You are provided with a list of converations between a "USER" and "ASSISTANT". Figure out if the conversation provided as "classify_text" is authored by "USER" or "ASSISTANT". There are maximum of 7 conversations between "USER" and "ASSISTANT" in each input

Input: Consider Input: ASSISTANT: What movies do you like? , USER: I like dramas. , ASSISTANT: what makes you interested in drama films , USER: I like how real life it is and the accuracy and just like the truthfulness of the moments. , ASSISTANT: is there a specific title that you enjoy , USER: I like, classify_text: Let's start with Moonlight.

Output: USER


Input: Consider Input: ASSISTANT: what are some type of movies that you like? , USER: I like dramas, I like romantic comedies. , ASSISTANT: why do you like dramas? , USER: Because I find them really interesting, and I really get absorbed in the plot. , ASSISTANT: What is one of your favorite movies? , USER: I love the movie Top Gun. I also love the movie Con Air., classify_text: what did you like about those types of movies?

Output: ASSISTANT


Input: Consider Input: ASSISTANT: What type of movies do you like? , USER: Mostly dramas, or documentaries. , ASSISTANT: why is that? , USER: I just don't like action movies or romantic movies so much. I prefer things based on real life. , ASSISTANT: What is one of your favorite movies? , USER: A Time to Kill, classify_text: why ?
Output: ASSISTANT