In the WP1, Simon Grant created the concept of Learning Episode, expression that we transformed into Digital Learning Episode to speak about what the dashboard was supposed to collect, analyze and put to good use (“valoriser”)… The DLE has become the central component of the MyLK project and we can read the following definition on the MyLK’s blog ( :

“A learning episode is a set of one or more periods of time during which a learner is engaged in learning. It can have a larger or smaller granularity: one greater learning episode may comprise several lesser learning episodes e.g., a MOOC may have a designed structure and outcomes, a playlist on YouTube, a video, etc. A learning episode is defined by the combination of the identity of the learner, and either or both of the learning outcomes identified, and the period or periods of time spent learning. For best definition, all three will be specified.

A future learning episode may not have firmly fixed dates and times. A past learning episode may not have the learning outcomes fully defined. Learning episodes range from formal experiences, as arranged and managed by a learning institution, as in a course of study leading to a qualification; to informal experiences documented and identified by a learner as having intended or actual learning outcomes.

A continuous learning episode is a single experience with no interruption. Examples of continuous learning episodes would be a lesson, a lecture, an experiment, a training session, or one session of a game. A composite learning episode is a set of learning episodes where the learning outcomes are related, whether by intention or by accident. Examples of composite learning episodes would be a course of study, a project, the creation of a creative work, a period of employment.

For a composite learning episode to exist, there must be some reason to believe that the learning outcomes of the constituent learning episodes are related. There could potentially be other aspects of relationship between the constituent episodes.”

It should be noted that a DIGITAL Learning episode is obviously a learning episode achievable through digital tools (online or offline)

To organize the development of MyLK platform – whose purpose is to scan and analyze a learner’s DLE to enable him/her to present or “benefit from” (“valoriser”) the Learning Outcomes and Competences (LOC) that he/she has developed thanks to this digital experience – we will have to clarify this definition to analyse a DLE so that it can be translated by computer.

The universe of MyLK’s DLE

Before beginning to decompose a DLE, it is important to specify the universe in which we work. The digital world is big today, it includes the Internet sphere but not only. When I work on a Word document for an internship report, we can also speak of DLE. Indeed, if we agree to say that my internship experience is not a DLE since my training is not related to the digital tool, the written formalization of this experience (verbalization, synthesizing, analysis) is also excellent learning experience facilitated, to some extent, by the digital tool. But, in the MyLK platform project, it’s difficult to hope being able to scan to this kind of episode. The reasons are many, but the first that we can mention is the place in which it was held DLE: this is the non-Internet digital sphere.

Indeed, both for reasons of feasibility but also of ethics and respect for private life, it seems complicated to analyze the DLE lived outside the Internet sphere.



  • First premise: to be relevant to MyLK, a DLE must happen in the Internet sphere



Then, it should be noted that until now when we quoted in the DLE example, they were all linked to an Internet resource. These are the most obvious learning episodes since it is fairly easy to target the learning itself and therefore the LOC developed during this learning that are more knowledge and skills acquired through the content of the resource as skills developed (even more so when the learner is in “consumer”, it is less true when it is in “producer / creator”). However, there are SLAs that develop skills and thus are more valuable on an educational plan. For example, when looking for work, puts in place procedures or a protocol of actions which improve over time since learns to look for work (eg typing the right keywords on the search engines, identify relevant ads, go seek the necessary information to prepare a résumé / cover letter / interview …). Conceptually and theoretically, it is indeed a DLE within the meaning of MyLK since it happened on the Internet. However, for technical reasons it is difficult, at least initially, to scan these procedures for their meaning on the MyLK platform. So we have to limit the monitoring and analysis of DLE to those directly related to resources and less to actions (eg watching a video, write an online article, read about Facebook …).



  • Second premise: under MyLK a DLE is necessarily “linked” to a digital resource that we call a DLR.



We must now set the DLR. We can consider that the DLR is made from three elements:

  • A content that includes a topic, theme, knowledge … We call it Digital Learning Content or DLC;
  • A medium (video, article, post …). We call it Digital Learning Medium or DLM
  • A platform or a provider that broadcasts the resource that we call Digital Learning Platform or Provider, in both cases we speak of DLP.


  • Third premise: under MyLK, a DLR is defined by the combination of its DLC, DLM and DLP.


So we said that DLE was “linked” to its DLR and we defined a DLR. However, the term “linked” is not satisfactory and must specify how a DLE is “linked” to DLR. This “link” is achieved through an activity can we call Digital Learning Activity. If we limit DLEs to those “linked ” to a DLR, the number of ways that a DLR can be linked to a DLE is limited.. Here is a list probably not exhaustive of these DLE made on a pedagogical level (DLAp):

  • watch (video)
  • make (a video)
  • read (an article, an e-book, post, etc.)
  • write (an article, an e-book, post, etc.)
  • do an exercise (MOOC, online tests, etc.)
  • discuss (on a forum or social networking)
  • share (via email or social network link to an article, video, etc.)
  • etc.

However, if these activities exist on an pedagogical field, how to translate them into computer language? We can give a possible translation by making a direct analogy between pedagogical action and its computer translation; eg “watch video” computationally match actions “click on the player / player turns pages”). But how to ensure that the person who clicks on the player and turns pages is actually watching the video? It is therefore assumed that a learner clicks the player and turns pages is watching the video. If there are particular cases where this assumption will prove false, we can consider that metadata, thanks to the theory of large numbers, this case is so marginal that it will not influence the statistics.


  • Fourth premise: under Mylk, pedagogical action that links the learner to a resource (DLAp) corresponds to its computer translation or DLAp = DLAc



Here is a synthetic scheme for defining a DLE in computer terms (that’s a smiling one !):

Next steps before starting the software development phase:


  • DLAc :


      • List all DLAp
      • Make a classification of these actions in terms of learning effectiveness (eg write is worth a read)
      • Make a table set DLAp / DLAc : we can start from
      • List the various data that we can collect related to these actions


  • DLM:


      • List the different existing DLM
      • List the various data that we can collect related to those DLM


  • DLP :


      • List the various existing DLP
      • List the various data that we can collect those related to DLP


  • DLC


    • List the various data that make up or define a DLC (theme, subject, author, critics, etc.)