Data for Good Series: A Hands-on Tutorial on Using Machine Learning for Evidence Synthesis

Evidence synthesis is a critical aspect of research where various pieces of scientific literature are collected, analyzed, and synthesized to make informed decisions or draw conclusions. In this talk, Dr. Max Callaghan from Mercator Research Institute on Global Commons and Climate Change will provide a detailed walkthrough of how machine learning techniques can be applied to assist in this process.

The talk covers the following key aspects:

1. Retrieving Literature: The process begins with retrieving relevant scientific literature from bibliographic databases. This step involves searching and collecting a wide range of documents, such as research papers, reports, and articles, that are related to climate policies.

2. Building Classifiers: Machine learning plays a crucial role in this talk by enabling the creation of classifiers. These classifiers are designed to filter and annotate the collected documents automatically. They can categorize documents, assign relevant keywords or tags, and help in sorting and organizing the vast amount of data.

3. Displaying Results: The talk provides insights into how to visualize and display the results obtained from the machine learning classifiers. It covers the generation of static graphs, which can help in conveying important patterns and trends found in the literature. Visualizations are a powerful way to make complex data more understandable.

4. Interactive Website: Additionally, the speaker demonstrates the creation of an interactive website that allows users to explore the mapped literature. This website can be a user-friendly interface that enables easy navigation through the literature, searching for specific topics or keywords, and gaining a comprehensive overview of the available research.

Finger food and light refreshment will be provided for the event. 

About the speaker:

Max Callaghan is a researcher at the Mercator Research Institute on Global Commons and Climate Change in Berlin. He studies large collections of text about climate change using Natural Language Processing (NLP). A large part of his work is focussed on developing applications of machine learning that help researchers produce evidence synthesis about climate change. He also works with texts from social media or political debates in order to understand societal discourses around climate change. Currently, he is working on:

• The intersection of climate and health
• Climate impacts, especially in the context of vulnerability
• Policies and measures to mitigate climate change

He received his PhD from the University of Leeds in 2022.