WEBVTT

00:00.950 --> 00:08.300
OK, so this particular project is actually an text processing project, so in this particular project,

00:08.300 --> 00:15.000
we will be walking the text and this is the data set which we have for jobs, which is basically the

00:15.000 --> 00:16.490
IMDB dataset.

00:16.830 --> 00:22.540
This contains movie reviews and there will be 50000 reviews, just media.

00:22.650 --> 00:25.080
So we have 50000 text reviews.

00:25.920 --> 00:32.820
And all along with that, we have one volume of sentiment which contains value, either positive or

00:32.820 --> 00:33.270
negative.

00:34.170 --> 00:43.830
So you will have to convert this entire text into some kind of vector, which we have already known

00:43.860 --> 00:45.570
as part of the MLP session.

00:46.350 --> 00:54.300
And people have to convert the sentiment into either a label included or, you know, dummy form that

00:54.300 --> 00:55.310
is 014.

00:56.400 --> 01:02.520
And then he can simply apply any classification module on top of it.

01:03.480 --> 01:08.520
So I hope they will be able to implement this based on the knowledge which we have got during the end

01:08.520 --> 01:09.210
of the recession.

01:10.560 --> 01:17.430
And in case you need any help, you can go ahead and have a look at the implementation which has been

01:17.430 --> 01:17.900
shared.

01:18.210 --> 01:25.920
But I would suggest to try this at your own first and after that, have a look at the solution once

01:25.920 --> 01:27.050
you have tried it.