WEBVTT 00:01.660 --> 00:05.350 Hello, everyone, and welcome to the world of natural language processing. 00:06.130 --> 00:11.560 So let's see what natural languages are in our daily life. 00:11.890 --> 00:18.370 We do communicate with each other and all those communication happens inside the natural language. 00:18.670 --> 00:24.370 Heider in terms of tax, in terms of speech, we are all surrounded by the text. 00:24.640 --> 00:28.020 Like we keep sending e-mails as a mess. 00:28.270 --> 00:29.620 We keep reading Web page. 00:29.950 --> 00:33.700 And everything has been surrounded by texts. 00:36.270 --> 00:38.070 Now, just think about the speech. 00:39.300 --> 00:47.060 We speak to each other like we watch television, listening to the radio, like we use the smartphone 00:47.100 --> 00:53.370 views the Aleksa, we use the Google now or Microsoft Cortona. 00:53.850 --> 01:00.570 So every where there is a natural language, in one hand, natural language processing is all about 01:01.530 --> 01:07.650 finding some hidden, meaningful insight out of those natural languages. 01:08.610 --> 01:15.610 And this close is all about finding those useful insights from the texts, how you can do prepossessing 01:15.660 --> 01:21.450 from the texts, and how you can apply some advanced machine learning and a deep learning algorithm 01:21.870 --> 01:24.090 on the top up tax data. 01:27.130 --> 01:31.540 So natural language processing is a field of artificial intelligence. 01:32.020 --> 01:40.060 And here, mainly humans and computers are interacting in terms of natural language, not in terms of 01:40.270 --> 01:41.410 computer languages. 01:42.790 --> 01:46.390 Now, NLB natural language processing, a sort of revision of. 01:47.730 --> 01:55.560 Natural language processing, which has an ability to go to machine like how machine can read, how 01:55.560 --> 01:56.730 Machine can understand. 01:57.450 --> 02:01.680 And something like derives some meaning from the human languages. 02:01.980 --> 02:05.570 And that is what the main object to behind and epiphany. 02:06.910 --> 02:14.080 Most of this natural language processing technique relying on machine learning and deep learning related 02:14.080 --> 02:16.030 algorithm to find those meaning. 02:16.440 --> 02:20.530 And we learn about a lot of machine learning and deep learning implementation. 02:20.770 --> 02:23.460 Horner Text Data in discourse. 02:25.040 --> 02:30.890 So let's see some of the applications of natural language processing like machine translation. 02:31.700 --> 02:32.450 So, as we know. 02:33.880 --> 02:37.210 Google Translate what I would say, Google machine translation. 02:37.540 --> 02:45.370 It can do the translation between hundreds of different paths of human language with such a high accuracy. 02:45.700 --> 02:52.530 And that become possible with a recent architecture of neural machine translation kind of architecture. 02:54.550 --> 02:55.750 Speech recognition. 02:56.230 --> 03:05.080 So in our daily life, we keep using of a smartphone, our smartphone can recognize your wise with very 03:05.080 --> 03:06.650 high accuracy nowadays. 03:07.030 --> 03:10.620 So your Alexa have publicity. 03:11.020 --> 03:18.850 Google now or Microsoft Cortona, they can detect and recognize your voice with very high accuracy. 03:19.240 --> 03:23.800 And that's all become possible because of natural language processing. 03:24.250 --> 03:27.910 So speech recognition is one of the very hot topic nowadays. 03:28.330 --> 03:31.900 And it is among the very large part of many of the products. 03:35.330 --> 03:36.440 Sentiment analysis. 03:36.470 --> 03:42.050 So in this particular cause, we will see implementation of this sentiment analysis with various different 03:42.050 --> 03:42.470 views. 03:43.520 --> 03:48.740 So finding some positiveness or a negativeness involved in attacks. 03:50.150 --> 03:55.900 And it will help us to check whether customers are satisfied with your services are not. 03:56.090 --> 04:03.990 So based on the reviews provided by the customer, we get the information like how satisfy a customer 04:03.990 --> 04:04.310 is. 04:05.880 --> 04:07.110 Question answering system. 04:08.580 --> 04:11.060 So Apple City, Google, OK? 04:11.510 --> 04:13.520 They're all state up. 04:13.560 --> 04:15.600 I question answering system. 04:16.200 --> 04:21.330 They communicate Vitas in terms of human language in their replies. 04:21.420 --> 04:23.610 Also in terms of human language. 04:23.970 --> 04:26.940 So they're kind of a virtual assistant for us. 04:27.420 --> 04:32.890 And this is one of the very heart applications nowadays in an LP field. 04:33.210 --> 04:37.920 A question answering system automated summarizes in. 04:39.270 --> 04:46.020 Nowadays, there are lots of tax available on Internet and we don't have time to read every single piece 04:46.020 --> 04:46.580 of tax. 04:46.890 --> 04:53.610 So this automated summarization will give you a summary in a very brief tome so that you will get a 04:53.670 --> 04:59.940 IDL, something like a 5000 word article into just maybe 200 word articles. 05:06.360 --> 05:06.630 Chad. 05:07.740 --> 05:12.900 So headboards are currently operate on several channels. 05:13.650 --> 05:15.050 It will be on Internet. 05:15.270 --> 05:18.150 It will be available on our mobile application. 05:18.180 --> 05:20.520 It is available on a messaging platform. 05:21.820 --> 05:29.980 And nowadays, many of the businesses are implementing those chatbot so that our virtual assistant will 05:29.980 --> 05:33.160 interact with us based on your cratty. 05:33.490 --> 05:36.440 He just cannot solve this query then only. 05:36.460 --> 05:41.860 It will be forwarded to a legal human text classification. 05:42.240 --> 05:42.550 Heider. 05:42.660 --> 05:44.890 Text is spam or a ham. 05:45.490 --> 05:46.390 Heider, Texas. 05:46.510 --> 05:46.870 Good. 05:47.170 --> 05:47.530 All right. 05:47.710 --> 05:48.370 We'll be right back. 05:48.790 --> 05:51.070 So that is also one of the main application of. 05:52.000 --> 05:52.630 An AP. 05:56.100 --> 05:57.270 Character recognition. 05:58.070 --> 06:05.820 So out of your image, what exactly a character is that is the character recognition, spell checking. 06:06.180 --> 06:12.990 So every time we do a Google search Google, we keep giving you suggestions that what you should search 06:13.330 --> 06:15.750 with your spelling correction also. 06:16.170 --> 06:21.330 So spell checking isn't one of the very important application of natural language processing. 06:21.780 --> 06:28.390 You can go to a best application available on Internet for spell checking like a family app. 06:29.730 --> 06:33.240 It has an ability to find all those grammatical check. 06:33.570 --> 06:35.520 Scan your all the text. 06:35.700 --> 06:38.910 Find all your mistakes, typos in a sentence. 06:39.180 --> 06:45.300 Is there any structural problems associated inside tax and many more things beyond that? 06:45.450 --> 06:47.760 It can detect out of your tax. 06:48.280 --> 06:48.660 All right. 06:48.690 --> 06:55.290 So that is all about the some of the few application I have listed, but many important application 06:55.740 --> 06:57.850 about the natural language processing fee. 06:58.460 --> 06:58.830 All right. 06:59.730 --> 07:00.810 See you in the next video. 07:00.810 --> 07:07.920 We'll get down to my installation part and straight into how to do some pre processing step with your 07:08.090 --> 07:08.760 tax data.