This post is a part of a series, as explained here.
As I already wrote, our workshop proposal for Applied Machine Learning Days was accpeted. We were certainly very happy and felt obliged to celebrate:
Below is the disclosure of what we promised to deliver (direct copy-paste of the proposal).
Workshop title: Machine Learning for News: Theory, Applications and Visualisation in Python
Duration: full-day
Workshop summary: The increasing availability and throughput of information flowing through social media outlets or internet-based news poses challenges for both the users and journalists. Users need to navigate through different news sources and, ideally, fact-check the content. Journalists face the challenge of summarizing often complicated and extensive data sets, and verifying that what is reported is indeed factual, all in a timely manner.
In this workshop, we want to equip you with the tools that will help you to quickly make sense of large streams of text and voice information, to both be better at reporting it and to defend yourself against misinformation. The visualisation tools and machine learning techniques we show can also be used by anyone, and journalists in particular, to deliver a better content.
This full-day course consists of two parts. In the morning, we will first provide a crash course on Python, and cover various packages for static and interactive visualizations. Then, we will give an introduction to machine learning in the context of natural language processing, and to neural networks, which will prepare you for the afternoon hands-on session.
In the afternoon, you will have the opportunity to explore the current real-life data set from WikiLeaks. Our project will incorporate hands-on tutorials for different use cases, such as text summarization, text retrieval and inference from knowledge bases. Finally, we will show you how to wrap the results in an intuitive user interface.