Skip to main content

Using AI to Build Vocabulary: 3 Approaches

by Brent Warner |

Vocabulary can be tricky in the classroom and can often feel dull and repetitive for students. Though a lot of vocabulary-focused edtech has its place, many students have felt let down or frustrated by tech-based tools because they are overly rigid with definitions of words. This makes sense from a technical perspective: Students are expected to put in a synonym or definition, and if it matches the information in the database, they’re right. If it doesn’t match, they’re wrong.

Let’s imagine we want our students to show that they understand the word teacher. Traditionally, they would be asked to identify a picture of a teacher in comparison to pictures of people in other careers. Alternatively, they might be asked to provide a synonym: professor, instructor, lecturer, etc. All of these are fine, but the techniques rely on passive recognition of the pictures or synonyms provided, and research shows that students hold knowledge longer with active retrieval. Some edtech tools have dealt with this by providing a full thesaurus of options, and any one of the options would be considered “correct” in quizzing. In these cases, if a student chooses a word like faculty or scholar they would still be marked as correct.

But — what happens when a student’s creativity or way of thinking veers away from the predefined answers but is still very much correct in their definition? This is a great spot where AI is going to give students the opportunity to show what they know, not just that they can regurgitate information. Imagine now that a student defines teacher as “a person who works in a school and helps you learn.” In the classroom, any of us would accept that definition, but in terms of edtech, only with AI can students test themselves on their own and still get flexible enough comprehension to accept the wide variety of possible choices.

1. Broad Prompting

To give you a sense of broad prompts, you can check out “20 Quick Chatbot Prompts You Can Use With Your Multilingual Learners Today.” Note that the following prompt doesn’t provide any details on the type of quizzing, or other specifics. Students can be as loose or specific as you determine is necessary, and this “broad” example is very loose. This is the broad prompt I used to start a basic interaction:

I am an intermediate ESL student trying to learn new words. I'm going to provide you with a list of vocabulary words. Please quiz me on the words until I show an understanding of 80% of the words. Make sure that I understand 80% of the words on the list, not 80% of the questions you ask. You will continue until I am successful or until I close the chat. [PROVIDE LIST]

Depending on your chatbot, there may still be problems with accurate tracking of numbers (i.e., at what point the student hits 80%), but the bigger goal here is to allow flexibility in responses. Here’s a sample of the conversation I had:

Chatbot
What does "achievement" mean? Can you use it in a sentence?

Me
It means to get a goal. Passing this test is a big achievement

Chatbot
That's correct! "Achievement" refers to something accomplished, often by effort, skill, or courage. Your sentence perfectly illustrates its use. Well done!

Next question: What does "challenge" mean? Can you use it in a sentence?

Me
It means to try something. I will challenge this new soda.

Chatbot

You're close! "Challenge" can indeed involve trying something, but it specifically refers to a task or situation that tests someone's abilities. It often implies difficulty or the need to overcome obstacles.

A more accurate usage might be: "I will challenge myself to run a marathon next year."

2. Narrow Prompting

Though I like the idea of allowing chatbots to interpret your meaning and be more flexible, there are times when you need more precision in your vocabulary studies. In these cases, you can, of course, ask chatbots to provide much more focused training that more closely reflects the specific nuances of any vocabulary you’re working on. This is the narrow prompt I used to begin:

I am an intermediate ESL student trying to learn new words. I'm going to provide you with a list of vocabulary words. When you have the words, please create a paragraph that uses synonyms of the words in the list. Please bold and number the synonyms, then challenge me to find the word from the original list that matches the synonyms you provided. Make sure to mix the order of the words so they do not match the order I provide them to you. If you need to change the word form (for example, using the noun form of the synonym when the original is a verb), make sure you clarify the difference in your question. Please ask me about one word at a time, in the order they are presented in the paragraph. [PROVIDE LIST]

Here, I’m a lot more specific about the type of activity I want it to make, and my output will come only from the list I provided. Depending on my needs, I could have asked it to make a multiple-choice quiz, a cloze activity, or anything else that aligns with my outcomes for the day.

3. Customization

The final part of the prompting puzzle is simple but powerful. Have students include a sentence or two around their personal interests, favorite celebrities, school majors, hometowns, or anything else that they know and care about. Have them add a quick line to their prompts, like:

The paragraph should be about Superman

Or

Focus the content to talk about baking.

This short and simple addition placed into any prompt will let your students guide the conversation around their own interests, and, with a little luck, you may not even need to remind them to do the activities because they’ll be champing at the bit to get started!


If you start looking at ways that you can bring AI into your students’ vocabulary acquisition and combine them with Naashia Mohamed’s excellent post on scaffolding vocabulary learning, you’ll see that there are opportunities to increase all three of Nation’s “requirements of learning” for your students while better customizing the content and deepening their engagement.

Consider the way you can use these ideas as a springboard for your own customization around vocabulary learning in your classroom. If you’ve got tweaks, additions, variations, or altogether better approaches, please share in the comments — we’re all here to help each other!

About the author

Brent Warner

Brent Warner is a professor of ESL at Irvine Valley College in California, and an educational technology enthusiast. He is co-host of the DIESOL podcast, the only podcast with a specific focus on EdTech in ESL. He frequently presents on the crossroads of technology and language learning, focusing on student engagement and developing learner autonomy. Brent likes his coffee black and his oranges orange. He can be found on LinkedIn at @BrentGWarner.

comments powered by Disqus

This website uses cookies. A cookie is a small piece of code that gives your computer a unique identity, but it does not contain any information that allows us to identify you personally. For more information on how TESOL International Association uses cookies, please read our privacy policy. Most browsers automatically accept cookies, but if you prefer, you can opt out by changing your browser settings.