 In this module we will talk of the different document types or the data structures which are supported by NoSQL and what are the products. Now remember or try to understand that in the context of NoSQL we are working with those kind of data sets or which have a structure changes. The structure changes over a period of time and the relationships are not defined. Now if you tell this to a relational database architect they will just raise their hands. They will say this is not possible. Of course relational model was not meant for this kind of data structures. That is why we have NoSQL. So let's look into more details. So what we have over here are different types of those data structures. So if you are wondering whether NoSQL is just a niche solution or an increasingly mainstream one the answer lies in this and subsequent slides and modules. So it's time to talk about recent trends and how you can use NoSQL databases over and above the traditional RDBMS approach. So over here we have this triple graph store over here which we will discuss and we have this key value over here and this document we will discuss in the next module. I think it's module number 144 okay. So there is a lot of things there are a lot of interesting stuff over here and let's look into more detail. Instead of storing data in a row for fast access data is organized for fast column operations. This column centric view makes column stores ideal for running aggregate functions. You see that all the column is over here. Column store each occurred think row in a RDBMS environment doesn't require a single value per column. Multiple values are here. Multiple values are here. Instead it's possible to model column families. For one of these column families consist of several fields. One of these column families may have multiple rows also. So this is kind of hard to understand or digest by someone from the domain of the relational model but this is what we have in the context of the no sequel. So what is the advantage of this columnar approach? You don't require always the fields right because it's in real life all the fields may not be available. So there is no blank padding right as opposed to a relational model and that has a result of storing the data in less space saving in space. And of course with a single key I can read multiple records through the column and there is no join. That is the biggest advantage over here that there is no join. There's no join is needed because join takes time. So let's look at the other key value store. So I have this keys over here okay and the beauty of this approach is I have this keys for this user and this user they are same but I also have this user over here. So I have arbitrary set of data okay. Of course I can have for example the salary over here. Salary can be okay say for example what is the salary. So I can have the salary over here so I can have arbitrary set of data and it can be integers it can be text and the benefit of this approach is very high query. High speed why the speed is higher the reason being that I just access this and directly go over there. So again there is no database access because whenever there is database access I have to go and run a query and maybe perform a join with this approach I don't have to do all of this. Let's move ahead triple and graph stores so I have this subject and I have this predicate predicate is the property and property of what and this is the value of the property. So with this triple approach I can do those queries and I can perform those relationships which is not possible in a relational model and for which I have to do a lot of work over here in a triple graph store I can develop quite complex relationships which I can show you in the next slide and I can run complex queries over those relationships which is not sequel which is not sequel let's look at that example also. So I have these relationships together so from one triple okay from one triple now I have this a whole graph this is a directed graph I have this directed graph and I can add lot of information in it and I can also query this directed graph also and I can find the relationships for example like this relationship over here I can discover these relationships I can discover these relationships over here which are quite difficult to implement and then query in a conventional RDB MS model. So this is all I have for this module thank you very much.