 next library is called NLP Toolkit. NLP stands for Natural Language Processing. Basically its purpose is that when you talk to a lot of social media tools, you write emails, you type, or even speech recognition, we write in a natural way. Even if you are a native English speaker or not, the language in which you speak or you write, and just keep in mind that what we are talking at this time, we are not talking about English grammar, we are not talking about some academic writing. We are talking about the data available to us, we have to process it. Basically we need its toolkit for that. Now as I told you, whenever you write an email or a letter, you go and comment on Facebook or on any other social media platform, when you comment on it, you go to your V-Logs or V-Logs, you type in that, that is natural. For that you have to think a lot about grammar, how to use it, how to use it, how to use it, whatever is your natural flow, you write in it. So basically this is our library, its purpose is that you use it, and in two situations it is basically used. A text data, as I said to analyze the text, the analysis of the text is very critical again. The data is totally unstructured, you don't even know what it means, what it means to write, what it means, what it wants to say, and what it will understand when the person will read it. So this is the biggest problem in communication. The intention of sending a message, and the understanding of reading the message, these two things can be totally different. Basically the different situations we have, the social media is being used by everyone, a lot of people use it, which are involved in negative social activities, which can be a risk for your society, and a lot of people do good work, there are a lot of different social setup, people stay in it, they use their different social media. So there are two very important subjects, one is called psychometric analysis. The person who is writing, the type he is talking about, you analyze it in similar ways, what he really wanted to do, what he really wanted to achieve, what message he is giving to someone, some message can be your code, the code he has used, so the psychometric analysis of that data is very important, and this is not only in this, but it is also used in call center performance, that they have a different solution available, that you convert the voice recognition data into text, and perform analysis on it. Similarly when your unstructured data is there, you use natural language processing tool kit. For this, the different features, like I have told you, for learning we use it, we want to learn about what someone has written, what someone has said, we have to learn it, then we have to understand it, because you have read one thing, you have written on it, you have written on it, you have written on it, so what have you understood from it? So in the first step, you have seen what is written, then you have to understand it, so that you can understand the receiver of the message, or your system basically, as a data scientist, who has to perform analysis on it, so after that, the content, your whole content, what was it, then how you have to reproduce it, when it is produced, what is it, because on it, there are all psychological models available, the way we have, that is also a model of statistics, similarly there are models of psychology, and that is also in statistical forms, A, B, C, X, Y, Z, it is not an imaginary thing, and when you collect the data, then you can apply the model on it, and based on these models, even your recruitment, when there are big firms, when they hire their staff, they do their analysis, they see what their personality type is, what kind of person is fit for this type of role, there are many many applications, where we can use this type of NLP tool kit, of course, I mean this is no brain, analysis and visualization, whatever you have to do, to collect your data, on it, you have to perform these two functions, otherwise you are doing something different, these are some of the steps in it, which are very important to understand, tokenizing is basically what we did, during one of the other mapping, so tokenizing is basically what we have to do, the different words, we assign them to token, first I assign token, this is minus 10, this is plus 10, now the different words, if we have to do first, that is love, it means it is a very very positive word, it might be somewhere here, that its number is plus 7, its token is its ranking is plus 7, but if it is hate, it might be somewhere here, some words, first one word is going, or go, it doesn't mean anything, it doesn't mean negative, it is positive, it can be here, so basically you tokenize it, then you see that you have to filter the words, you see what your result comes out of it, then you see that the different speech, the text, the different parts of it, then you search them, then you relate them with each other, if in the beginning someone has written in the first part of the content, in the second part and in the end, then all these things, when you analyze the words and the different ones, then you correlate them with each other, then you create your own decision, or you create your own opinion using some of the psychological model, then you say that the people who have written the comments, their comments are risky, suppose, or there is a negative impact in it, that it is blocked, or some work is done on it, or it should be given to the state departments, or some action is taken on it, so this is very very important, you can easily understand, that your natural language processing is very important, especially in the era in which we are spending our lives, and in the coming time, as you will be going to your practical life, the importance of these things, will increase, similarly the associated risk will also increase, similarly NLP has some other important features or functions, you can use them during your analysis, like named entities, this named entity can be the name of a city, or a building, it can refer to anything, but as it is known, you can understand that named entity it can be a known, represented by a known, what is its name, what is its share, what is its human being, similarly when you have analysed it, determined it, then you have analysed the text, in the last slide, after giving the weightage of tokenisation, that it is negative, it is positive, it is neutral, it is so negative, it is so positive, neutral is neutral, so based on this, you analyse the text, you use the tokenisation, the other features we discussed, the link is based on that, then this is important, this is also okay, but the dispersion plots, when you have a text message, a conversation, based on that, you can determine how the conversation is going, how the conversation is going, whether it is neutral, whether it is a special event, whether it is a special place, or this is just for the sake of discussion, or there is a festive season, first of all, there is Eid, Christmas, there is this kind of occasion, in that people talk about other things, for God's sake, there is a problem, there is a flood, there is a war-like situation, the sentiments that are exchanged, they are different, so basically, when you plot it, then you can easily understand, and when it is plotted, you can divert into some suggestions, perspective analysis, which we say, that you did a forecast, just what we had, different types of analysis, that you have the second data, similarly you have the text data, you have processed it, you have plotted it, and based on that, then you have a statistical model, based on psychological variables, and then it will tell you, how this thing is going, you will forecast something, and when that forecast comes to you, based on that, using some preventive measures, you can prevent it, facilitate it, or do something else, so this is the whole process, the way you have your column and row on the table, similarly your text data, or unstructured data, you can do your analysis, so the different starting from descriptive analysis, then you can do forecasting and perspective, you can do everything based on it,