Portland Community College | Portland, Oregon Portland Community College

This content was published: October 8, 2018. Phone numbers, email addresses, and other information may have changed.

Is it time to try intelligent agents?

Posted by

It’s the beginning of your busy fall term, and you have so many things to do!

Wouldn’t it be great if you could get D2L to do some of your work for you? You can – by using intelligent agents.caricature of a sleuth with magnifying glass

Anyone who has completed the Online Instructor Orientation (OIO) since 2015 will remember the lesson about creating a no-show policy for your online course. One recommendation for your no-show policy is to email a student who hasn’t entered the course by Tuesday or Wednesday of week one. An intelligent agent can monitor a student’s activity in a course, detect when the student has NOT entered the course after a certain number of days, and then email the student with a customized message!

One caveat: setting up an intelligent agent is NOT an intuitive process – at least I did not find it to be intuitive, and I use D2L almost every day. Like a lot of D2L tools, the intelligent-agents tool has just enough options to overwhelm you, without providing a clear path to your destination. So let me be your guide.

Here’s a very simple example of one use of an intelligent agent, just to get you started. If you like the way the agent works, you can get really sophisticated and have the agent do all kinds of work for you (see “Further reading” below).

Email a student who hasn’t entered the class during week one

First of all, I’d recommend you try to follow the directions on D2L’s resource page, which will remain correct during the next D2L update.

When you reach the part about Agent Criteria, ignore everything except the Course Activity section. You’ll want to specify the number of days of inactivity before the agent kicks in. In my example (below), I have set one day, since I’d like D2L to email any student who hasn’t entered the class on the first day of term.

course activity setting in D2L intelligent agent

The next section – the “Actions” section – is where you get to decide how often to contact the student. You can email just once – the first time the condition is satisfied – or you can email the student repeatedly, depending on the schedule you set up (in the “Schedule” portion, which comes later). In this situation, I’d recommend you email just once – when the student hasn’t entered the course by your no-show deadline. The setting here is fairly straightforward:

actions for D2L intelligent agent

Now comes the really fun part of the “Actions” setting: you get to compose the email the student will receive!

The key here is to use what D2L calls “replace strings,” which are variable names in curly brackets. A common example is {firstname}, which tells D2L to search the database and find the user’s first name and replace the curly brackets with the user’s first name. Intelligent agents use a bunch of special “replace strings” such as {InitiatingUser}. This string tells D2L to find the name of every user who initiates the condition (doesn’t enter the class) and put that user’s email address in the “To” line. Super cool!

The other thing I highly recommend is to put YOUR OWN email address on the Cc line. This way you’ll get a copy of every email that goes to every student, which has several advantages:

  1. You can see which students are not keeping up, without having to consult the Classlist;
  2. You will have confirmation that the intelligent agent is doing its work. (You can get the same confirmation in the tool itself, but there’s nothing like seeing the email that goes to the student.)

The third and final setting, called “Scheduling,” was the most confusing for me, since I didn’t understand why I need to set a schedule for the agent to run when I’ve already said, in “Actions,” that I want the action to run “only the first time the agent’s criteria are satisfied for a user.” The thing to know about this section is that if you don’t enable a schedule, the agent won’t run unless you run it manually. I’m leery of setting a schedule that will run daily, so I’ve set the agent to run annually on Oct 2nd, the second day of term (I’ll need to remember to update this setting manually when I teach the course again next term).

schedule setting for D2L intelligent agent

This section allows you to hassle your students, since you can have the agent run daily, if you want. But I’d recommend against nagging students too frequently. If they are checking email, they will see your message.

Now that you’ve created your intelligent agent, you may well wonder: is this thing going to work?

Turns out that some smart person at D2L has come up with a “practice run” feature that allows you to see if the agent works:

practice run feature in D2L intelligent agent

The practice run won’t send actual emails to students, which would be disastrous, but it will provide a message telling you that the agent has run successfully. And if you set up your email to Cc yourself, you’ll get a nice email saying that the agent has run and in which course.

Give intelligent agents a try!

And if you’ve already tried using an intelligent agent in this way, I’d love to hear how it went for you and what you learned (leave a comment below).

Further reading

Andy explained, back in March 2016, how to create an agent that nags students who haven’t read your syllabus. Looks like the D2L interface has evolved a bit since then, but much of the explanation is still relevant and helpful.

The great Barry Dahl curated a wonderful collection of resources on intelligent agents in the Brightspace Community.