What you need to know (and where to learn it)
In order to follow the Academy lessons, you will have to have some basic knowledge of the unix shells (Bash or similar). This is out of the scope of this tutorial, but a quick web search will get you all the information you need.
If you do not know where to begin with, try having look at this tutorial.
You will also need to have and know how to use a text editor which can save simple, non formatted text files. There are countless alternatives on every platform, just pick your favourite one.
Starting and using Grakn
From the directory of Grakn, run the command
./grakn server start to start Grakn. As soon as you see the Grakn logo, you will be able to connect to the Grakn.
There are several ways to interact with Grakn. We will mostly be using 3 of them: the Graph visualiser, the Dashboard Console and the Graql shell.
Let’s load some data before we start to interact with Grakn. You can download the data from here.
Unzip, go to the subfolder
short-training. From there, run
./load.sh PATH/TO/GRAKN-DIST academy, where
PATH/TO/GRAKN-DIST being the path to Grakn core.
This can take a moment.
To access the graph visualiser, with Grakn running in the virtual machine, simply point your favourite browser (Google Chrome is preferred to use Grakn) to
localhost:4567 and you will be presented with the following screen.
This is the Graph visualiser. In this mode, you can query your Grakn knowledge graph and the results will be presented as a graph. To try it do the following:
In the top right corner click on the
grakn tab (this is the Keyspace selector); a drop-down menu will appear. Click on
academy to select the Keyspace where our knowledge graph has been loaded
In the box on the top of the page (the Graql editor) write
match $x isa company; get; and return. The result of the query will be visualised in the main area of the page.
GOOD TO KNOW: Keyspaces
In the Grakn terminology, a keyspace is an isolated storage layer where you can store a knowledge graph. This way you can have several knowledge graphs in the same running instance of Grakn without them interacting and steping on each others’ toes.
The Graph visualiser is a good tool to visualise (you guessed it) the graph structure underlying your knowledge graph. It is useful to explore the graph, to find connection and to better understand the result of our reasoning engine thanks to the explanation facility (don’t worry if some of these terms sound obscure: they will become clearer during the rest of the course).
When the results to the query are too many, though, too much information is on your screen and the Graph visualiser becomes less useful. For this reason, there is a way to automatically limit the number of results that appear on your screen when you launch a query in the graph visualiser. You can choose anything you want, but 2 is a good number to start experimenting with.
GOOD TO KNOW: Clearing the graph.
In order to clear both the graph and the Graql editor in the dashboard, just SHIFT+Click on the (X) sign next to the editor.
When you have too many results or in general for some type of queries that do not return results in the form of a graph, it is not advisable (or in some case not even possible) to use the graph visualiser. In those cases, you need the Console. To access it click on the “Console” tab on the left of your screen and run again the same query as before to see how the result looks like.
In order to launch the Graql shell, from the command line, in the folder of Grakn core, run
./graql console -k academy. This will launch the Graql shell in the keyspace “academy”, that contains the data we are using during this course.
Once again, type
match $x isa company; get; and hit return to launch the query. As you can see the results looks very similar those in the dashboard console.
To quit the Graql shell, just type
exit to return to the command line.
Other entry points
The graph visualiser, the console and the Graql shell are not the only ways to interact with Grakn. There are lower level ways to control it programmatically, like the REST endpoints and the Java API, but those are out of the scope of this article. They are treated more in depth in the documentation that I highly encourage you to consult.
What you have learned?
If you have made it this far, you should have a working distribution of Grakn running in a virtual machine, and you should know how to access the main entry points to the system: you should know how to access and query Grakn from the Graph visualiser, the Dashboard console, and the Graql shell.
You are now ready to get your hands dirty and start working with Grakn.
In the next module you will start learning about Graql, the language of knowledge that is used to control the different components of the software stack. As you will see, it is very readable and easy to learn and use. If you want to have a look around before going on, head to the documentation portal