I’d like to make the case for a ‘Chat’ entity in FTM - a container for multiple Message entities, similar to how a Document is a container for Pages.
I think it could be very helpful when representing chat conversations and social media scrapes. I also believe that, long term, it would help with building a nice chat-like representation for these kinds of datasets into OpenAleph, something that feels very natural to a user.
Issues I currently see:
What if the chat message is not text?
Does the Chat entity contain metadata (who’s in the chat, from when to when, chat descriptions, etc)?
What level granularity should this metadata have (who read which group chat message when, etc)?
I know that this has been a topic of discussions before, looking forward to hearing what other people think!
I really want this as well Scraping social media and instant messaging apps is so heavily relied on and useful right now. And if the social media platforms refuse to give us viable tools to inspect the content (like proper Ad Libraries and a tool like CrowdTangle used to be), then we have to do it ourselves.
I’m thinking about this as replicating the UI of a chat as much as possible. Additionally, search through such stuff should also offer a UI experience which takes you not only to the message but also the messages before and after - in the case of a chat thread.
In the case of social media apps, metadata should be displayed in a way that mimics the UI we already experience: likes, shares, comments, reactions etc.
Super exciting feature from my point of view as well.