Comment Analyzer

Drupal as a Content Management Framework supports commenting content out of the box. In the free and hanseatic city of hamburg we have a Drupal based system for collecting informal civil contributions. To collect the data we have a Drupal Backend and a custom decoupled frontend implemented with Vue.js. 

The content created by the citizens of the city can range from a hundred entries or more. Each of these contents as well as the comments are analyzed manually by one or more clerks. To simplify the process of comment analysis we have developed a Drupal module: comment-analyzer.

This module shows the administrator important information that helps him to analyze the comments correctly. 

The presentation will show how we proceeded to develop our analysis model, which Drupal features like the Batch API we used, how the integration with different open source frameworks worked and an brief introduction to Natural Language Processing (NLP). 

The purpose of this presentation is to demonstrate how NLP can be used to analyze comments on a Drupal site. We will discuss the different techniques that can be used, such as sentiment analysis, topic modeling, and text classification. We will also showcase the benefits of using NLP especially to make the free and hanseatic city smarter.

In conclusion, this presentation will be a great opportunity for Drupal developers and business owners to learn about the benefits of using NLP to analyze user comments. Attendees will leave the conference with a better understanding of how NLP can be used for e. g. to improve user engagement, detect spam comments, and understand user sentiment.