Hi, I have a question about the data labeling: why the root tweet/reddit post is labeled?
For example, in the "charliehebdo" topic of the tweet-english data, the root tweet with id: 552783667052167168 is labeled as "support" in the train-key.json file.
This is also true for the reddit dataset.
I believe that: only repost should be labeled, because it shows some stance towards the root tweet/reddit post.
And the task description is like this: "Therefore, we introduce a subtask where the goal is to label the type of interaction between a given statement (rumourous post) and a reply post (the latter can be either direct or nested replies). "
Could someone tell me why ?