A、Collecting all the matches. B、Tagging the matches by hand. C、Giving an effective pattern to the coach. D、Generating a playlist

The crowds arrive daily at Flushing Meadows Park in Queens. It’s the US Open and all eyes are on the top tennis players in the world. With 254 matches just in the men’s and women’s singles tournaments, it’s a lot to keep up with, but behind the scenes, IBM technology is doing just that.
    “So this gets updated at the end of that point. We see how Watson evaluate at that point.”
     With so much data to sort through, IBM is employing its artificial intelligence software Watson to help. In addition to delivering scores, Watson also scans game footage.
    “It listens to the roar of the crowd. We’ve trained it to look for fist pumps and gestures of the players as well as to look for facial expressions.”
    All of that contextual data is used to measure the momentum of any given match after each shot.
    “A few years ago, we didn’t really perceive video as data, but it really is.”
    “There’s 17 courts, so it’s a lot of action; it’s a lot of video for the editors to digest. So now that you can quickly identify the most exciting moments and turn it around in near real time for fans.”
    These highlights aren’t just for fans. Coaches also use them to prepare players.
    “In the past, it was a really labor-intensive process where the matches had to be manually tagged, and by tagging, I mean somebody goes through the video footage and makes a note of every forehand, every backhand, every unforced error. It takes hours and hours.”
    Watson’s artificial intelligence processes and indexes video in minutes, freeing up valuable time for personalized coaching.
    “If the coach has a certain pattern in mind that they think will be effective in the next match, we can generate a playlist that shows the player executing those patterns and really reinforce that in their mind.”
    Players’ facial expressions and body language can also be correlated to their performance.
    “Even the way their grunting, there’s certain trends and you know sometimes, the counterintuitive, some players may play better when they get a little bit angry, when they get a little upset.”
    With the help of artificial intelligence, perhaps getting a little upset will lead to more upsets on the court.
    Question 19. What does Watson refer to according to the speaker?
    Question 20. What should be done in the past when there was no Watson?
    Question 21. What is said about the players who are a little bit angry?