I’d try the CSV imports first if that’s the collection information, unless it really contains data protection information then don’t. After try the others containing JSON with watched and ratings etc. I would imagine that the TVTime exports would have some kind of ID linking the content of the separate files such as an IMDB ID etc that Trakt would be able to use to link the separate bits of data together.