Supervised learning on sources of information with different importance

I am trying to classify customer support sessions using supervised machine learning.

In each customer support session I have 3 bags of information. 1. The title of the customer's complaint 2. Information about the device the customer was using 3. Text of the chat session with the customer support agent

In each customer support session, there are 6 different classes. Is it better to: 1. Train a classifier on each bag of information and have them vote on which class the session belongs to? 2. Put all of the information into a single set of features, and train a single classifier to determine which class the session belongs to? 3. Other?

Topic aggregation feature-selection nlp machine-learning

Category Data Science

1
answered at

Its better to do supervised classification then unsupervised clustering since you have the right labels for the response.

Now in supervised classification approach 1 is better than approach 2 because in approach 2 you will the reducing the already available information by combining features.

Approach 3 is advised where you will create more features from the given 3 features by use of NLP processing, sentiments, tags, chunking, time taken, text length etc and then selecting the features those provide more information gain using feature selection approaches.

About

Geeks Mental is a community that publishes articles and tutorials about Web, Android, Data Science, new techniques and Linux security.