Page 18 - AI

Activity 10

Review it with AI

When you use social media or apps, you must have seen that people write
comments and reviews on various topics, products and services. These reviews
are very helpful while making decisions, right?

Sometimes, these reviews can run into thousands and the text may be too long to
read and make sense of. Not to mention, they can be in various languages. How
to make a decision in such a case?

Can you imagine how AI can help in understanding these reviews?

Remember, we studied about Natural Language Processing (NLP) in Chapter 2.
NLP can be used to make sense of these reviews and understand them. We can
use NLP on a text to classify it as positive, neutral or negative, based on words

that might appear in the text. This is sometimes called sentiment analysis.
Provide a range of online reviews for students to view and decide if they are positive,
neutral or negative. What words provide students with that overall impression?

Use the table below and classify these words into positive, neutral or negative.

Great Hate Like Okay

Not Bad Fantastic Don't Like Best
Worst Boring Exciting Wonderful
Unusable All right Amazing Dependable

Ordinary Interesting No good Impressive

Now, classify your analysis and put it in the table below:

Positive Neutral Negative

Discuss your results with your friends and see if they match. You must understand
that sentiment analysis is not always reliable. Mostly while training a dataset in
AI, words and phrases would be classified as positive, neutral or negative through

supervised learning. You must remember that Supervised learning is the process
providing AI with a lot of labelled training data for training it.

ArtificiAl intelligence 17