Research Tools
Discourse Analysis
Semantic discourse categories assigned by Claude Haiku — 14 DH-informed categories classifying how commenters frame their arguments.
Comments Labeled
44,412
of 487,977 total
Coverage
9.1%
of eligible comments
Categories Active
14
of 14 discourse types
Avg Confidence
0.807
classifier confidence
Model
Claude Haiku
multi-label classification
Search & Filter
Meta-Commentary
3,846 comments
CNN still here?
CNN still here?
Comment about media presence (CNN) in the video or comments section, not about immigration policy itself.
Meta-Commentary
0.90
@snjspring
Mar 4, 2025
wow 2 million in 4 hours
wow 2 million in 4 hours
Comments on the video itself, noting a metric (2 million in 4 hours) likely referring to views or engagement.
Meta-Commentary
0.90
@thanosianthemadtitanic
Mar 4, 2025
Can't argue with that .
Can't argue with that .
Generic agreement with a previous comment without adding substantive argument or position.
Meta-Commentary
0.90
@dominicthompson1909
Mar 4, 2025
This comment section is racist.
This comment section is racist.
Direct criticism of the comment section itself rather than engaging with immigration topic.
Meta-Commentary
0.90
SmartVidsEz
Feb 5, 2026
F** you you remove Hindi dubbing, 🤬🤬🤬🤬🤬🤬🤬🤬
F** you you remove Hindi dubbing, 🤬🤬🤬🤬🤬🤬🤬🤬
Complaint about video platform features (dubbing removal) rather than immigration policy or content.
Meta-Commentary
0.90
ShrivasSahil
Feb 4, 2026
? well said
? well said
Brief affirmation of a previous comment's point without substantive content, meta-engagement with other commenters.
Meta-Commentary
0.90
@gracielacano6706
Mar 4, 2025
Can you shut up and let us here the questions/answers??? I hate these dumb as commentators!!!
Can you shut up and let us here the questions/answers??? I hate these dumb as commentators!!!
Criticism of the video's commentators and request for better audio quality, directly addressing the video production.
Meta-Commentary
0.90
@darrylmathews7783
Mar 4, 2025
0:39 as an American I can’t make it make sense.
0:39 as an American I can’t make it make sense.
References a specific timestamp in the video and expresses confusion about video content.
Meta-Commentary
0.90
@newbreezy100
Mar 4, 2025
Os vídeos estão demorando muito para serem dublados por IA para português 😢 🇧🇷🇧🇷🇧🇷
Os vídeos estão demorando muito para serem dublados por IA para português 😢 🇧🇷🇧🇷🇧🇷
Comment about video production delays for Portuguese AI dubbing, directly addressing the channel's content delivery rather than immigration policy.
Meta-Commentary
0.90
JuniorKirchoff
Feb 11, 2026
????????From the USA
????????From the USA
Minimal content indicating geographic location and emoji reaction, functioning as meta-commentary on the video or discussion.
Meta-Commentary
0.90
@dsworld8721
Mar 4, 2025
Omg what a speach, ouch
Omg what a speach, ouch
Brief reaction to the speech itself, commenting on its quality and emotional impact.
Meta-Commentary
0.90
@zdravkoavdalovic3131
Mar 4, 2025
Not we...him.
Not we...him.
Brief correction or disagreement with a previous commenter's use of 'we' versus 'him'.
Meta-Commentary
0.90
@danielpulvermacher7707
Mar 4, 2025
You must learn to BUT OUT when news is happening in front of you. Youbreoeated what we just heard while the Prime Minister was actively answering questions. Your analysis is less important than the news.
You must learn to BUT OUT when news is happening in front of you. Youbreoeated what we just heard while the Prime Minister was actively answering questions. Your analysis is less important than the news.
Commenter critiques the video creator's editing and commentary choices during the news event itself.
Meta-Commentary
0.90
@melstewart3310
Mar 4, 2025
??????????????????
??????????????????
Question marks alone suggest confusion or commentary on the video/discussion itself rather than substantive engagement.
Meta-Commentary
0.90
@ashleighjaimaosborne3966
Mar 4, 2025
2:05 this guy gets it
2:05 this guy gets it
References a specific timestamp in the video and expresses agreement with a speaker, focusing on the video content itself.
Meta-Commentary
0.90
The_Hollow_Entity
Jan 27, 2026
This was soo soo good!!!!❤
This was soo soo good!!!!❤
Expresses approval of the video content itself with enthusiastic praise.
Meta-Commentary
0.90
@parittramandal2797
Mar 4, 2025
President Trump called him governor, so why can't he call him Donald ?
President Trump called him governor, so why can't he call him Donald ?
Questions the appropriateness of Trudeau calling Trump by his first name, referencing Trump's prior disrespectful address of Trudeau.
Meta-Commentary
0.90
@devindersingh5729
Mar 4, 2025
guy on the bike has no goddamn clue what he’s talking about just like every racist
guy on the bike has no goddamn clue what he’s talking about just like every racist
Dismisses another commenter's credibility by labeling them racist, focusing on the person rather than substantive debate.
Meta-Commentary
0.90
c0rps3bunny
Feb 9, 2026
The interviewer is insufferable.
The interviewer is insufferable.
Critiques the interviewer's demeanor rather than discussing immigration policy itself.
Meta-Commentary
0.90
i8fish
Feb 24, 2026
CNN gets less views than reruns of Fear Factor WTF ARE THEY LISTED ON PAGE 1 for?!?!?! Garbage fake news and bias beyond belief, it’s not news it’s propaganda
CNN gets less views than reruns of Fear Factor WTF ARE THEY LISTED ON PAGE 1 for?!?!?! Garbage fake news and bias beyond belief, it’s not news it’s propaganda
Comment criticizes media platform algorithms and news source credibility rather than addressing immigration policy directly.
Meta-Commentary
0.90
@jobegerald6996
Mar 4, 2025
Compare Videos
Select 2–4 videos to compare their discourse profiles side by side (% of labeled comments).
Discourse Taxonomy
Category Distribution
Primary discourse type assigned to each comment
Classifier Confidence
Distribution of confidence scores across all labeled comments
Discourse Over Time
Top 6 discourse types by month — how framing evolves over time
Category Co-occurrence
Which discourse types most frequently appear together on the same comment
Top Channels by Discourse Type
Identity Assertion
CNN
2,313
Tyler Oliveira
1,991
RocaNews
467
JamCan
274
Yoke Immigration Canada
89
Solidarity
CNN
4,297
Tyler Oliveira
400
RocaNews
75
JamCan
36
CTV Your Morning
32
Policy Critique
CNN
3,445
Tyler Oliveira
551
Toronto Sun
340
CTV News
268
RocaNews
235
Humor / Satire
CNN
2,078
Tyler Oliveira
1,667
RocaNews
241
Global News
89
CTV News
68
Fear / Threat
Tyler Oliveira
1,549
CNN
1,332
RocaNews
321
JamCan
204
Toronto Sun
118
Meta-Commentary
Tyler Oliveira
1,702
CNN
1,478
RocaNews
168
CTV News
81
JamCan
77
Unverified Claim
CNN
1,480
Tyler Oliveira
820
RocaNews
203
Yoke Immigration Canada
203
JamCan
152
Moral Argument
CNN
1,733
Tyler Oliveira
704
RocaNews
205
JamCan
73
CTV Your Morning
23
Classified by Claude Haiku (Anthropic)
|
14-category DH taxonomy
|
Search
|
Discourse Quality (Perspective API)