5
$\begingroup$

My data exists of several restaurant menus (in text form). I would like to detect the category of an item on the menu which can be 'food' or 'drink'.

How do I call this kind of machine learning? Text classification?

I know javascript isn't ideal for this kind of thing but I would like to use brain.js or tensorflowjs to implement a proof of concept.

Thanks in advance!

$\endgroup$
1
  • 1
    $\begingroup$ This is a task for Natural Language Processing (NLP). $\endgroup$
    – Paul92
    Commented Jun 4, 2019 at 17:35

2 Answers 2

1
$\begingroup$

Menu is generally a list of items served in restaurant. Items broadly divided in to two sections:

  1. Food: Where all food items are listed
  2. Drinks: Where all drink items are listed.

If all restaurant menus that you have, listed items in similar way, I'm bit confused how you will use ML in this case.

But if restaurant menu includes recipe/ ingredients of the food/drink items, then yes I can apply machine learning in order to identify the category of an item. In such a case we can use NLP or simple text classifier(by dividing the recipe in some key word tokens and segregate between food and drink)model for prediction.

$\endgroup$
0
$\begingroup$

You can try to classify based on the presence of words in the name of the food/drink. For example, an item whose name includes "chicken" is probably going to be food, but if it includes "margarita" is probably going to be a drink (not necessarily, of course, because you might have items as "bloody mary with chicken broth" or "steak marinated in margarita", and that's where you will need your machine learning model for. Train your model in a comprehensive enough train set, it it might work.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.