Dear Forum,
I am an academic researcher writing an article using the new Twitter API v. 2. Among the data obtained through the Twitter API v2, I am using the context annotations data. I have two questions regarding it:
- I understand Twitter privacy concerns, but I would like to know if there is some technical detail about context annotations functioning and a source that can be cited about it. As far as I understand, it is an automated (trained?) classification algorithm, but if some more detail would be available it could be of help for researchers;
- I have a doubt about the interpretation of the domains and the entities associated with a tweet. I saw that one tweet can have multiple domains and entities associated, in some cases even 30 or 40. I wanted to understand: in these cases, the domains and entities are to be interpreted as hierarchical or as equal? In other words, the first domain and entity are the principle ones and all the subsequent ones are proportionally less relevant, or they are all equally related to the tweet? Or they are indented, meaning that the subsequent domains and entities are sub-categories of the firsts?
1 Like
