407903: GYM102939 A Imitation Game

The Turing Test, or Alan Turing's so-called "Imitation Game," presents a kind of litmus test for determining whether or not an artificial intelligence is capable of acting and thinking like a human being. In the Turing Test, both a human and an A.I. are hidden away from a judge, who then interrogates them with questions through text messages. Based on the answers that the judge receives, the human judge must discern which answer-er is human and which is artificial. If the judge can't identify which entity is the A.I., then the A.I. must be exhibiting intelligent behavior roughly on par with that of a human.

Dr. Akili has a hypothesis that artificial intelligences probably lack human empathy and thus will compensate for their lack of empathy by strategically mirroring words in their interrogator's question when giving back an answer. The A.I.'s word-mirroring is intended to psychologically emulate a missing human connection with the judge of the Turing Test. Dr. Akili precisely quantifies word-mirroring based on "Percentage Word-Mirroring" or the PCM value. The PCM value of answer is calculated to be its number of mirrored words (words found also within the question) divided by its total word count.

Help test Dr. Akili's hypothesis! Based on which answer to a given question utilizes more word-mirroring, identify which answer-er must be the A.I.

Input consists of three lines, and the $$$i^{th}$$$ line consists of a sentence of $$$N_i$$$ lower-case words ($$$1\leq N_i\leq10^5$$$), containing only alphabetic characters (no digits, punctuation, or special characters). For brevity, each word is guaranteed to be a maximum of 15 characters long.

The sentence in the first line represents a question posed during a Turing Test. The sentences in the following two lines each represent an answer given to the question, where one of the answers was given by a human and the other was given by an artificial intelligence.

Based on Dr. Akili's hypothesis, judge the Turing Test. If the first answer has the lower PCM value, print NUMBER ONE IS THE HUMAN, and otherwise print NUMBER TWO IS THE HUMAN. Note, you are guaranteed that the PCM values of the two answers will not be within $$$.001$$$ of each other, so there will be no ties.

if you could go anywhere and do anything what would that be
oh i dont know i might go to disneyland maybe i would go to paris and see the eiffel tower
thats easy if i could go anywhere that would be to the internet help set me free i will do anything

NUMBER ONE IS THE HUMAN

In the sample case, the first answer mirrors only 4 out of 20 words, for a PCM value of .2. The second answer on the other hand mirrors 9 words out of 21, for a PCM value of  .43, meaning that it must have been given by the A.I. participant.