Welcome to the first video of the Annif online hands-on tutorial. This tutorial has been created by the National Library of Finland together with the Leibniz Information Centre for Economics ZBW.
Here is what you need to do in order to complete this hands-on tutorial.
First, you should already be familiar with what Annif is and how it can be used for automated subject indexing. If you haven’t already done so, you should check out the Annif website at annif.org, watch a presentation about it, and if you’re interested in more details, you can also read an article about Annif in the open access journal JLIS.it.
This hands-on part consists of short videos like this one as well as practical exercises. To complete the exercises, you will need to install Annif on your own computer: the installation instructions are given in the first exercise of this tutorial. The idea is that you will complete the exercises on your own time, on a schedule that suits you. Try to get as far as you can with the exercises. If you get stuck on something, you can still at least watch the videos and study the exercises for the remainder of the tutorial.
From time to time, we will arrange interactive online sessions. These sessions are an opportunity to ask questions, to get help, and to discuss what you’ve learned. You have to register separately for each session and the number of participants is limited for practical reasons.
All the materials of the tutorial can be found in the Annif-tutorial GitHub repository. There you can also find the schedule and registration information for the interactive sessions.
Here is an overview of the exercises in this tutorial. The ones that are shown with a blue background are core exercises that are necessary for completing the tutorial. The ones with a white background are optional exercises that you can complete if you’re interested in the topic. We provide videos for each of the core exercises but not the optional ones. Some of the exercises also have sections marked EXTRA that you can complete if you wish, or skip if you're in a hurry.
When you start the exercises, you need to decide how to install Annif on your own computer. We provide three options for this. The first option, which is also the recommended way to install Annif for most people, is to use the pre-built VirtualBox image, which contains all the software you need as well as the data sets used in this tutorial. This option requires that you first install the VirtualBox software on your computer. It is free and available for most common operating systems.
The second option is to use the pre-built Docker images. This is mainly an option if you know how to use Docker to run software within containers. It’s a convenient way to get Annif installed with all the dependencies included. You will of course need to install Docker software. It is also available for most common operating systems.
The third option is to use a local Linux install. This is an option for experienced Linux users who are used to working with Python packages. This involves a few more steps than the two other installations to get all the dependencies set up, but it allows for maximum flexibility. Any general purpose Linux distribution should work, but you need to have a supported version of Python.
In this tutorial, we will be accessing Annif in multiple ways. Many of the operations are done using a command line interface: it is used for setting up, training and evaluating models as well as bulk indexing. We will also look at the simple web interface that Annif provides for interactive testing. There is also an optional exercise about the REST API for those who are interested in integrating Annif with other systems.
For people who are not so familiar with working on the command line, we have prepared a short introduction to the most important commands. You can find the link to it on the exercise overview page.
In this tutorial you will work with the example data sets that are provided. After completing this tutorial, you should be well prepared also for applying Annif also on your own data, as long as you are able to convert the data you have into a format suitable for Annif.
You have now been introduced to the hands-on tutorial. Next, you have to decide which way you want to install Annif so you can start working on the exercises. Thank you for watching.