1
00:00:04,500 --> 00:00:07,570
In this lecture, we have
seen a particular application

2
00:00:07,570 --> 00:00:10,230
of sentiment
analysis on Twitter.

3
00:00:10,230 --> 00:00:13,940
However, the area of sentiment
analysis is much broader.

4
00:00:13,940 --> 00:00:17,770
Over 7,000 research articles
have been written on the topic.

5
00:00:17,770 --> 00:00:21,570
Hundreds of start-ups are
developing sentiment analysis

6
00:00:21,570 --> 00:00:22,610
solutions.

7
00:00:22,610 --> 00:00:25,630
Many websites perform
real-time analysis of tweets.

8
00:00:25,630 --> 00:00:27,870
For example,
"tweetfeel" shows trends

9
00:00:27,870 --> 00:00:30,300
given any term, and
"The Stock Sonar"

10
00:00:30,300 --> 00:00:34,820
shows sentiment
and stock prices.

11
00:00:34,820 --> 00:00:38,980
Let's talk about text
analytics a bit more generally.

12
00:00:38,980 --> 00:00:41,210
Sentiment analysis is a
particular application

13
00:00:41,210 --> 00:00:42,420
of text analytics.

14
00:00:42,420 --> 00:00:45,750
In general, the critical
aspect of text analytics

15
00:00:45,750 --> 00:00:47,960
is to select the
specific features that

16
00:00:47,960 --> 00:00:51,030
are relevant in a
particular application.

17
00:00:51,030 --> 00:00:56,660
In addition, it's important to
apply specific knowledge that

18
00:00:56,660 --> 00:00:58,360
often leads to better results.

19
00:00:58,360 --> 00:01:03,470
For example, using the meaning
of the symbols or include

20
00:01:03,470 --> 00:01:05,390
features like the
number of words.

21
00:01:08,450 --> 00:01:10,870
Let's finally discuss
the analytics edge

22
00:01:10,870 --> 00:01:13,640
that we have seen
in this lecture.

23
00:01:13,640 --> 00:01:16,170
Analytical sentiment
analysis we have seen

24
00:01:16,170 --> 00:01:20,800
can replace more labor-intensive
methods like polling.

25
00:01:20,800 --> 00:01:24,150
Text analytics can also deal
with the massive amounts

26
00:01:24,150 --> 00:01:27,789
of unstructured data being
generated on the internet.

27
00:01:27,789 --> 00:01:29,590
Computers are
becoming more and more

28
00:01:29,590 --> 00:01:31,740
capable of interacting
with humans

29
00:01:31,740 --> 00:01:34,550
and performing human tasks.

30
00:01:34,550 --> 00:01:37,560
In the next lecture,
we'll discuss IBM Watson,

31
00:01:37,560 --> 00:01:41,789
an impressive feat in the
area of text analytics.