment (if the row is a response to the parent video) or reply (if the row is a response to a parent comment).

valid
Binary indicating whether the row had sufficient data for analysis (e.g., sentiment scores).

=====================================================

FILE NAME: Comment-ID-Cross-Reference.csv.zip

Cross-referencing data for connecting Comment-Sentiment.csv IDs to each comment's unique YouTube ID.

COLUMN NAMES:

comment_nid
A unique ID for the comment.

comment_id
The YouTube unique ID for the comment.

cardinality
The order that the reply comes in the thread.

=====================================================

FILE NAME: Video-ID-Cross-Reference.csv

Cross-referencing data for connecting Comment-Sentiment.csv IDs to each video's unique YouTube ID.

COLUMN NAMES:

video_nid
A unique ID for the video.

video_id
The YouTube unique ID for the video.

=====================================================

FILE NAME: Video-Data.csv

Codes for videos.

COLUMN NAMES:

video_id
The YouTube unique ID for the video.

format
Whether the video's format was a typical TED Talk or an Animation.

gender
The gender of the video's presenter (not available for Animations).

language
An identification of non-English videos.

valid
Binary indicating whether the row had sufficient data for analysis (e.g., open for commenting, English).

=====================================================

FILE NAME: Comments.csv.zip

Textual data for comments.

COLUMN NAMES:

videoId
The YouTube unique ID for the video.

id
The YouTube unique ID for the comment.

parentId
The YouTube unique ID for the parent comment.

authorDisplayName
The YouTube display name of the comment author.

likeCount
The number of likes the comment received.

publishedAt
The time the comment was published.

publishedTS
A timestamp version of the publishedAt column.

replies
The number of replies the comment received.

isPublic
Whether the comment is publicly visible.

textDisplay
The display version of the comment text.

textPlain
The plain version of the comment text.