1
00:00:00,000 --> 00:00:00,367


2
00:00:00,367 --> 00:00:01,450
ZAMYLA CHAN: Tweet, tweet.

3
00:00:01,450 --> 00:00:04,500
What sentiments do your Tweets evoke?

4
00:00:04,500 --> 00:00:07,450
In Tweets, in this problem,
we're going to allow the user

5
00:00:07,450 --> 00:00:10,210
to pass in a Twitter user handle.

6
00:00:10,210 --> 00:00:14,680
And based on that handle, we're
going to retrieve that user's Tweets,

7
00:00:14,680 --> 00:00:18,620
evaluating whether those Tweets
are generally positive, negative,

8
00:00:18,620 --> 00:00:22,390
or neutral, and printing them
out with the appropriate colors.

9
00:00:22,390 --> 00:00:27,280
Green for positive, red for
negative, and yellow for neutral.

10
00:00:27,280 --> 00:00:28,880
So what do we need to do?

11
00:00:28,880 --> 00:00:32,930
Well, first, we want to ensure
proper usage, get the users' Tweets,

12
00:00:32,930 --> 00:00:36,670
initialize Analyzer, and then
finally, analyze those Tweets.

13
00:00:36,670 --> 00:00:40,600
In order to ensure a proper usage,
take a look back at the implementation

14
00:00:40,600 --> 00:00:43,660
of Smile to get a hint there.

15
00:00:43,660 --> 00:00:46,690
Then in order to get
the Tweets, we're going

16
00:00:46,690 --> 00:00:51,820
to look into helpers.py, where the
function get_user_timeline allows

17
00:00:51,820 --> 00:00:54,970
us to retrieve a user's Tweets.

18
00:00:54,970 --> 00:00:57,740
We want to check whether
that is successful or not.

19
00:00:57,740 --> 00:01:02,300
So make sure to look at the function
to see what it returns if it fails.

20
00:01:02,300 --> 00:01:05,650
So for instance, what if the
user name that the user provided

21
00:01:05,650 --> 00:01:08,920
doesn't exist or perhaps
is a private account?

22
00:01:08,920 --> 00:01:13,060
If it's unsuccessful, then you'll want
to exit with an appropriate statement

23
00:01:13,060 --> 00:01:15,140
to the user.

24
00:01:15,140 --> 00:01:17,710
Next, you'll want to tokenize the Tweet.

25
00:01:17,710 --> 00:01:19,660
We're, again, going
to look at Tokenizers

26
00:01:19,660 --> 00:01:22,300
as we did before in Analyzer.

27
00:01:22,300 --> 00:01:25,930
And these Tokenizers are part
of the natural language toolkit.

28
00:01:25,930 --> 00:01:28,780
Specifically, there's
actually a TweetTokenizer

29
00:01:28,780 --> 00:01:33,670
that allows us to split the Tweet
into a list of words, shorter strings.

30
00:01:33,670 --> 00:01:37,180
Once we have that, we can
instantiate our Analyzer

31
00:01:37,180 --> 00:01:40,960
and iterate over every token,
scoring every token as either

32
00:01:40,960 --> 00:01:42,940
positive, negative, or neutral.

33
00:01:42,940 --> 00:01:46,660
Then finally, that sum is going
to indicate whether that Tweet

34
00:01:46,660 --> 00:01:50,000
is positive or negative.

35
00:01:50,000 --> 00:01:52,100
That's all the hints
I have for you today.

36
00:01:52,100 --> 00:01:55,830
My name is Zamyla and this was Tweets.

37
00:01:55,830 --> 00:01:57,979