How to properly implement the hidden layer of the neural network?

Question

Python

Hello!
I am writing a simple neural network to recommend articles on a travel site. Each article is tied to a city and a country or just a country. The input layer of the neuron will contain the latest articles (the cities whose articles were liked by the user will have the value n (number) of like articles).
How can I make it so that in the middle layer the cities are generalized into countries and at the end I know whether the user should recommend this article. (I will sort articles with this neuron and give the user the most suitable ones)

PS This topic is recent, so feel free to get better if something is wrong

Answer 1

What's wrong?
"cities were generalized into countries" - what does the neural network have to do with it ???
"I will sort articles with this neuron and give the user the most suitable ones" - and what does the neural network have to do with it?

Answer 2

in your case, a neural network is not needed. because the features are obvious (city, country) and can be determined algorithmically (from city -> country).
tag all articles, and then just sum up the number of tags in articles liked by the user