Not only since our conversation with Javiera Atenas and her keynote at the ILIAS conference in Lucerne we know that data needs to be "tamed". Thus, in e-learning it is important to clarify which data we want to collect at all and how we deal with this data in the long run.

ILIAS 6 now supports the e-learning standard xAPI for the first time. This standard facilitates the integration of external learning activities into your LMS. The focus is on the logging of user activities. However, xAPI is not known for data economy. This is because all learner data is to be collected on a central server - regardless of whether it is needed or not. The data collection therefore quickly becomes very extensive. The standard does not provide for the data to be deleted at a later date.

If you want to observe learners over a long period of time and collect learning status data from a variety of devices and applications, xAPI offers many interesting possibilities. Organizations who want to minimise problematic data accumulation, on the other hand, are unsure whether and how they should use xAPI. Fortunately ILIAS 6 offers an elegant solution for them. It allows you to use the many possibilities of xAPI without losing control of your data.

The technical background

As reported, xAPI is not only used in an LMS environment. Many other application scenarios are also conceivable – for example, in the form of learning games on a smartphone. All learning data that is generated in this process is stored centrally in a so-called Learning Record Store (LRS). ILIAS and other xAPI-compatible systems can access the LRS together.

There are already some good open source LRS that we can use for our xAPI application. One of them is called Learning Locker. It can be connected as standard in ILIAS 6. But its use also implies that data about learner activities is stored externally. What can an institution do to keep control over its own data?

LRS via proxy in ILIAS avoids unwanted data accumulation

In ILIAS the LRS is connected via a proxy. This proxy sits between the learning content provided in ILIAS and the LRS. In this way it pseudonymizes the data flow to the LRS if required. In the settings of ILIAS it is possible to prevent the sending of the name and e-mail address of learners. Instead a combination of user ID and a unique ILIAS platform ID can be used.

If you want to prevent user profiles from being created across multiple objects based on the user ID, you can also generate an encrypted ID for pseudonymization. This ID is composed of the user ID and the respective object ID. In a next version of the proxy it is also planned to further reduce the activity data sent to the LRS.

The problem of data deletion remains: Since this is not provided for in the xAPI standard, data would theoretically remain in the LRS forever. Fortunately there is a solution for this as well. Learning Locker allows to create different LRS types - which can be deleted again. It is therefore possible to use different LRS types for different user groups or contents in ILIAS and to delete them when the data is no longer needed.

If an LRS type is deleted, the corresponding data is still retained in the learning progress of ILIAS. If, on the other hand, data is deleted exclusively in ILIAS, the data on the LRS can no longer be assigned to a person due to pseudonymisation.

Thanks to this innovative implementation, the use of xAPI in ILIAS is possible without any problems - even at institutions and companies that want to minimise data accumulation.