Business Unit: VisionWorks (Big Data & Cloud)
Title: Sentiment Analysis of (Bi-)Monthly Journal De Journalist
Technology: Python, NLP, Sentiment Analysis

Context & Concept:

Perform research into the tone and evolution of articles in the magazine ‘De Journalist’, published by ‘Vlaamse Vereniging van Journalisten’(VVJ). They will deliver the data in pdf’s(around ten years, monthly to two-monthly editions).

You will focus on the interviews and opinion columns. The results should make a clear distinction between these types or articles.

Goal
You will develop processes in Python for:
• Processing and recognizing the content of the pdf’s
• Storing and providing data in such a way that makes ad hoc requests and visualizations easier

You will conduct a sentiment analysis based on the articles, whilst focusing on the articles and opinion columns, and will derive answers and assumptions on the following questions:
• Are the interviewed people either positive or negative about journalism and media?
• Can we mark an article as ‘rather positive’ or ‘rather negative’?
• Are there editions in which there are more negative than positive articles?
• How do the different editions compare to each other? Is there an evolution over the years, are there specific phenomena linked to time or events?

• Candidates should be following an IT-related study. Knowledge of Python is a plus.
• The internship will take place in our offices in Mechelen. Candidates must arrange their own transport
• Candidates can work autonomously without daily guidance