4

I am working on a natural language processing data problem and I have selected some keywords from it as features. I want to group them under one heading. But I can't find any method or algorithm to do that? I tried topic modelling using MALLET. It groups similar words into a cluster. I want a heading or subject attached to them.

Is there a way to group similar words under one topic and assign them a title?

There are a number of different features that need to get one topic like picture quality camera selfie resolution that should come under camera. Need something that groups these things I enter words and it gives me the topic.

Dawny33
  • 8,226
  • 12
  • 47
  • 104
Maha Kamal
  • 101
  • 1
  • 3
  • 1
    Possible duplicate of [How to give name to topics created using LDA?](https://datascience.stackexchange.com/questions/9663/how-to-give-name-to-topics-created-using-lda) – Emre Sep 25 '17 at 18:14

1 Answers1

2

If you have already selected the keywords you want grouped, why not write a function that finds all occurrences of words in the list and replaces it with your one, core word?

What you're describing sounds like a more advanced version of stemming.

CalZ
  • 1,653
  • 6
  • 14