Artificial Intelligence Meets eLearning
On a recent project, we needed to develop a customer service eLearning course that could act as a live customer, i.e. accept natural language statements, process them, and respond accordingly. If you’ve ever developed a conversation-based eLearning course in the past, you probably created a limited set of questions and corresponding responses, each one branching out to a different slide/layer based on the selected option. While this approach works in many situations, it significantly limits what the learner can “say” to a customer in the training. Here, because the learner should be allowed to freely structure her statements, the branching approach cannot be used. In this article we will describe the approach we took to create the module in Articulate Storyline 360 and link it to the AI engine. If you wish to learn more about the project goals, you can read the AI in eLearning Case Study.
To solve the problem at hand, we turned to Artificial Intelligence. We decided to use ClueLabs eLearning AI engine to process the entries and provide appropriate responses. Other features that were important here included:
- Tracking whether the learner sticks to the script required by the company policy
- Determining whether a particular step of the conversation was executed on time, early or late
- Identifying irrelevant and duplicate questions/responses
- Customized feedback for each step of the conversation
- Recording all conversations for management to review and utilize in further coaching
- Using data collected by AI to predict the outcomes of the training
- Training the AI algorithm to improve understanding based on the collected data
Step 1 – Prepare the assets
Import the graphics and lay out all elements such as text and buttons on the slides.
Step 3 – Set the variables
Use the SetVar and GetVar functions to set the player variables before the response is processed.
Step 4 – Process the response from AI
Step 5 – Add triggers to show checkboxes when the steps of the conversation are executed.
For each required step in the conversation process, add a trigger to show the checkmark when the step is executed.
Step 6 – Add a trigger to take the learner to the feedback slide when End Conversation is clicked.
When the learner decides to end the conversation, she will be taken to the next slide to receive the feedback.
Step 7 – Create states for each step.
For visual feedback, each step will have 3 states: green=executed on time, red=missed, yellow=executed but not on time.
Step 8 – Create states for description labels.
Add simple shapes with text to display descriptions for each step.
Step 9 – Create text feedback for each step.
For each feedback text box, create 3 states with different text based on the conversation outcome.
Step 10 – Use Storyline triggers to display the correct states.
Based on the data received from AI engine, display the states of the objects corresponding with the outcome.
Step 11 – Have another satisfied customer.
In summary, artificial intelligence in eLearning has an excellent potential to improve learning outcomes and create engaging learning experiences. What’s really amazing is that we can keep using the tools we know and love (such as Articulate Storyline) to create such experiences. Feel free to contact eLearning Company if you need help integrating advanced features such as AI, data analytics, natural speech processing, data insights, etc. in your eLearning products.