Discovering Language Superpowers with NLP

Slide 3

Introduce learners to the ground rules, lesson norms, and team roles. If this is your first time conducting an Eddy lesson, check out 'How to set up an Eddy Classroom' for more information.

Slide 4

Start the discussion by asking learners whether they have been in a situation where Google auto-completes their search query by providing suggestions. Provide them with an example as shown below. Allow the learners to share personal experiences if they would like to. 

Slide 5

Engage in a fun activity by asking the learners to predict what Google would suggest for the following: Why is the world _________? Once the learners have responded, type it out on Google and look at the auto-suggestions. Check if any of the auto-suggestions matched with the learners’ responses. 

Slide 6

Ask the learners how they think Google comes up with these suggestions. 

Slide 7

Mention that Google is an artificial intelligence tool, which analyzes millions of responses to generate suggestions or makes meaning of the phrases provided by adding related words through Natural Language Processing. 

Slide 8

Introduce the concept of Natural Language Processing as mentioned in the deck.

Slide 9

Explain the concept of Natural Language Processing by drawing comparisons with a mind map. Just like how drawing a mind map is used to organize and represent information in an interconnected manner, NLP uses algorithms and models to analyze and interpret information for a computer. However, encourage the learners to think of ways in which a mind map is different from NLP, such as difference in purpose and output. 

Slide 10

Let the learners know that they will be playing a game called Semantris Explain to them that it is a word association game, which requires them to follow simple steps, which are shown below. 

Slide 11

Once they are done playing, ask a few reflection questions (mentioned in the slide).

Slide 12

After listening to their responses, explain briefly how NLP works by identifying a list of related words. Remind them of how Google generates meaningful automated suggestions by identifying related words. 

Slide 13

Mention that to identify related words or form meaningful sentences, NLP requires textual data or prompts, such as words, phrases, or incomplete sentences. 

Slide 14

Now, introduce the learners to their main mission, which requires them to help Tobi (their virtual friend from Pictoblox). Let them know that they will be using Natural Language Processing (a feature offered by Pictoblox) to help Tobi recommend a few movies based on their moods. 

Slide 15

Remind the learners of the ground rules. Help them form groups of three and tell them to carefully look at the demo of how to download and use Pictoblox before getting started with the activity. To understand how to download and use Pictoblox for this activity, access the tutorial for Pictoblox. 

Slide 16, 17

While dividing the learners into groups, assign them the roles of a project manager, resource manager, and a time manager. If this is your first time assigning roles in an Eddy classroom, visit 'How to set up an Eddy classroom' for more details.  

Slide 18

After providing them the demo for one category of movies (action-packed movies), let the learners know that they would be creating two more categories, which are ‘feel-good’ movies and ‘science-fiction’ movies. They will also identify moods related to these genres and use adjectives to describe their moods. Help them with the activity in case they get confused about how to use  Pictoblox. For activity specific instructions, refer to this tutorial. If the learners are confused about what movies to choose to train Tobi, provide them with the movie list

Slide 19

Once the learners are done with the activity, encourage them to reflect on their experience by asking the questions as shown in the deck. For the third question, the acceptable answers include Siri, Alexa, or any chatbot. Reiterate how Siri, Alexa, or chatbots use Natural Language Processing by making meanings of sentences and identifying related words. 

Slide 20

To reiterate that NLP algorithms engage in meaning-making by relying on grammatical rules just like we do, let them know that they will be doing a small activity called Word Weavers, where they will be arranging jumbled words to form meaningful sentences. You can use the Word Weavers Worksheet and provide a copy of the same to each learner.