1
00:00:00,920 --> 00:00:02,890
- [Instructor] For our last demo of some

2
00:00:02,890 --> 00:00:04,240
of these Watson services,

3
00:00:04,240 --> 00:00:08,170
let's take a look at their
personality insights service

4
00:00:08,170 --> 00:00:11,940
which is capable of looking
at unstructured text,

5
00:00:11,940 --> 00:00:16,240
such as somebody's
tweets or Facebook posts

6
00:00:16,240 --> 00:00:17,330
or things like that

7
00:00:17,330 --> 00:00:22,330
and analyzing that text for
their personality traits.

8
00:00:22,460 --> 00:00:23,780
So, if you take a look

9
00:00:23,780 --> 00:00:26,380
at IBM's documentation on this service,

10
00:00:26,380 --> 00:00:30,120
they say that you can gain
insight into how and why

11
00:00:30,120 --> 00:00:33,100
people think, act, and
feel the way they do.

12
00:00:33,100 --> 00:00:36,750
So they use various linguistic
analysis techniques,

13
00:00:36,750 --> 00:00:39,740
they use personality theory,
and other capabilities

14
00:00:39,740 --> 00:00:42,450
to analyze that unstructured text.

15
00:00:42,450 --> 00:00:44,180
And one way that you might be able

16
00:00:44,180 --> 00:00:45,800
to use information like this,

17
00:00:45,800 --> 00:00:48,860
and of course, nowadays with
all the privacy concerns,

18
00:00:48,860 --> 00:00:50,950
you may not want to do things like this

19
00:00:50,950 --> 00:00:55,110
but you could potentially
use this type of a service

20
00:00:55,110 --> 00:00:57,820
to target your product advertising

21
00:00:57,820 --> 00:01:00,470
for instance, at the
people who are most likely

22
00:01:00,470 --> 00:01:02,410
to purchase those products.

23
00:01:02,410 --> 00:01:04,540
So, they have, of course, a demo of this

24
00:01:04,540 --> 00:01:06,470
along with many of their other services.

25
00:01:06,470 --> 00:01:08,600
Let me switch over to the demo.

26
00:01:08,600 --> 00:01:10,600
When you go to the demo page,

27
00:01:10,600 --> 00:01:13,300
you'll see that they have some options

28
00:01:13,300 --> 00:01:15,790
for trying the service
with tweets and replies

29
00:01:15,790 --> 00:01:18,670
for some people that are
well-known in the world

30
00:01:18,670 --> 00:01:20,250
like Oprah Winfrey.

31
00:01:20,250 --> 00:01:23,950
You can supply a body
of text or if you want,

32
00:01:23,950 --> 00:01:26,410
you can log into your own twitter account

33
00:01:26,410 --> 00:01:29,270
and they'll pull your recent tweets

34
00:01:29,270 --> 00:01:32,010
and analyze based on that.

35
00:01:32,010 --> 00:01:34,780
So, what I did just for demo purposes

36
00:01:34,780 --> 00:01:38,240
was try out the Oprah
Winfrey Twitter account.

37
00:01:38,240 --> 00:01:40,370
So if I go ahead and select that,

38
00:01:40,370 --> 00:01:42,020
looks like I have to
refresh the page here,

39
00:01:42,020 --> 00:01:43,900
so let's try it again.

40
00:01:43,900 --> 00:01:46,060
So select that and click analyze,

41
00:01:46,060 --> 00:01:49,670
and now, when it finishes analyzing,

42
00:01:49,670 --> 00:01:52,960
they give us back what they
call a personality portrait

43
00:01:52,960 --> 00:01:55,550
based on the number of words they analyze,

44
00:01:55,550 --> 00:01:57,900
they say this is a very strong analysis.

45
00:01:57,900 --> 00:02:01,380
Then, they give me a
summary of Oprah Winfrey

46
00:02:01,380 --> 00:02:03,700
and her personality traits.

47
00:02:03,700 --> 00:02:06,100
Not only in paragraph form,

48
00:02:06,100 --> 00:02:09,090
things that she's likely
and unlikely to do,

49
00:02:09,090 --> 00:02:12,260
but also a whole bunch
of interesting statistics

50
00:02:12,260 --> 00:02:17,260
as well about her personality
consumer needs and values.

51
00:02:17,340 --> 00:02:21,030
So for instance, this 94% here

52
00:02:21,030 --> 00:02:25,500
says that Oprah Winfrey is agreeable

53
00:02:25,500 --> 00:02:29,680
and more so than 94% of the population.

54
00:02:29,680 --> 00:02:34,550
And she's conscientious more
so than 81% of the population.

55
00:02:34,550 --> 00:02:36,160
So as you can see,

56
00:02:36,160 --> 00:02:40,140
this is kind of a more abstract concept

57
00:02:40,140 --> 00:02:43,860
in terms of the analysis of
the text that's going on.

58
00:02:43,860 --> 00:02:48,290
This is again another example
of natural language processing

59
00:02:48,290 --> 00:02:51,560
but at a much higher
level than the techniques

60
00:02:51,560 --> 00:02:53,400
that we demonstrated for you

61
00:02:53,400 --> 00:02:56,330
in the natural language processing lesson

62
00:02:56,330 --> 00:02:59,000
and even in a couple of the demos earlier

63
00:02:59,000 --> 00:03:01,670
in this presentation from IBM Watson.

64
00:03:01,670 --> 00:03:04,200
Now, I just wanna point out
before we finish up here,

65
00:03:04,200 --> 00:03:08,260
again, if you go to the
Watson services listing

66
00:03:08,260 --> 00:03:10,770
on their website and
click browse services,

67
00:03:10,770 --> 00:03:12,640
you'll see what all the services are

68
00:03:12,640 --> 00:03:15,010
and you'll see all the demo links

69
00:03:15,010 --> 00:03:17,600
for the services that
I've shown you so far

70
00:03:17,600 --> 00:03:20,580
and the other ones that
are available as well.

71
00:03:20,580 --> 00:03:22,420
I didn't cover all of those here

72
00:03:22,420 --> 00:03:24,250
and if you are interested,

73
00:03:24,250 --> 00:03:25,250
you can go ahead

74
00:03:25,250 --> 00:03:28,153
and take a look at some
of those other demos also.